SEO for eCommerce – The Beginners Guide to Keyword Research

SEO is a complex science, however its principle is simple: if you want as many people to find you organically as possible, which means simply from match pages of search engines, you must tell the search engine what your site is about.


However simple this may seem, for several reasons it is not.


This post’s primary focus: You need to understand how your audience searches, what words they use when searching your products, what they search when you may have a chance to get them shopping.

Once you have understood this, it is time to apply search engine optimization on your website, communication and advertising campaign based on this.


We will guide you through the following steps:

  • Why is it essential to find your keywords?
  • Basic concepts
  • Place of keywords with Google today
  • Step Zero: You must learn about your customers
  • …and you must also learn about your competition
  • How to make lists?
  • How to expand your list?
  • Learn from your competition
  • Refine your list
    • Number of searches
    • Relevance
    • Purchase tendency
    • Competition
  • Watch your architecture as well
  • Do not expect immediate success


Why is it essential to find your keywords?


If you search something in Google, the algorithm usually has to sort out hundreds of millions of sites what to show you.

The winner is always the one with the best position. Over half of the clicks are divided among the first three matches, one third of the clicks go to the matches shown in the first result page.

It is said that you might as well hide corpses in the second result page of Google. Only the desperate dare to wander there.

The reason for this is that years of sophisticated development and hundreds of indexes have led to the algorithm being extremely precise in determining which sites have real value, which sites actually have the content the user would find useful.


Of course, search engine optimization is not only about setting the keywords – if you wish to get a good position, the following are equally important:

  • your site should be fast,
  • you should have quality content,
  • your system should not break down by a bit of overload, etc.


In case of a car it is also not only about the wheels to be able to speed on the highway. The engine is also important, the fuel and that the car is streamlined. But one thing is sure: you are going nowhere without the wheels.

Just like without the right keywords.


Basic concepts


Before getting deeper into more complex issues, let’s discuss what the most important basic concepts of search engine optimization exactly mean.




Keyword : In search engine optimization, keyword means a word or expression that briefly characterizes the content of a website. The keywords are included in the metadata of the website, that is how webmasters help the search algorithms in matching the sites to the appropriate searches.


Long tail keywords

These are keywords that are very rarely searched (e.g. on a monthly basis), they are mostly special long expressions or even complete questions. The reason they have an individual name is that in search engine optimization they have a huge role. Searching just one word, the user has a much smaller chance of getting an exact match, therefore most people enter multiple words at the same time (for example: “womens black jeans”) or they set specific questions or sentences (“How much does a pair of black jeans cost?”).

How important is long tail? If only your matches appeared in all existing search engines to the most popular 1000 keywords, around 89.4% of the traffic running through search engines would still not reach you. This means that a vast number of searches are made up of long tail keywords, based on the principle: ‘Look after the cents and the euros will look after themselves’.


Search traffic


Search traffic : shows how many people searched one specific keyword in one month. It is advised to target keywords which are often searched – the problem, of course, is that you’ll face tough competition, as all your market competition will target the popular expressions. In this case it is recommended to target long tail keywords with medium or low traffic, but can be targeted very precisely, therefore you’ll have a larger chance for success because of the smaller competition – we will discuss this later.



Speaking of which, what exactly does competition mean in respect to SEO? Of course, all sites you wish to beat by optimization. You probably have no chance to appear on the first page with simple words like “book”, “clothes” or “car”.

The top of the results pages are usually “granted” to companies who have already spent enormous time and effort to keep their positions. It is extremely difficult to compete with their resources, and it is also needless: there are more cost-effective solutions than this.


Place of keywords with Google today


A couple of years ago SEO was much simpler: you took the appropriate keywords, placed them on your site, and you were more or less done. The algorithms watched fewer variables, therefore it was a widespread solution to apply keyword-clouds (keyword-stacks) making the website basically unusable and ugly.

This has obviously led to the user experience not being very high. If you searched an expression, you did not get the best matches for your page, but for the one that was better ‘optimized’. It was possible that you could hardly find any really useful content on the first couple of search results pages.

This is why Google and other search engines have changed their search methods, and have come up with much more complicated solutions.

Nowadays, they primarily watch how the users behave. If a website is basically well-optimized for an expression, it does not necessarily guarantee a good position. It is also essential that the users’ behaviour reveals that it is really useful what they have found after clicking the link. They simply get what they expected.

Of course, keywords are inevitable parts of the whole process, this is where it all starts from.

Since the Hummingbird update, the intention of Google is not only to try and give good matches, but to find out our objectives during the search.


If we type in “dollar euro”, we can immediately see the current exchange rate.

If we type in a mathematical formula in the search field, we get a calculator with the result shown.

If we insert a geographical name, Google Maps will be at the top of the list.


Step Zero: You must learn about your customers


In order to find the right keywords, you must dig deep into the communication of your target market.

Random brainstorming can also be fruitful, but researching simplifies your work, and you can learn about ideas that you would never have thought about yourself.

First of all, you must set segments interesting for you.

The segment is relatively simple, it is revealed by your experience and market survey. 40-50-year-old men working as office employees is a segment, for instance. Within this, you must create ‘buyer personas’.


Buyer persona, in other words the sample customer, is a semi-fictitious personality: has the most characteristic statistic features of the given segment. You give it a name, set the age, marital status, number of children, what he/she does in his/her free time, what content he/she consumes and so on.


After this, you must observe how each of your buyer personas communicate. You must find the places where they speak to each other, or even to experts like you, and you must pay attention to them.

Read the incoming customer emails, visit online forums, professional groups on social portals. Go to conferences, events where you can find your target audience.


Pay attention to how they communicate.

  • What words do they use?
  • What style do they use?
  • How do they ask questions?


…and you must also learn about your competition


The way your target audience speaks about a product is not the only important thing, as it might happen that they will search differently. This is why you must also watch how your competition speaks about products like yours. What expressions do they try to plant in common knowledge? What words do they focus on?

How do they refer to it offline? What is said in ads, on billboards, what do sales representatives say about the products? All this affects how your potential customer will search for your product.

As a result, there is the phenomenon that two-thirds of those searching for a product because of an advertisement end up not buying from the one who originally paid for the advertising. Others are better at optimizing on the net, and ‘by-pass’ the traffic.

tips Example : Imagine, you pay millions for a TV spot (by which you can reach your relevant audience only at a price of an enormous waste coverage), the two-thirds of those searching on the net for your product as a result of your commercial are dragged away by your competition – only because they were better at optimization.


How to make lists?


Take a seat with your team (if any), or with professional acquaintances of yours, and start a brainstorming session.

Write down each expression that comes to your mind that the users may use relevant to your site.

  • What search expressions do they type in when considering your products?
  • What are they interested in when they are actually searching for your content?
  • How do they put it in words?


Do not waste time with single terms because of the competition mentioned above. Target expressions compiled of at least two words. Use long tail keywords as well, e.g. you can put 5 or more words in what you think your target audience will type in the search bar.


How to expand your list?


The easiest way is to see what Google recommends for you. Take your list and search everything you have listed. If you scroll down to the bottom of the match page, you will find the related searches, among which many can be useful for you.

Of course, you can use smarter tools than that. One of these is Übersuggest, which recommends potentially effective search expressions with the help of Google. The method is simple: the service takes the given keyword, then runs through all the letters in the alphabet to see what permutations most often occur in searches.

To make it simple: for a given keyword you get all search expressions that include it and are most often used.

The most popular solution, of course, is using Keyword Planner.

It gives you the possibility to check how many searches are initiated regarding certain expressions from your compiled keyword list in a month, how big the competition is, meaning how useful they may be for you. Besides, the tool also shows you related relevant expressions – ones which are more often searched, or where the competition is smaller, or perhaps simultaneously both.

When using the tool, you must first create an AdWords account. After signing in, you can find Keyword Planner under Tools. Once you open it, you must choose the first option, which is ‘Search for new keywords using a phrase, website or category’.


keyword planner google adwords


After filling out Targeting, add each keyword from your previously compiled list, then click ‘Get ideas’.

On the next screen, below the graph you will find the necessary information on the ‘Keyword ideas’ tab. (If you are not a paying AdWords user, you will only see numbers of scale.)


keyword planner research adwords


In the first column you will see the searched phrase and the relevant search engine expressions, in the second you will find how many searches have been initiated regarding the given phrase in the geographical location you have set, according to the language target, within a month.

You can refine your filter by only showing relevant terms where the competition is low, meaning where you have better chances to dominate the market. From this list you can easily select the keywords which best describe your website’s content.

However, you should be careful using the tool: experts say Google tends to glorify the number of searches initiated to the keywords upwards. The reason the system works this way is that it is the interest of the search engine company to illustrate more and more phrases favourably, as AdWords customers will only pay for advertisements on the given keyword if they see it has high potential. The system does not explicitly lie, it just does not always show the exact truth.


Do you have a search bar on your website?

If so, it is splendid, use it. Watch what words and phrases people use, who have already found you but have not quite found in your site what they are looking for.

Why is this useful? As users we all have experience about how various search engines work. Those placed on websites are by far not as intuitive as Google’s algorithm, so you want to be as specific as possible. So that the simpler search engine will be able to come up with the relevant subpage.

In other words: in the search bar you should enter what you think will help to find what you are looking for. This way your users will tell you how to optimize.

Use Wikipedia

At first it might sound strange, but I will explain what I mean.

Wikipedia pages perform awesomely in search engines, if there is a page related to the given keyword, it will most probably be within the first three matches. You can find loads of information on these sites. Related names, concepts, events, that is everything that might be relevant in connection with your original keyword.

Look at the entry ‘London’, for instance. Here, you will find a number of phrases, like:

  • Capital of the United Kingdom
  • River Thames
  • Great Britain
  • Londinium
  • Tower of London
  • cultural centre, etc.

If you want to optimize a website in terms of tourism, and already you set your buyer personas, you now have gained a good number of perfectly relevant search phrases.


Use Pinterest as well

You might have read our article about how to use Pinterest in ecommerce. Now we will see how it helps search engine optimization.

You can do this with the help of Promoted Pins: simply start planning a campaign, enter one of your main keywords and a relevant phrase, and see how Pinterest will flood you with suggestions. These will mainly be long tail phrases that are recommended for you based on user habits.


Learn from your competitors ‒ the power of product categorization


Visit online stores offering products similar to yours – preferably the ones with higher success. Examine how they categorize their goods.

You do not need to worry about main categories, as they contain keywords where you probably have no chance of getting a high rank, unless you sell in a niche market.

You should rather look at what sub-categories, or sub-sub-categories they have in their online store.

And since customers usually search using category keywords rather than specific product or brand names, there is a gold mine here to discover!

For example, in the end you get a route, something like: Laptop -> Acer -> Business

You can get brilliant ideas from routes like this, as your last category site search phrase in this example would be “business acer laptop”.


Don’t stop here if your product is somehow special, try to include it in your search phrases. Is the food organic? Is the car eco-friendly? Hand made, imported? If you include these expressions to your long tail phrases, you will have a much smaller competition to fight.


Refine your list


After you have done the necessary brainstorming and research, you have a good chance of still having dozens of keywords.

For optimization, you should determine the most relevant ones for you, about 5-7 items. Also maintain a bigger list with two-three dozens of keywords, which might also ‘come handy’ – the ones to which you can produce useful content, to which you can optimize your subpages, and so on.

You must observe several aspects of whether a keyword is really valuable for you.


1) Number of searches

You don’t want to optimize for keywords which are only searched two-three times a month. This is the way to count: it is not sure you will get the best search results position, and even if you do, only a third of those who click through (not the ones initiating the search!) on average will end up on your site.

This will be the audience where you must reach the highest conversion level possible with an excellent site, good products, good offers, good sales system. Calculate with cart abandonment, those who leave without finishing the transaction, and so on.

Also consider that most expressions are seasonally more or less searched – in some cases without any specific reason, so you cannot even safely forecast the trend. In case of “Christmas sweater” it is understandable there are more matches in December than in May, but there are everyday, all-season products, not related to holidays, which also show similar fluctuating results throughout the year.

Decide if you will have enough potential customers to make it even worth the effort.

2) Relevance

When we say the keyword should be relevant for you, of course, we mean it should be relevant for your target audience.

If you optimize a site based on this, it will reach a good position in the search engine results and your audience will click on it ‒ will they get what they expect? If not, Google will also see this – the bounce rate will be high, and the rate of those staying on your site or even click through within will be low.

This will lead to losing your gained position, as Google very closely watches user behaviour. The algorithm weighs heavily whether real quality and relevant content is to be found on a site.


The keyword must very precisely match your product, otherwise your conversion rate, the number of your potential customers will be lower.

tips Example : If you want to sell a home user Acer laptop, but optimize to the phrase “acer gamer laptop”, because more people search for that, you immediately generate a mass of disappointed visitors. Disappointment is one of the worst emotions you can wish for in ecommerce.


3) Purchase tendency

If you reach large traffic, it is half success. If you really provide relevant and quality content on your site – whether it is a professional article, or just a product description – you will most probably not slip off the top of the list. Even if your offer is good (you have good products, you have written good descriptions to them, you apply proper Call to Action messages), you will not have problems with your conversion, either.

Nevertheless, there are some terms for which you can reach higher conversion rates.


These can best be divided into four big categories:

Keywords signalling purchase intention. If the search phrase contains “price”, “discount”, “shipping”, “purchase”, “buying” or any words similar to these, it shows obvious purchase intention. As a result, a user facing a good offer is much more likely to convert on your site.

Keywords related to products. Such as “best” something, specific products or brand names, phrases like “compare”. These show that the user is currently collecting information. They are less likely to buy immediately, but can be pinned down by content such as a post about comparing within a product category or discussing the reasonable price of a product.

Information seeking keywords. “Product test”, “how to”, “what should I do” and similar phrases. All these illustrate that the customer is learning about the field, the product. They want information, specific knowledge.

Navigation keywords. Company and brand names. They show targeted search, the user searches for specific brands – these can be useful if the company, the brand or even the product you are selling is quite well-known, widely acknowledged. You should not optimize to brand names of the competition, experience shows they do not work well.

Finally there is one more category: the keywords of the hesitating visitors, those who are reluctant to buy. Among these are “free”, “free download” and others. No purchase intention is linked to these: the users search free options instead of buying. Therefore it is useless to fight for them: it is difficult to convince them to buy anything from you. They can perhaps make good audience.


4) Competition

If everything else is perfect, you can still discard the search phrase. And this is what you should do if you see that it would make an enormous effort to fight the sites currently possessing the best positions of the current search results page.

Most search engine optimization tools are capable of assessing the competition – you can also find details about this in Google Keyword Planner, just as in popular SEMrush, as well. The latter, for example, shows you in percentage, how “difficult” a search phrase is in respect to organic matches, meaning how hard it is to make a good ranking for it.

In case of large competition, optimization will be more expensive. Even if you want to reach the top of the results page thanks to organic searches only, meaning you do not use advertisements, but wish to appear among the “normal” search results, above a certain level you will have to employ an SEO expert for the job, generate loads of content, test, and further optimize.

However, if you find a long tail phrase the value of which is unexploited by others, but apparently the users use it, you will have a much easier job.

In most cases you can easily check yourself how optimized the first pages on the list are to the given phrase. Simply observe how precisely the keywords appear in the displayed matches – it is especially clear in case of long tail expressions, if you get poorly optimized matches.

Of course, thanks to Hummingbird and Google, ever since keeping in mind the quality content, the exact match of keywords is not as important as it used to be. It is easy to see, as it is only one variable out of 300 – but if you smartly optimize your website to exact long tail phrases, it will always give you advantage against the competition.


Watch your architecture as well


Not only your subpages, but also your website as a whole must be centred around a given semantic core.


What the hell does that mean?

Basically, it means that the main keywords must be phrases close to each other, and the rest, the less significant keywords must organically be related to them. You will not be penalized if it is not the way you build it up, but Google, Bing and other search engines will be less likely to see your website as a coherent whole, all this leading to damaging your position.

Do not only use relevant phrases, but synonyms as well. Regard Google as a foreigner only hardly speaking your language, who has to be exactly told and overemphasized what you want to say (even in English).


Do not expect immediate success with keyword research and SEO

It is no wonder that SEO is an individual profession. Optimizing is a quite complex task, and keyword research is just the first step – content processing, further market research, refining are yet to come.

This is why you should not expect to get to the top of the match lists immediately. It might happen that even with the best optimization it will take you weeks or even months to reach the desired position, as users react more and more and better and better to your site, as you perform minor modifications in the face of the results.

However, you will not get anywhere without optimization. Your competition optimize, they “occupy” organic traffic, which today is and will be your largest and cheapest source of traffic for a couple of years. You do not want to get out of this, otherwise you are not a top-notch e-merchant.



9 reasons why your visitors bounce off your website

Your website already has interesting content. The design matches the latest trends. You have fantastic images to show off your products and thanks to your marketing initiatives, traffic has started to increase.

Have you asked yourself the question why most of your visitors are still leaving your website? Why is the bounce rate increasing? Why is it that only such a small section of your visitors generate leads and become your customers?

Problems like this occur daily. There is a saying: “If you build it, they’ll come”. Unfortunately, when we want to convert visitors, the job is not that easy. We have to consider a lot of unique factors when it comes to online conversions. Such as correct optimization.

You noticed that your website’s visitors quickly move on from your website without making any conversions? The truth is, there is a lot of room for errors in the background that we have to pay attention to if we don’t want to lose conversions. If you answered the above question with a yes, it’s time you examined if you make the mistakes mentioned in this article. These mistakes lead to the increase of your website’s bounce rate.

We collected the nine reasons that can result in the increase of the bounce rate. If we pay attention to them, fix them, or use them, then the problems will become avoidable and you can keep the visitors you gained with the time and money you spent on achieving your goals.


We’re going to go through the following topics:

  • Content is difficult to read & the design is outdated
  • Too many adverts and options for subscribing to a newsletter
  • Autoplay
  • The website loads too slowly
  • The website is not responsive
  • You haven’t made the advantages of your products clear enough
  • Something is wrong with directing traffic to the landing page
  • Navigation on your website is difficult
  • Lack of call-to-action (CTA)


9 Reasons Why the Bounce Rate Might Be High


  • 1) Content is difficult to read & the design is outdated


First impressions are based on looks. This is also true when it comes to websites which makes web design a very important factor. If your website still has the design elements of its first version and also lacks optimization, it is time for a major transformation!

Naturally, we are not only talking about colours, graphic elements, and images. Fonts, text colours, and their backgrounds are also important parts of the design. All of these factors influence the legibility and the reception of the site’s content.

Difficult-to-read content could later lead to the increase of the bounce rate.
Other than Comic Sans, you can use any font, there are no real rules for this. You can achieve desirable results by having contrasting colour combinations and clean, uncomplicated Sans or Serif typefaces.

tips Tips: A few ideas and a little help to make choosing the font that matches your marketing and corporate identity the most a bit easier: when it comes to size, we should always use a larger font. This can provide a much better user experience for your visitors both on mobile and on PC. For titles, the font size should be 22 or bigger and for the body (the main part of the text on your site) should be 14 or bigger.


  • 2) Too many adverts and options for subscribing to a newsletter


If your website gets its income from the ads displayed there (it works as an advert based business model) then the ads have to stay. But the fact that you need ads does not mean that every inch of the website should be covered by them! On the contrary!

A fundamental part of achieving conversions is trust so it would be better to limit the number of ads and places where they are displayed. It is better if ads are not the first things the visitors see and ads should never be more prominent than the relevant content.

This suggests that optimizing the website is of utmost importance.

A solution for this could be an ad-free template or a custom developed website. The visitors want to see valuable content and avoid distracting elements such as:

  • pop-up ads and texts
  • flashing images

It is recommended to avoid making the newsletter subscription forms too intrusive or pushy. Consider how many forms you need and where to put them on your website! Expedia won an extra 12 million dollars by erasing one field from its form.


  • 3) Autoplay


To tell the truth, I can think of a few more irritating things when it comes to a website than videos that start automatically. This makes me want to close the page immediately.

Today’s online customers have gained the privilege to decide when and how to consume online content. If you flood them with content they don’t need, you will lose them as potential customers in no time – especially if they haven’t made any purchases yet.

Website optimization can help you with moderation as well.


  • 4) The website loads too slowly


The website’s load time is very significant with regards to website optimization and performance. Kissmetrics have carried out a survey that talks about this in detail. Here are some basic facts from their analysis:

  • It is a fundamental factor for 47% of the customers that the page loads in 2 seconds.
  •  If the load time is more than 3 seconds than 40% of the visitors leave the website.
  • After 3 seconds of waiting, customer satisfaction decreases with about 16%.

Do you agree now that website optimization needs to be on top of the to-do list?

It also has to be mentioned that if your page load time increases, Google will also like it more. It is known that pages with quick load times get ahead on the search results list which means:

  • visitors stay for longer on the website,
  • they view more pages,
  • and the bounce rate decreases.

The first to factors contribute to the website getting further up on the Google search results list.


  • 5) The website is not responsive


After Google updated their mobile algorithm, non mobile friendly websites fell behind on the organic search list.

Moreover, Mobify did some research that showed: if your website is not optimized for mobile than you can lose a lot of customers: according to the survey, 30% of people making purchases from their mobile disrupt the process when they realize the site it not mobile friendly.

Turning your website responsive may seem uncomfortable at first but you can’t let your traffic and sales decrease because of it. Think about it: more and more people are shopping on their phones to save time.


  • 6) You haven’t made the advantages of your products clear enough

In order to assist your customers through the sales channel, you have to provide clear product information and highlight why your product is the best.

For example let’s see the case of the original Apple iPod. We’ll show two possibilities of how the iPod could be described on the website:

For the early buyers of Apple, it was important that they could carry 1000 songs wherever they went. The 1 GB storage and the MP3 format were only secondary factors.

Product functions often disappear behind such advantages. It is very important to understand and work with this phenomenon and in order to do this, you have to identify the conversion problems on your website.


  • 7) Something is wrong with directing traffic to the landing page

Those companies that only have one or two landing pages gain less subscribers compared to those in possession of several landing pages.

A landing page is one of the surest ways of building an e-mail list. Because:

  • it only focuses on one problem
  • clear and have no frills
  • it offers something valuable to the visitors in exchange for their e-mail address, for free. Something that the visitor would even be willing to pay for.

It is important for a company to have several landing pages because it makes categorizing potential customers based on their interests easier. As a result, you can offer everyone exactly what they want.


How can I decide who wants what?


Simply because you’ll be in possession of the following information:
You’ll know what the given visitor clicked on:

  • which link
  • which advert
  • which landing page they got to

Simple as that.

If they are on, elegance is the most important to them.
If they are on the landing page, their priority is sportiness.

This does not mean that they are only interested in that one thing, but it is the most important for them so that’s what you have to highlight on the landing page. This is what the title will be about, the main image, the main advantage, etc.

Here’s another problem you have to pay attention to.

In a post on the Crazy Egg blog, Cody Ray Miller describes his experiences on visiting the website.

You would think that typing this address into your browser would lead to the Nissan car brand’s website, wouldn’t you? Instead, it takes us to the website of a small family business selling small computer parts. Accidentally, the manager’s surname is Nissan.

Ask yourself the question: if you’re on the hunt for a new car on the internet but find yourself on a website selling computer parts, would you still insist on the brand you picked or would you broaden your interests?
Of course not. You’d rather keep looking.

What happens if we break our promise and we lead the visitors somewhere else?

The visitor gets excited and for a moment believes that they found the best solution to satisfy their needs. However, when they arrive to the landing page, they realise that the information they are looking for is not available, the page is in need of optimization. What do they do now? Obviously, they leave the page, thus increasing the bounce rate.

To be honest, you lost them forever as potential customers because it is enough to take advantage of their trust in you once.

Never promise anything that is not in accordance with reality or anything you can’t keep because these will only result in:

  • the increase of the bounce rate
  • having a low conversion rate
  • the number of visitors will not be relevant as people only open it and leave right away because they didn’t get the information needed.


  • 8) Navigation on your website is difficult


You’ve come across with this problem before too, haven’t you?

You arrive on a website looking for some specific information and find the following:

  • an infinite and impenetrable online maze
  • that is impossible to make sense of

Obviously, you leave the website in no time and look for another one where you can find the information you need and what’s been optimized.

If navigation on your website is too complicated, it can cancel out the positive effects of Search Engine Optimization and even lose you potential customers.

What can you do in this case? The best is to look at your website as if you were an outsider, one of its visitors:

  • If it was your first time on the website, what would your expectations be regarding the information displayed?
  • How much time would you spend trying to find the information you need and the answers to your questions?

If you consider this and realise that it is in fact complicated to navigate through your website then it is time to restructure it. However; in case you’re uncertain, it’s still not a problem as there are many services available for you such as:

  • user testing,
  • heatmapping,
  • and mouse tracking analyst softwares (Mouseflow, Hotjar).

These tools can help you optimize the bounce rate.


  • 9) Lack of call-to-action (CTA)


According to Small Business Trends’ 2013 survey, 70% of smaller B2B websites did not use CTAs. This is very interesting because it would seem that CTA is obvious for everyone.

Then why do they leave this part out? Simply because they’re concerned about drawing the visitors’ attention to making purchases.

Truth is, however surprising it might be, that customers will rarely make purchases (or do anything else) if you don’t motivate them to do so through optimizing the website. This can be done in blog posts or on the product catalogue page.

tips Comment: To achieve the optimal bounce rate, we have to keep in mind the company’s profile, its target audience, and the website’s conversion. A blog with 80% bounce rate counts as good. It is important to know that this metric is usually used incorrectly. For example, did you know that for the web analysing software Google Analytics, bounce rate means the proportion of single page visits on the landing page compared to the total number of visits. So this rate should always be examined with the context of the subpages and the different conversions in mind.)



In this article, we aimed to gather the most important factors that might lead to the visitors quickly leaving the website.

It is not by any means comprehensive and we did not go into details either, but it is important that:

  • these should be improved as a package, not one by one,
  • during the optimization of the website, it is recommended to consider and correct the errors to make them a part of an all-round marketing strategy.

Strategic thinking should always be in the spotlight if the aim is to develop, measure, analyse, and optimize a website. This way we can see how the different parts are connected and make up a whole.



Use the AARRR Funnel Analysis to achieve more ecommerce sales

As you go from Acquisition to Revenue during the process, less and less clients remain on the different levels. As a consequence, the diagram of the process forms a funnel; that’s where the name funnel analysis comes from.


A little funnel analysis history


Elias St. Elmo Lewis created the AIDA model (Attention, Interest, Desire, Action) in 1893, based on his own sales experience. This model concentrates on grabbing attention, then interest, creating desire and ultimately, and encouraging taking action.

The model is built on advertising psychology and its main purpose is the analysis of offline marketing tools. Right until around 2000, the use of the model was extremely popular among marketing professionals, however, due to the spread of online communication and websites, a new model appeared which was created in 2007 by Dave McClure. This was the AARRR model, also known as AARRR funnel analysis.

Of course all this doesn’t mean that the AIDA model is no longer in use (on the contrary!), it’s just that it approaches the process from another aspect. Due to the rapid changes of the online world and because of the experiences gained based on these changes, newer and newer models and ideas are born.

The AARRR model focuses mainly on online marketing and data-driven marketing, and it’s also ideal for the monitoring and optimization of conversion (mainly transactional) processes.


The AARRR model analyses the users through 5 measurable dimensions:

  • A: ACQUISITION ‒ Where do visitors come from? How much does the acquisition of a visitor cost (e.g. CPC, CPA)?
  • A: ACTIVATION ‒ Have the clients gained a sufficiently positive experience when visiting your online shop? Have you managed to turn them into users?
  • R: RETENTION ‒ Will they return to your communications channels? How often do they use your product/service? How often do they purchase from you? How high is the churn rate?
  • R: REVENUE ‒ Do users purchase as well? How high is the cart value? How much is the revenue per user? What kind of profit increasing tools do you use (cross-sell, up-sell)?
  • R: REFERRAL ‒ Do users like the site and/or product enough to recommend it to others? Are the customers satisfied?


Different indicators measurable by measuring tools can be defined in each dimension of the AARRR funnel analysis. After the definition of the indicators, you will be able to measure and then to optimize these quantified and scaled values, thus increasing your efficiency.


According to McClure, these measurable data may be the following:

  • quantitative,
  • qualitative,
  • competitive – it allows comparison with the data of the competitors.
  • comparative – it allows comparison with your own data.


Diagram of the AARRR model


The model can be illustrated in several ways.



Log into Google Analytics and then to Reporting Suite. Here you’ll find the illustration options of the model defined by you under the menu Conversions/Targets/Channel illustration (it’s necessary to create properly set targets and access routes for the channel illustration).


google analytics funnel

Example for Google Analytics channel illustration



Another kind of illustration is the bar chart illustration. The majority of the analysis software applications give preference to this.


sales funnel in ecommerce

Example for bar chart channel illustration


You can use various free of charge (Google Analytics) and paid (Kissmetrics) pieces of software in order to increase the efficiency of your website (generating more registrations or higher revenue). After choosing the most appropriate analytical tools, you will by all means have to integrate the code of the measuring tools into the ecommerce store. This step is essential because this way you will be able to measure all the indicators and levels of the funnel model.


Now let’s see the dimensions of the AARRR funnel analysis in more detail




The first dimension of the AARRR funnel analysis collects the data in terms of the acquisition of visitors. You will certainly have to determine the channels for the acquisition of visitors. Such channels may be for example social media, organic searches, email campaigns, display campaigns and the channels of the different advertisements.

After having determined the channels, continuous auditing of the obtained results is indispensable. The number of acquired visitors and the cost of the visitor acquisition determined on the basis of the channel, may be appropriate indicators. The purpose of this stage is to have as many relevant visitors on your website as possible, and to achieve that with the lowest possible marketing costs.




After the users have found your online shop, the second dimension of the model, activation, comes in, the purpose of which is the increase and the measurement of their activities. You can measure the activities of your visitors (and later your users) with various analytical measurement tools. Such tools are Google Analytics available to everyone, and the different heat mapping analytical tools (e.g.  Mouseflow).

Key indicators of the activation stage may be: number of page views, time spent on the given page by the customers, or the number and cost of conversions targeted by the company.


Conversion indicators may be, but not limited to, for example the following (depending on the company and the profile):

  • registration,
  • logging in to the online shop account,
  • subscription to the newsletter,
  • launching the payment process,
  • viewing certain product pages,
  • adding the product to cart,
  • viewing videos.




Even if the 80/20 principle, elaborated by Vilfredo Pareto, (according to which 80% of the effects come from 20% of the causes) will not necessarily be valid for your company, still, the visitors and the users of your ecommerce store who showed some sort of a commitment and loyalty (in the scope of the already mentioned conversion indicators) and returned to your online shop play a vital role.


tips Fact: Reaching the already acquired visitors involves a much lower cost than reaching those who have not yet visited your website.


You can measure for example the following key indicators in the third stage of the AARRR funnel analysis: ROI (Return on Investment) indicator of remarketing and email campaigns, churn rate, and the rate of returning visitors compared to the total number of visitors of your website.




The most important indicators from the point of view of all companies may be revenue, and the amount of the realized profit.

In this dimension of the model the measuring tools help you determine the scaled values of the customers. Such values may be:

  • average cart value,
  • the cost per acquisition of one customer,
  • income per one user,
  • the time elapsed between the first visit on the website and the purchase
  • and the amount of the monthly revenue itself.




The fifth and last dimension of the funnel model is based on the viral spread, the meaning of which is that your visitors or clients share their views and the pieces of content and/or products offered by your ecommerce store, in the form of Facebook posts and/or YouTube videos for example.

Then later these pieces of content reach the other users without involving additional costs (that’s what’s called viral spread). Important key indicators of this stage may be:

The number of shares on Facebook and the numeric value (from 1 to 5 stars) of the ratings,indicators of customer satisfaction,viral indicator (how many new customers arrive thanks to the referral of your already existing customers), viral cycle (how long it takes until the person to whom the online shop was recommended becomes a customer thanks to the referral.)


The referral stage does not necessarily have to be the last one, since the visitors can share their opinions on the ecommerce store or on one of its products even during the first visit.


It’s also possible to change the order of the elements of the AARRR funnel analysis created by Dave McClure.Let’s take as an example that visitors can also share their positive or even negative experiences without purchasing, so referral doesn’t necessarily come to the last stage of the analysis.

The order of the dimensions of the funnel analysis presented in this article is not set in stone, however, it looks like this in most of the cases (based on observations).

In addition, it provides guidance concerning which points are worth defining and analysing with the help of the metrics and key indicators (KPIs) determined based on the goals of the company.

In general, elements of the AARRR model interact; they are connected in a coherent system with each other, and you can analyse the results of this connection most efficiently with a cohort analysis showing the data of several months.

One of the disadvantages of the model is that it is primarily suitable for the illustration of linear processes. On the other hand, its advantage is that it has recognized the macro steps which the visitors take towards the purchase.

However, the order of these changes vary according to both the type of business and the target group, but the order of the steps can be completely customized to the company profile/target group.



Can Magento really handle over 500,000 products? If yes, how?

In an effort to give some useful hints, let us try to find an answer to some more precise questions:


  • How many products the different versions of Magento can manage easily by default, hosted on an average server, with and without extra optimizations?
  • What makes the difference between Magento versions in terms of catalog performance?
  • What are the main bottlenecks for a big catalog in Magento, and how can Magento be scaled to handle even more products?
  • What are the common mistakes that one can make when handling a huge Magento catalog?


How many products the different versions of Magento can manage easily by default, hosted on an average server, with and without extra scaling?


  • Magento CE 1.9.x safely manages ~10,000 – 25,000 products in most cases, without much extra care*. On the other hand it can manage 100,000 – 200,000 products, or even more via heavy, system wide scaling, resource tuning and code optimization.
  • Magento EE 13.x. 14.x safely manages ~100,000 – 200,000 products in most cases, without much extra care*, and even 400,000 – 500,000 or even more with proper scaling, optimization and server resources.
  • Magento 2 CE safely manages ~ 100,000 – 200,000 products in most cases, without much scaling and extra care, and even 400,000 – 500,000 or even more with proper scaling, optimization and server resources.
  • Magento 2 EE is designed to be able to manage even more**, depending on some highly advanced enterprise features like database sharding, job queues, advanced mysql and web server topologies and proper resources.


All these are very rough numbers, of course, referring to a catalog with a few attributes, categories and simple products. The figures are based on Magento’s own performance tests and our own experiences. The figures may vary greatly by the server setup, software and resources, too.


What is also very important to note is that due to Magento’s database design, there are some especially massive aspects that act as multiplication factors to get to the real number of products Magento really works with in the background.


Some of most important elements that have a huge impact are:

  • number of Magento shops/languages
  • number of product attributes
  • number of categories and depth of the category tree
  • number of configurable/bundle products
  • number of customer groups with different product prices


All this means that as far as catalog management is concerned, a Magento catalog with a few thousand products in a heavyweight catalog setup, for example ~20 languages, diverse attributes and category structure, mostly with configurable products and multiple customer price groups, may equal to a Magento catalog with 100,000+ simple products in a single store.


These are the features that make Magento really flexible, but they come at a cost.


What makes the difference between these Magento versions in term of catalog performance?

versions of magento


  • Magento CE 1.x, including 1.9.x and
    • – indexing, especially URL and search indexes are not optimized for large catalogs  (this also applies to EE up to 1.12.x).
    • – No Full Page Cache (FPC) is available by default
    • + some frontend optimizations like javascript + css merging, CDN support
  • Magento 1.x EE
    • + FPC is available, which speeds up catalog browsing and saves a lot of server resources
    • + From version 1.13+.x incremental indexing is introduced whereby products that were changed or added will be re-indexed in the background by cron jobs.
    • + From version 1.13+.x full reindex processes are highly improved as well and work well even for large catalogs.
    • + Solr search engine is available by default
  • Magento 2 CE
    • + Inherits incremental indexing feature from Magento EE 1.13+.x
    • + Inherits FPC from EE 1.13+.x and Varnish frontend cache is added as a choice of FPC. The requests that are served by the Varnish cache never need to reach the Magento application servers, which reduces the load on the web nodes while dramatically improving the response time.
    • + Browser cache is utilized for session data caching
    • + Checkout process is improved greatly
    • + Async order and product updates
    • + Client side optimizations like minification, js resources bundling, caching static content, image compression
    • + PHP 7 is supported by default. PHP 7 may have even 200% performance gain over PHP 5.x by itself.
  • Magento 2 EE
    • + Has all the features of Magento 2 CE
    • + Solr (2.0) and Elasticsearch (2.1) for search
    • + Database sharding separating catalog, checkout and order business domains  is available
    • + Mysql cluster and Multi-master mysql setup is supported
    • + Job queues introduced for advanced background data processing (deferred stock update as the first implementation)


What are the main bottlenecks for a big catalog in Magento and how can Magento be scaled to handle even more products?

magento with bottlenecks


  • Hosting – a big catalog obviously needs more resources, a VPS with plenty of memory and a multicore processor is a must, a multi-node server setup may be needed.
  • Server softwareNginx as a web server and PHP 7 and Mysql 5.6 or equivalent (Percona/MariaDb) are highly recommended, even for Magento 1. Fine tuning of these pieces of server software for Magento is essential in this case.
  • Product Import – optimized and tailored product import and update is one of the key elements in this case. Tools that enable batch database updates are extremely helpful. Any method or tool that uses single product updates is a huge potential bottleneck. Magmi is one of the best choices here.
  • Some other rules of thumb:
    • Save only what has changed
    • Use dedicated resources for import processes
    • Separate price, stock and basic product data import
    • Re-index only products and product data that needs to be reindexed    
  • Indexing  – Indexing in Magento is a second step of saving product data and it is the trickiest part when having a huge catalog. It contains of a series of processes to copy product data from database tables optimized for data storage to tables optimized for different aspects of frontend data access. Since indexing is “only” needed for Magento frontend features, it is possible to separate indexing from product save or import. In the newer versions of Magento 1 EE and Magento 2 CE and EE incremental background indexing is introduced, which speeds up working in the admin but may still not be ideal for large volume product updates.
    • Indexing in Magento 1 CE is one of the greatest bottlenecks.
      • URL indexing tends to provide the most issues, partly because it can be bloated to millions of records, partly because it runs for a long time and in Magento 1 CE it is not optimized for big catalogs.
      • Beyond a certain catalog size and number of Magento stores the increased overhead of product flat indexes will outweigh its benefits so it tends to be better not to turn on this feature for a big catalog.
  • Catalog search
    • Default MYSQL fulltext search is resource greedy, indexing and the actual search on the frontend tends to be slow and its accuracy and the relevance of the result set is fairly poor. So it has to be replaced, even in Magento 1. There are some good, even free alternatives for Magento 1 to replace MYSQL search with Solr, Elasticsearch or Sphinx. Magento 1 EE has Solr and Magento 2 from 2.1 on Elasticsearch by default. Some 3rd party Elasticsearch and Solr 3rd extensions can replace the default MYSQL layered navigation engine at the same time, which can be a huge benefit.
  • Full Page Cache
    • Full Page Cache (FPC) is a mechanism whereby html pages generated by the server softwares are cached as a whole. Next time the same web page is required, the cached version is returned without the need to regenerate the content. In Magento 1 CE has no FPC by default and in Magento 1 EE caching is managed by Magento code itself. It saves a lot of resources and results in greater speed, but the best way to do FPC is by implementing caching in a layer in front of Magento without the need to touch Magento at all when a cached content is served. This is accomplished by Varnish caching in Magento 2. Though Magento 1 CE has no FPC and Varnish, there are good extensions to implement these features, too.
  • Application Cache
    • Magento heavily relies on different types of config and application caches, Redis, a memory based, scalable application cache is highly recommended with tag management. The extension to handle Redis is built into Magento starting from the latest version of Magento 1 CE.


To sum it up for different Magento versions:


  • Common bottlenecks are product import and indexing in the backend, search and layered navigation in the frontend. Checkout and Order management are also important factors, but these are more strongly related to the number of visitors and transactions.
  • There is a big chance that batch product import is to be tailored and optimized in all Magento versions – feature rich and optimized product import is still to be solved even in Magento 2 EE.
  • Magento 1 CE is the platform least prepared to handle big catalogs, but through heavy scaling, proper hosting and 3rd party extensions that provide better indexing, search and caching, it can be considerably upscaled. Indexing may still remain a bottleneck, however.
  • Magento 1 EE has a bunch of optimizations already in place, to implement Varnish cache and fine tune the server architecture are the best candidates to scale it up even further.
  • Magento 2 CE is designed in a way that it should be prepared to serve middle-sized businesses, too, just like Magento 1 EE. Functionality wise it lacks enterprise features, such as store credits, better CMS management etc., but as far as performance is concerned it is on pair with Magento 1 EE – or even better due to Varnish. An obvious way to scale it up is to utilize built in optimization options, fine tune server architecture and resources and implement Elasticsearch or Solr for catalog search.
  • Magento 2 EE targets enterprises even beyond the middle-size range and aims to offer a highly scalable architecture that utilizes the benefits of cloud computing.


What are the common mistakes that one can make when having a huge Magento catalog?


In addition to failing to missing the points above, there are some additional mistakes that one can make when having a huge Magento catalog.


  • Underscaling – Maybe the most important advice here is that it is very important to build a live Magento store in a way that it has plenty of system reserves and there is a proven way to scale it up quickly if needed. Performance tests throughout the development cycles are very useful so that the limits of the system are known.
  • Too many / poor quality 3rd party or custom extensions  –  3rd party extensions are not always built for and tested to be performant for large catalogs and even a small oversight in the extension design can result in performance disasters.  
  • Lack of proper monitoring – System monitoring is essential so that issues and bottlenecks can be discovered and eliminated in time.


Some technical details behind the hints


In the section below we try to provide some details on Magento concepts and terms for the better understanding of the subject.


Data Storage in Magento


One point is that Magento is very complex and its features are virtually maxing out the limits of a PHP/MYSQL based system. Products are really massive objects, with many aspects to them from its really versatile attribute scheme through prices, images, categories, product options to different product relations.

What is more, all these properties may have as many different layers of values as many stores and languages are used. On the top of that, these aspects are potentially extended by a number of 3rd party extensions. All this means that saving and updating products in the database is a complex action.


Indexes – Preparing for Magento’s Frontend Data Access


The other point is that the structure of the database tables where product data is primarily saved is optimized for flexible data storage but, due to this complexity and the nature of relational databases, not optimized for the different types of data retrieval at the same time. To save the day, Magento introduced the so-called index tables that are populated by the process called indexing.

Indexing in Magento: is a process whereby data is copied from database tables optimized for data storage to tables optimized for frontend data access

Most indexes, like URL, category/product relation, price, and stock and attribute indexes, as well as search index are pretty much essential for a stock Magento to work. There are some other indexes that are optional such as catalog flat and product flat indexes, which flatten the EAV and multi store data to dedicated store tables and single product rows.

Indexing in Magento has two faces. On the one hand, it enables Magento’s most powerful features to work and optimizes data access. On the other hand, it makes even more complex, time consuming and resource greedy to store product data in a usable way for the frontend.


  • Product Attributes Index – By default, it is essential so that layered navigation works. It copies product/attribute options data to a table structure that is optimal for finding products based on the different attribute options they have. The number of attribute options are multiplied by the number of stores/languages Magento has.
  • Product Price Index – By default, It is essential so that sorting and filtering by product prices work. Any Tier prices, configurable, bundle and products complicate this calculation to a great extent since the minimal price of complex product types depends on the price of the constituent products. The number of Magento websites and price groups act as further complicating factors here.
  • Catalog URL Rewrites Index – By default, it is essential so that SEO friendly URLs and redirection from old URLs to new ones work. The depth and size of the category structure, the number of old URLs and stores/languages has a huge affect on this table. The size of this table can be blown up to millions of records and keep continuously growing even in the case of a catalog with a few thousand products and few languages. In Magento 1 CE and Magento 1 EE up to 12.x this is definitely the single most problematic indexer as far as big catalogs are concerned.
  • Category Products Index – optimizes product filtering based on categories by storing catalog-product relations in a separate table. Again, it is an essential index for the frontend. The fact whether layered navigation is used for a category (“is anchor”) or not has an effect on these product relations.
  • Catalog Search Index (fulltext) – It is essential for default MYSQL-based search to work. It merges the text of product attributes and option labels of individual products so that it is searchable by the MYSQL fulltext engine. Here again, as many stores there are, so many index rows are created for a single product.
  • Stock Status Index – This is for calculating the fact whether product is purchasable in Magento, which may be governed by the mixture of some global, website and product level values and settings.
  • Product Flat Data Index – This index is optional and was added at a later point in the evolution of Magento to speed up product listing/sorting and spare server resources. The way it works is to copy product attribute values which normally could only be retrieved by huge queries joining multiple tables into a flat structure with only one record per product and store.


Database design


  • Magento EAV
    • EAV stands for Entity-Attribute-Value. Is a dynamic attribute management pattern which enables adding/removing/modifying product attributes like colour, manufacturer etc., without changing the structure of the database tables. This is a very powerful and user friendly feature and Magento has a number of configurational options for custom attributes out of the box.
  • Magento multi store
    • It is a unique Magento feature that a single Magento database can manage multiple stores in a way that attributes can have dedicated store level values that may override default values. This, again, is a highly user friendly approach and makes it easy, for example, to create different language versions of the same catalog only by changing the description of the products and label of the product options.
  • Magento indexing
    • Indexing in Magento is a process whereby data is copied from database tables optimized for data storage to tables optimized for frontend data access. Different indexes are used for different types of data access. Indexing processes are built to speed up the shop and generally provide a huge benefit, however make data storage more complex and resource greedy. Especially in the case of full reindexing, which is sometimes inevitable, and any Magento shop should be prepared to be able to perform a full index rebuild, it may take a long time and require huge resources to reindex all product data.



SummaryMagento has gone through a great evolution since the beginning and by now with Magento 2 out it is certainly not only the most feature rich open source online store system, but it is probably the most performant as well. If it gets the care and resources it needs it can manage huge catalogs with great success.



Ecommerce trends 2017 – get prepared now and conquer the market

No matter which market your online shop is targeting, no matter what kind of product you are selling, you should expect your turnover to increase significantly in the next few years, because the number of customers on the internet is also growing, and the market is expanding – provided that you’re actively monitoring the current trends.

If you start to get prepared for 2017 now, you still have enough time to plan your most important marketing campaigns, to come up with solutions how you’re going to transform your ecommerce store in line with the most recent user habits, and how you’re going to use the new technologies.

And if you’d rather wait a little longer, you’ll inevitably lag behind your competitors.


We’re going to talk about the following topics:

  • You don’t have to be an innovator in every field
  • Real time analytics and reach
  • Mobile penetration is continuously spreading
  • Micro moments
  • Make the customer path story-oriented
  • The era of the virtual salesmen will come
  • The perfect user experience will become a basic expectation
  • Social media takes control
  • Omnichannel – communicate everywhere at the same time
  • Quicker delivery
  • Maximizing the number of interactions
  • Let’s summarize the trends


You don’t have to be an innovator in every field


Trends are called trends exactly because we’re not talking about occasional, suddenly emerging phenomena here, but usually about predictable processes.

For this very reason, we can write an article already in autumn 2016 about what trends are expected to dominate the year 2017, and which would most probably be of great significance, as the market was clearly heading in those directions.

This also means that you don’t have to be afraid that one of your competitors will outpace you in almost every field just because he actively monitors the market.


tips Important: In any case, the most important thing is to satisfy the needs of your own customers, and you’ll have to get to know them to be able to do that.


(We’re going to talk about this in more detail shortly – because it does matter how you treat them, what kind of information you’ll ask for and store.)

It’s quite possible that your audience will require only a couple of elements of this list, and absolutely not all at the same time. It’s also possible that they don’t have any other significant need than you deliver them the ordered products quickly and reliably. On the other hand, they will make purchasing decisions based on that.

Therefore, you shouldn’t look for where you could introduce some kind of innovation based on your own preference. You should always consider what kind of path the customers purchasing from you follow, and where you could simplify and also raise the quality of that process, at what point you could help the user and how you could ensure that they choose you the next time as well.

Bearing that in mind, let’s see which are the emerging trends in the field of technology, marketing and sales, the opportunities of which you should at least explore when planning your strategies for the next year.


Real time analytics and reach


With the help of the most recent analytical methods, you can follow your customers not solely based on their past behaviour. Software applications, like Kissmetrics, offers, are able to tag the visitors and follow them real time continuously through the different channels, trying to get them with retargeting or other methods to the predefined target conversion.

Nowadays you may know much more about what your users, potential and actual customers do in your ecommerce store and why they do that – you’re able to see where they click, where they move on to after leaving you, and you are able to guess their motivations based on the data, and you can target them continuously in line with that with the most appropriate message in the given situation.

Continuous active reach has a major significance – now that you have the proper tools, and the analytics are also at your disposal, it can certainly be declared that the customers, to whom you don’t only send a couple of newsletters as a first contact, are much more valuable to you.

tips Useful tip: According to a survey, customers that are reached continuously, visit online merchants 44% more often and spend more (373 dollars on average) per purchase than those, who are not communicated with continuously (289 dollars).

There may be a lot of forms of reach – and if you follow the modern online marketing approach, you shouldn’t try placing banners on sites with high traffic. Because that’s simply a waste of money compared to the truly effective methods.


Let’s see some of these methods:

  • You can create questionnaires, thus collecting testimonials and data, and then publish them later.
  • You can answer their questions in posts, or even by creating video content.
  • You can send them newsletters, follow-up letters, and a great deal of various emails.
  • You can also reach them through social networking sites by creating closed groups, or by sharing content through traditional channels or even by retargeted


The list can be extremely long – you can contact your customers and visitors through a lot of channels, some of which we are soon going to discuss in detail.


Mobile penetration is continuously spreading


We’ve already written about how more and more people purchase through various smart devices as the devices become more and more widespread, and as the trust of the users increases. Overall, however, the conversion rate is still low compared to the desktop.

On the other hand, this will not at all be the case in the next few years. Based on the researches of Gartner, e-merchants are expected to increase their revenues stemming from targeting mobile phone users by even as much as 50% in 2017.

Gartner draws attention to another interesting fact: next year the revenue generated by digital mobile assistants may reach even as much as 2 billion dollars. These practically represent the technologies which facilitate the potential customers’ path to conversion, either by the automatic filling in of some data or as an intelligent virtual assistant.

It’s an interesting fact that while an average user grabs his mobile 150-200 times a day and looks up information (e.g. local shops), and he practically realizes most of the purchasing process micro moment by micro moment on the mobile – even if the conversion will take place on the desktop in the end –, only 14% of the companies pay attention to exploiting mobile as an ecommerce platform.


Who will rule the markets? The previously mentioned 14% that recognize before their competitors do that this is an area in which an enormous growth potential lies and which is continuously expanding.


The transformation that the near future will bring, will be radical, though its signs are already visible today. Mobiles will become accepted as a means of payment, the purchasing process will almost entirely move to the smart devices, and the desktop will fade into the background – which at the same time means that you will be able to reach the users anytime and anywhere with a well targeted relevant message. Just like it also means that they will be able to make the decision to purchase anywhere and anytime.



tips Conferences you may be interested in: To keep up with recent trends, attending conferences can be of enormous help. Here’s a very nice list (created by Red Stag Fulfillment) of the conferences to be held around the world in 2017 covering different topics, e.g. conversion optimization, mobile commerce, international ecommerce and many more.



Micro moments

eCommerce trends and micro moments 2017


Google’s 2015 concept, simultaneously with mobile ecommerce, will most likely reach its peak next year.

The micro moments concept is quite simple: the purchasing path followed by the users, which could well be controlled traditionally, has fragmented into small pieces, i.e. moments, by now, as the search for the information, the collection of the data necessary for the consideration of the decision, the purchase, as well as the content consumption itself are all done on our mobile devices at random moments.

If you want to control these moments, you have to do an excellent job in several fields at the same time.

  • In mobile search engine optimization: if the user searches for a product or a service that’s relevant for you during his path, he will find your pages.
  • In content creation: you don’t necessarily need to show the product right away, the potential customers many times just simply collect information on the mobile phone. They see something on the street that they would like to have, and they become curious about what it is capable of, how much it costs, whether it’s on special offer somewhere in the city, and so on. Consequently, you need to create pieces of content that answer these questions, and which also rank high on the search result pages.
  • In segmentation. You need to know who your target group is, what they think about, where they are, how they use their phones, what they search for. Because this is the only way you can make sure you elaborate a strategy that can stick with every micro moment that’s valuable for you.
  • And this segmentation doesn’t only have to be done based on only demographic and other similar data: you have to know the intention of the users that you may guess of course based on their activities. It’s not that you have to know in advance what a given potential client is about to do – however, you need to recognize the patterns, namely how the customers with various intentions usually get to you (or to your competitors), and you have to enter into that process at the appropriate point.


The first steps that you need to do, if you really want to dominate the market in the mobile era, are:

  • Get to know your customers – make segments and create buyer personas. You should know exactly what and how they purchase, what they search for – and more. How they think, what they like reading.
  • Collect all available data.
  • Know where your customers live.
  • Whether they like using mobile or desktop
  • How they like to pay.
  • What kind of content they consume.


Make the customer path story-oriented


Storytelling: is one of the oldest and most efficient marketing tools, nevertheless, its new renaissance came with the expansion of content marketing within online marketing.

Content is clearly capable of not only selling but of improving your indicators in practically all areas – provided you create it smartly. And this, in most of the cases means that you establish some kind of a personal relationship with your customers, thus evoking such emotions in them that will finally lead them towards conversion.

Find the point where you are capable of making the users commit themselves emotionally. Where you can show them how well a product worked for people similar to them, how much easier, happier it has made their lives.

Don’t try to influence logic at first, but rather try to evoke some kind of an emotion instead, and raise attention with that, and then reinforce the commitment of the customer with that later.


The era of the virtual salesmen will come


We’ve already mentioned virtual assistants – one of the biggest predictions for 2017 is that virtual personal assistants will burst into the market, who will assist the customers at every step of the purchasing process.

At the same time, it’s likely that personal, face to face selling will play less and less of a role – people selling in offline stores will be replaced by automated systems, which can often inform the customers more effectively.

Salesmen will work online instead, in order to answer questions that automated virtual assistants can’t, or even to fine-tune sales processes.


The perfect user experience will become a basic expectation


User interfaces are becoming simpler and simpler, just as well the purchasing processes – users are getting more and more used to the simplicity (e.g. tap, swipe, pinch, drag and drop etc.) that mobiles and smart devices in general provide them with.

This user experience, which a lot of ecommerce stores are optimized for, influence expectations quite significantly. The mobile has an impact on the desktop: if you can purchase on the mobile with only a couple of finger movements, the long checkout process on the desktop will soon also seem uncomfortable.


Surveys show that the users long for simpler processes in every field: simpler information search, simpler purchase, support, commenting etc.


At the same time, the difficulty lying within is that you’ll need to be able to transmit all relevant information to them as well – and these are usually available in quite complex systems.

Therefore the solution is to work with such intuitive automated systems which are able to define, based on the behaviour of the users, what their current intentions are, what sort of information they need and what kind of activity they are most probably going to carry out – and which systems offer all this to the users without them having to look for those at all.

Earlier, online stores tried to persuade those “wandering into” their shops to purchase by showing them the most diverse product range, like in a bazaar, since that’s how they had the biggest chance of introducing them something they would actually like.

Today, in the age of the minimal user interfaces and of the advertisement or banner blindness, this doesn’t work (or rather it has never been an effective technique).


tips Tip: The truly effective technique is to actively monitor from the very first moment how the visitors act in your ecommerce store ‒ to try to gather, even passively, all possible information, and to direct the users based on their behaviour, by showing them personalized offers – always one at a time.


Social media takes control


The role of the social networking platforms in ecommerce is significant even today.There is a great deal of people who get even their shopping tips from their friends, or from brands present on the social networking platforms. The role of social proof, of social reinforcement is even more important.

This is valid for not only the goods offered online, but for the pieces of content as well. While earlier search engine optimization played the main role, as you were trying to catch the organic traffic arriving from Google’s search engine, nowadays optimization has to be more complex than that, and it should target social networking sites in the first place.


A resourceful e-merchant can find several services that will help him stabilize his presence on the social sites, automate the processes in a way that both the click-through rate and the return increases because of that.

Nowadays, sales on many platforms can be outsourced – it’s possible to purchase directly on Facebook, Pinterest or on other platforms as well, where you can sell products and services in a way that the users don’t even have to leave the site. And this is even encouraged by the platforms, since it’s in their interest that while the entrepreneur pays for the advertisements and gives a commission on the sales, the users stay on their sites and don’t leave for an external ecommerce store.


Omnichannel – communicate everywhere at the same time

social media trends in eCommerce 2017


Of course, the fact that the social media is present nowadays as a new starting point, doesn’t mean that you should neglect any other of your channels. It’s the contrary.

If you really want to be there for your customers at all times (not in a disturbing, but in a useful way), you need to use all available tools to communicate with them and to offer them pieces of content that will most probably be relevant and useful for them.

Pieces of content that will solve their problems even by themselves, that will answer some of their questions – and that can somehow be tied to your area of expertise or to your products.


tips Suggestion: You should be the one they feel free to ask instead of asking for advice from their friends concerning a given issue, and also the one who they specifically expect to be approached by with useful or helpful information or tips.


This is an important piece of advice even because according to Google’s data 85 percent of the users start the purchasing process (for example by collecting data) on a given channel and by the time they get to conversion, they switch to another one. Consequently, you have to be present everywhere, so that they can have the opportunity of doing that.


Quicker delivery


Customers expect that the ordered goods actually arrive to them in the shortest possible time. In case of the American customers this often means same day delivery: more and more ecommerce stores undertake that they deliver what you order on the same day.

Obviously, this is not possible in every case – it would often be physically impossible to fulfil same day delivery.

Large companies are testing several different methods to shorten the 5-7 day shipping that otherwise feels almost like forever, to a 1-2 day delivery.

Amazon’s experiment with the drones for example is one of the best-known of these, however, this has several important obstacles – on the one hand, due to the complicated regulations and licensing procedures, on the other hand, because this represents a way too expensive solution for a small enterprise.

Another similar solution that’s currently in the testing phase is crowd delivery, in other words the solution where the company doesn’t use the postal service or a courier service, but they simply let the community deliver the package to the addressee within the shortest possible time. However, a very well established and wide customer audience is necessary for this, with whom trust has already been built previously.


Advertise what you’re actually selling


On this blog you could already read about the importance of not only your text content should be authentic, but also visual content – product photos, among others.

This is just as well valid for the various advertisements. For example if you create a channel in the social media, where you can communicate with your customers, it’s worthwhile filling that up (let’s say on Pinterest) with things that they can actually find in your online shop, or that they can even purchase right there immediately with a very little search (without being too pushy, of course).


Maximizing the number of interactions


The very reason why the merchant has a difficult job with the building of trust in online shopping is because the users don’t have the opportunity to get to know the product personally.

They don’t miss salesmen standing next to them asking whether they can help them – nowadays virtual assistants offer a more than satisfying solution for information services.

No, the problem is that the customers can’t take the product in their hands, can’t feel it, can’t smell it. They can’t make sure that it is exactly what they need.

Today, innovative ecommerce stores are testing various methods which can facilitate building up trust.


It’s essential that you try to present the product in every possible way:You shouldn’t use only text to present what it is, but you should use high-resolution, authentic product photos as well. You can provide an interactive opportunity, where the users can freely rotate the product in three dimensions, can zoom in on it as they wish, or they can even change its colour, and can put its pieces together themselves.


The so-called “magic mirrors” also provide an interesting concept: in case of an ecommerce store selling clothing products, after asking for some data (in this case a photo) from the users, you can show them what they would look like in a T-shirt if they bought it. Of course this technology is quite in its infancy, however, it will be interesting to see how successful those trying it will be in having their customers commit themselves and in convincing the hesitating ones.

There is also great potential in virtual reality. A virtually “tangible” product presentation may be way more effective than a simple video. The problem is that the tools making the exploitation of virtual reality possible can’t be considered wide-spread even in the United States, while in Europe and other markets several years will pass until the users have for example a pair of VR glasses in their homes.


Let’s summarize the trends

If we needed to guess, based on what’s described above, in which direction ecommerce is moving generally speaking, we can say that in the direction of the user-oriented approach.

The following guidelines are the most important for an e-merchant:

  • learn everything that’s possible to know about your customer
  • personalize the offer based on that
  • facilitate the purchase as much as possible

And use the most innovative solutions in order to achieve this – with two exceptions:


You should only use similar solutions if you and your customers really benefit from them. You should make sure that the change you want to introduce is not for its own sake. The best is if you ask for the opinion of your audience in advance every time you are thinking about some sort of a radical innovation, because your spending may turn out to be unnecessary in the end.


The other requirement is that you introduce only those innovations that can actually bring improvement: so that you can truly improve the process with that. A single drone in a strictly regulated air space for example will be unlikely to improve your delivery speed at all, and a VR application in a country where VR glasses can only be found in the offices of tech companies, would be a completely useless innovation, no matter how innovative the solution we’re talking about is.

If you want to know which solution you should consider, we can provide you with a single piece of advice in addition to that you should always pay attention to what the numbers tell you.

Use your own service: continuously test how it feels to communicate with your online shop, what is content consumption, search and purchase like.

Where you feel that the process is even slightly unstable or unclear, there’s space for optimization. And where some kind of a solution is entirely missing or a significant change is necessary for instance, there’s space for innovation.

On the other hand, you should always bear in mind that your only important task is: to make your customers’ life easier, and not to focus on helping them to spend their money in your online store. They will also be grateful for this attitude, and thus they will return to you later as well.