What to measure in ecommerce analytics and how to improve results?

There are loads of opportunities for the optimization of an ecommerce store – you can search for best practices, case studies, but in the end it will always be your own experiences, the behaviour of your own audience on which you can base your decision whether you will benefit from a specific modification or not.


What will we talk about in the following?

  • What to specifically pay attention to in Analytics?
    • Audience – who is visiting, who is shopping and how?
    • Acquisition reports, or in other words, conversion cycles on your website
    • Behaviour – why do customers do what they do?
    • Conversions – where, how, and if not, why not?
  • Why not to pay attention to everything at the same time?
  • Tips so that you can improve the numbers
    • Reduce loading time
    • Provide all information
    • Grant exclusive offers based on behaviour
    • Simplify the check-out process as much as you can
    • If something does not work, get rid of it quickly
  • Is that all you have to do?


Most ecommerce systems offer some kind of basic features for the collection of data, but if you want to build a more significant database, you will definitely need the integration of Google Analytics. Google Analytics, by all means, can guarantee you an incredible amount of information of which you need to pick out those few indices that you will monitor continuously.


What to specifically pay attention to in Google Analytics used as an ecommerce analytics tool?

Reports and indices in Analytics that are essential for e-merchants can be divided into four main groups. These are the following.


ecommerce google analytics



The primary factor that any website, dealing with ecommerce, has to measure precisely is audience. Age of the customers, proportion of the returning clients and similar information can all be found here – in order to be able to build any marketing campaign effectively and to be able to target the various buyer personas, you need to know these and act in line with them.

This is where you find the demographic data. It is well known that the age of the customers and their gender can significantly influence their shopping habits, just like their reactions to a certain campaign or message, just as well how effectively they can be targeted with the given product. You may modify your supply accordingly – if you find that a different kind of audience is visiting the online store than the one you originally targeted, you can invent such promotions, or you can introduce such products that are attractive to members of that audience.

You have access to five different dimensions under Audience > Demographic data, which are Age, Gender, Attraction categories, Active market segments and Other categories.


Geographical data may also be important in certain cases – a professional ecommerce store can, for example, use weather-based segmentation: they can offer umbrellas at locations where there is a storm taking place, sunglasses where there is sunshine, and so on. It is not a disadvantage either if you know in which cities and regions your products are the most popular or how the different geographical segments behave.


Acquisition reports

Acquisition reports provide more complex information, you can learn more about the acquisition-behaviour-conversion-cycle, briefly the ABC-cycle. This means that you can find out how you acquire customers, how they behave on the website after the acquisition and what kind of conversion patterns can be associated with them.

For example if you use AdWords advertisements, you can integrate their data into Analytics and follow through the complete life cycle of the customer from the first click to the conversion.



Behaviour is also extremely important – this is from where you know for example whether the visitors, the customers will return to the website. If a lot of the visitors return, but only a few of them purchase, this may mean that your products are attractive but small impulses, for example personalized discounts may still be necessary for the conversion. If the proportion of returning customers is low, you need to examine whether you re-activate them effectively, what kind of e-mail marketing campaign you pursue, whether there is negative feedback in connection with purchase completion or delivery.


ecommerce analytics user behaviour
Photo credit: E-commerce via photopin (license)



Behavioural data can be found under Audience, however, it is much more than a simple subsection of the menu. You can see how behaviour is influenced by page load speed (users do not appreciate if your site is slow, they tend to leave it even only after a few seconds), by content (how much time they spend on the site, how high the bounce rate is etc.), and you can also see how the users carry out searches (what keywords they use, how much relevant data they get as a result).



In order to be able to review conversion reports, you have to connect Analytics with the system of the online store – considering that you can apply lots of various payment modules or checkout techniques on the website. There are several subcategories under the conversion reports, for example customer behaviour, which shows what kinds of sites the customers visit, how they use the search engine – and this is from where you can pick out the information relative to cart abandonment, which may be especially important.


ecommerce analytics conversion


Cart abandonment is a problem that affects all ecommerce stores since a big proportion of the users who add products to the cart will not pay for it, will not order it in the end. There may be lots of reasons for that: the checkout process is too long, or you have not informed the customer in advance about the delivery price or you have given any other ambiguous information, or that they are just making a so-called list that they can find more easily later, when they are really willing to shop. If you monitor the numbers in the conversion reports, you will be able to significantly improve the conversion rate by the correction of the errors.

This is also where you can get information about how the individual products perform – which are the most popular, which are the ones that your visitors do not purchase, which ones do they usually buy together, and so on. It is easy to see why that is important – you can fine-tune the marketing or even the supply in case of the products that perform badly, or you can even offer in pack by default the products that are usually purchased together.


Why not to pay attention to everything at the same time?


The reason why you have to determine the indices that are actually the most important is because dealing with ecommerce analysis and optimization is an exhaustive and continuous task. Your audience, supply and site is constantly changing, so you can never declare it “finished”, or say that you do not have anything to do with it any more.


ecommerce analytics focus


But if you want to build every single data in the strategy, you will need a whole team to be able to do that – you will have to elaborate extremely complex plans, carry out the necessary modifications on the site according to these plans and, of course, to test everything. This consumes a great deal of time and resources, and for that reason it is best for you to identify the points most critical to you, and focus on those specifically.

It is naivety to think that this is an easy task for a large company that has its own online marketing and developer team at its disposal. If you want everything at the same time, you will simply not be able to draw the appropriate conclusions from the enormous piles of data.

This way not only planning and analysis will become work that can be realistically carried out, but the developer will also have an easier job and you might as well be able to run only a few A/B tests simultaneously and still end up with precise results concerning the outcome of the optimization.


tips Tips for improving the metrics numbers:


All websites are different – their content, their appearance, what they offer, the functions of the system may all influence how successful they are.

However, there are some basic policies, following which you can increase your conversion rate regardless of the audience and the technical solutions.


Reduce page load time

Customers are always impatient and this is increasingly like that in the mobile era. More and more people purchase through mobile phones, which means you have to guarantee that every single of your pages is loaded extremely quickly. If the users have to wait more than a few seconds, they will leave you and switch over to a competitor.


ecommerce analytics page speed


Provide all information

All companies are “naked” on the internet. Nowadays information can be searched easily, the users can ask for the opinion of others in a matter of seconds, they can ask questions in connection with your products or prices on other platforms. You must do your best to provide them with all necessary information so they will not have any questions at all – you have to give them an intuitive frontend, where they can easily find their way around, and to indicate such information like, for example, shipping costs on the product pages. Never let them be unpleasantly surprised because you lose customers if you do so, and thanks to public feedback, maybe even more than one at a time.


Grant exclusive offers based on behaviour

  • When you see that one wants to leave your site without having purchased anything, throw a pop-up message in front of him in which you grant a discount valid only there and then.
  • If a user regularly buys the same products in your store, send him or her an e-mail and offer regular, automatic delivery of those products without a need to further orders.
  • If one packs one’s cart full, but does not purchase, send him or her an e-mail that he or she can buy the given products with some discount.


There are infinite possibilities available and thanks to the data, you can follow precisely the behaviour of your every single visitor. All you have to do is to build such a database, to create such a system, in which you automatically target the ones showing given behavioural patterns with certain offers.

Once you know well how people use the search engine or what kind of products they like looking at, offer them products that comply with their preferences on every single page. And if you find that they are just looking around, redirect them to some of your contents which may convince them that it is worth shopping in your store at that moment.


Simplify the checkout process as much as possible

By all means use a method that requires the least possible effort from the potential buyers for making a payment. Do not lead them through five different pages, do not request their registration and do not ask even for their blood type. While providing all necessary information, within one or two pages ask for their data indispensable for payment and delivery, and then it is done.


ecommerce analytics checkout process


This will increase the conversion rate and ameliorate the user experience. A high proportion of cart abandoners do not convert in the end because the ecommerce store requires too much from them.

You should definitely ask for their e-mail address and marketing goes smoothly from there: you can find them later with individual offers and content tailored to them – but you should really leave this for later and should not want to shove everything down their throats immediately.


If something does not work, get rid of it quickly

There will be some coupons, promotions, offers and even some products that the users will simply not like – furthermore, there may even be some that are not worthwhile for you to keep for some reason.

Do not insist on keeping these, you should rather invent new ones constantly. You can invent unlimited quantities of creative customer acquiring methods based on customer behaviour, or you can find newer and newer solutions to increase the average value of purchases. However, this will work only if you have ears for the messages that the data send you.



This may seem more simple in writing, which is why it is important to emphasize that this is a permanent job: optimization based on ecommerce analytics must be uninterrupted, otherwise you will realize after a well done modification or campaign that your conversion rate is again in decline.

You have to be able to modify your online store quickly so that you can react upon unexpected situations as well (e.g. the appearance of a previously not reached major audience that has been directed to the site by an exceptionally successful viral content). You also have to train all of your systems simultaneously – your content, the ecommerce store itself, your AdWords advertisements, e-mail campaigns will all have to be harmonized in order to reach the most people and to actually make them convert – preferably several times.


ecommerce analytics coninuous work

Never stop. You should always keep an eye on ecommerce analytics…



3 replies

Trackbacks & Pingbacks

  1. […] Cart abandonment is one of the biggest challenges for the operator of an online shop. There is a good chance that those who add a given product to the cart will not buy it – at the moment the average rate is approximately 75%, but according to some estimations, in a couple of years it may even reach 90% in ecommerce. […]

  2. […] addressing everyone at the same time with the same message, you should start personalizing based on analytics and testing. Collect the necessary data about your visitors and show them specific messages which […]

  3. […] Cart abandonment is one of the biggest challenges for the operator of an online shop. There is a good chance that those who add a given product to the cart will not buy it – at the moment the average rate is approximately 75%, but according to some estimations, in a couple of years it may even reach 90% in ecommerce. […]

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published.