The analytical minimum or what to measure on the ground floor and what is the baseline measurement for us

Analytics and working with data have been a big topic in recent years, which is of course a good thing. Measuring and evaluating data is more or less there for anyone who has a business tied to an online environment. Both an owner of a few-page blog and an owner of a million-dollar e-shop should have a basic overview of what’s happening on their website or e-shop. What to measure and how to work with the data?

Project life cycle

Every project has a life cycle. 

It all starts with a kind of “kick-off” – that’s when you usually invest a few hundred in each of the marketing channels and get first hands-on experience with the systems. 

But right after that, you start thinking about investments = that’s when analytics comes into play. You need to know what customers are doing on your website, what marketing activities they are coming from, and you also need to define goals that are based on realistic foundations.

In some cases, you’re getting to the stage where you’re dealing with planning for several years ahead, you need documents for getting a loan, or you need to justify your marketing budgets to an owner or investor. 

And here, you need to have at least basic data for managerial and operational decisions.

Content web analytics

Content websites are quite different from e-shop solutions in terms of analytics. The biggest difference is the lack of direct purchase as the main conversion. Content websites can then be divided into two main types:

  1. a website offering content (and possibly wants to make money from advertising or partnerships)
  2. a website built on lead generation (enquiry forms, calculators)

Within the baseline measurement, we consider it crucial to determine the website traffic sources (organic, social networks), for which a basic insertion of the measurement code is usually sufficient. For larger projects, it is then suggested to look more at marketing channels and attribution of marketing investments (whether last-click brings the right visits) or to look at content grouping (grouping content by post type, category, tag). With content grouping, you can then find out in a few clicks which posts in the Holiday category (when it’s not clearly visible from the URL) are performing best.

For content websites that aim to generate leads, it’s a little different. Typically, we can measure some form of interaction (mortgage cost estimator, pre-quote, cost calculator, product configurator) and we are concerned with who is completing those conversions and we are possibly following up with that user. In the case of long purchase processes (heat pumps, building materials, cars) then anonymous user IDs can also be implemented as part of the baseline measurement. In this case, even after a long time, you will be able to pair a user to a single lead and look at their entire conversion path.

What objectives do we recommend to measure?

Some objectives are specific to content websites, but most of them can also be used for e-shops, with e-commerce solutions measuring the entire shopping process – cart, funnels, etc.

  • main conversions
    • registration or subscription (payment)
    • downloads (e.g. product sheets, especially for B2B)
  • micro-conversion
    • article readership (combination of static time or number of characters in the article + scroll depth)
    • engagement with video – launching, playing a part of the video
    • clicking on a photo
  • web search – serves as a possible inspiration for new content creation or feedback on something customers can’t easily find

E-shop analytics

With e-commerce, the main thing is to measure what I actually use. Often clients come to us with measurement set up for almost everything – such measurement can be expensive, marketing systems can have trouble deciphering the data, and of course, navigating through the data is also difficult. What I measure should match the analytical maturity of the organization or project.

TIP: If there are not many transactions in the e-shop, but the advertising systems need more conversions to optimize campaigns, you can take a step back in the purchase funnel and extend the measurement of, for example, cart entry and filling in the address. If even that’s not enough, jump to the “add to cart” measurement.

In addition to the objectives we described above for content websites, you can also measure e.g:

  • cart and the whole shopping process (funnel)
  • registration to the e-shop, newsletter subscription, discount coupons (I will also learn from what value the discount is interesting for the customer)
  • searching (according to the volume I am able to estimate the stock or the potential to stock new products)
  • booking and adding to the wish list

In the long term, we are concerned with the actual profit (the amount that remains after deducting transport, returns, gifts, labour, taxes and other costs), i.e. the so-called net margin that actually remains in your account. It is important to note that free shipping is only free to the customer.

We do not recommend measuring returns, cancellations and warranty claims within Google Analytics – we usually handle this in a separate measurement outside of Google Analytics. Similarly, it doesn’t make sense to import the data back, but to do your analysis of individual channels outside of Analytics (for example, through transformation and data handling).

We would also avoid measuring product impressions across categories (what is viewed but not clicked). You can’t do anything with this data manually – you need quite complex automation and machine learning to change product listings according to the data.

Evaluating the objectives

In all cases of baseline measurement, we recommend that you evaluate all measured objectives with a financial amount (e.g. reading an article 0,3 EUR, downloading a PDF 1 EUR, etc.). This will make it possible to compare them with the cost of website performance. The technique you use to calculate the value of the objectives can vary from your own feeling, to an expert estimate, or to an actual calculation.

For example: with a content website, you know what the click-through rate is and what the average profit per click is. Based on that, you know how much a single visit will bring you and how many you need to meet your objectives.

What do we mean by baseline measurement and how do we work?

Usually, we don’t let the client go any further before they tell us what they want to do with the data. Measuring anything without this information will be problematic not only because of the difficulty of setup and possibly zero usage, but also with regard to internet user privacy, so-called ePrivacy – putting a few remarketing codes on your site that run without user consent, in the “if by chance” style, is already frowned upon, and you are exposing yourself to potential penalties due to the European Union’s ePrivacy Directive, which will probably come into force next year.

On the baseline measurement, we recommend the following:

  • everything deployed via GTM – we always transfer everything to our clients anyway, so that the measurement is controlled
  • collecting traffic on landing pages – often a basic GA measurement code is enough (or a smarter GA4 code that can also measure scroll depth, pdf downloads etc).
  • two or three remarketing codes that you will use (usually Sklik, Ads, FB pixel) so that you can properly and legally collect audience data
  • cart measurement for e-commerce (possibly Enhanced Ecommerce)

With mobile app analytics, even baseline measurement is significantly more complex – measurement doesn’t work well by itself and needs more care.

What to take away from the article:

“No matter what you’re measuring, the biggest shift will be if you look at the data regularly. Rather measure fewer things, but track the progress of your chosen metrics and act accordingly.”

Regardless of the measurement setup (right or wrong), without looking at the data, the measurement becomes useless – something is measured, but no one works with the data, no one uses it for reporting or optimization, no one discusses the data, and there is no effect from it. It’s kind of like Schödinger’s cat – until you start working with the data, it can be both right and wrong.

Our point is that data in whatever form should help you streamline processes and resource utilization. These efficiencies can come in the form of both time and financial savings or increased turnover. The resources thus gained can then be used to further development, innovation or support the community in the company’s immediate vicinity.

The biggest mistakes in analytics

We have already written about the most common mistakes in web analytics. But let’s highlight the most important ones:

  1. No one works with the data
  2. I have basic measurements and my website works on Javascript (e.g. the page does not reload, there are only virtual pageviews)
  3. Lack of working with step funnel (e.g. through the cart)
  4. Everything is measured without visitor consent (more of a legal thing)
  5. Measurement remains active even on development versions (old, dev.) and for large projects, this can skew the data quite a bit
  6. Personal data is sent to various systems (for example, after a password reset, the email remains in the URL, which is recorded to Analytics)
  7. Payment gateways and booking platforms overwrite visit sources and attribute conversions for themselves (fortunately this is getting better and people are excluding it)

The topics above are dealing with the digital analyst position on individual projects which can be external or internal.

Scroll to Top