Treasure in the Data: Shoptet’s Journey to Analytics Maturity

Shoptet is the market leader in e-shop solutions in the Czech and Slovak markets. Thanks to its simple and intuitive interface, it has won up to one third of all online stores. It is one of the largest Czech SaaS products, and we had the honour of taking its services one level higher.

We worked together on a large project that affects how its customers’ data is collected and measured. It will therefore have an impact on around 37,000 active e-shops running on Shoptet.

Take a look at what changes we helped bring to the platform, what the challenges were and, finally, the results of the whole project!

The Brief from Shoptet

The project covered around 30 websites owned by the Shoptet platform, which help it offer a first-class experience to its clients – e-shop owners – and win new customers who want to try Shoptet. The cooperation started with the switch from Universal Analytics to Google Analytics 4. All of us across the marketing and analytics market felt this change in one way or another, and we are very glad that Shoptet chose us for this migration. The main business goal, however, was to change and innovate the process of collecting, processing and analysing marketing data.

Our main contact person in this project was the Digital Marketing Manager, Tomáš Hradecký. Tomáš oversaw the whole implementation on the client side and was the user of our outputs. Tomáš came with a brief that he needed to unify tracking across around 30 of the company’s websites. He also needed to check and fix the current setup of marketing and analytics tags. At the same time, the Cookiebot tool had to be implemented correctly to manage users’ consent to cookie processing. Last but not least, it was key to give the client clear outputs and the option to develop the tracking further, easily.

A large share of the work was also delivered by Vašek Jelen and his team from MeasureDesign.cz (organisation), who had been part of the project from its birth and brought us in as implementation experts. So we worked with Vašek on a joint project for Shoptet, so that we could pass on as much know-how and experience as possible. Vašek led and oversaw the strategic side of data collection and its subsequent processing in BigQuery using data pipelines. Our role as Archetix was to take care of the complete data layer implementation, Google Tag Manager and Cookiebot on all 30 Shoptet websites. We also handled project management.

What Did We Struggle With?

The most complex part of the project was that we worked with up to 30 websites. But that is our speciality, because we often come across it in practice, we have a lot of experience with it and a clearly defined system for dealing with it efficiently. It often happens that the individual websites were created in different years, on different technologies, by different development teams, and changing them can often be problematic.

In larger implementation projects, the biggest challenge is also getting all parties to agree on the goal and scope of the project. Shoptet is already a large and growing company with established development processes and many employees. It is understandable that implementing changes is not easy in such cases. When things start changing across the board in such a company (here in the context of changing the tracking on all websites), it affects a large number of people and processes. Even though a brief may seem very trivial, carrying it out is exactly the opposite. For example, the brief “set up 4 events on 30 websites” is very demanding precisely because it means 4 events across 30 websites, and that is no longer so easy.

Here we always try to adapt to the client’s processes. The client can appoint one responsible person and communicate all requests internally. Or we can get a direct connection to the development teams and consult the required website changes with them directly. At Shoptet, it was a combination of both approaches.

Project Goal

After several introductory analytics workshops with the client, a business brief was created. It set out the basic outputs of the project. The goal of the project was to unify tracking across all websites within one year, from the most important to the less important ones. It was also required to comply with the cookie regulations, which meant including a correct implementation of Cookiebot. On the key websites, the task was to add a data layer to the tracking and archive the historical tracking setups. Thanks to the data layer, we can get key information more accurately. It will be used further in product analytics, which Shoptet is currently working on. At SaaS companies like Shoptet, product analytics serves the long-term development of the product and keeps it competitive on the market.

“The introductory workshops with Vašek’s and Marek’s teams were exactly what we needed to make sure we were on the right track. Their professionalism and thorough attention to detail convinced us that we had chosen the right partners for our journey to setting up our whole web analytics. I have been excited about our cooperation from the start and I believe that together we will achieve all our goals.”

Tomáš Hradecký – Head of Digital Media – CEE

After the business brief, once the client had reached agreement across all departments and got the final GO!, the decision was made to start the implementation. The initial specification was created, which from the very beginning counted on the solution being used on multiple websites. In other words, it would be used in a universal way, so that even as new domains or new projects were added, everything would remain clear and simple.

In general, Shoptet manages a large number of lead-generation websites, plus other parts related to the administration of the e-shops themselves.

Splitting Shoptet Websites into Categories

For the lead-generation websites, we proposed using a data layer to unify the tracking of several key events that are common to most SaaS applications. The first such event is the start of a trial (the user’s first free registration in the SaaS application). This person can then become a subscriber, i.e. a paying customer, install add-ons, buy a paid plan and so on. Like other SaaS applications, Shoptet also uses forms on its websites. So another unified event was tracking forms for contacting support or subscribing to the newsletter. Based on this, over several iterations, a generalised, so-called semantic data layer was created.

The websites were then split into three categories:

• priority = should have full tracking, including the data layer, Cookiebot, correct handling of user consent and data processing;
• important = get everything from the priority category, but without the data layer;
• marginal websites = are available only to Shoptet clients, do not need a broad implementation and will have basic tracking.

A big topic in projects of this size is that if a company has several websites, there is a good chance that each will run on a different technology, and that was the case here too. A large part of our role was therefore communication with developers. We held several workshops with them and found out that some of the websites run on WordPress, some on a JavaScript framework and some on yet another platform such as Zendesk. Our advantage was that we knew what the output of the implementation should look like, so all that remained was to agree with the developers on how to implement it correctly on each system. At the same time, a number of workshops took place across the marketing team.

During the project, Shoptet changed its marketing agency, and we had to take that into account as well. We carried out the implementation of conversion tags and remarketing tags again, and did everything so that the tracking would be sustainable in the long term and anyone could find their way around it. Including careful documentation.

Results of the Shoptet Project

To manage the whole project, we chose the Asana tool. It allowed us to communicate smoothly and bring the project to the finish line.

Our main role was to handle the implementation, testing and documentation of the whole project. The output, after a whole year of work, was the unification of events across the other projects, regardless of the technologies used. At the same time, we took user consent into account in the tracking, so that in the reporting prepared by Vašek Jelen, the client can work with two datasets – the so-called anonymous dataset (contains no user identifiers and is used to determine the total volume of interactions) and the so-called consented dataset (used to evaluate specific campaigns and the interactions of users who gave their consent).

We then trained the whole team on the client side and handed over the implementation together with the related documentation.

“On top of that, we acted as expert consultants for the client on Google Tag Manager and Google Analytics. Our cooperation turned into a long-term partnership, because the client decided to outsource our services, know-how and methods. We are very pleased that this leader of Czech and Slovak e-commerce chose us, Archetix, as its experienced mentor in the analytics field.”

Says Marek Čech, director and lead analyst.

Final Words from the Main Project Stakeholders

The project is currently at the stage where procedures to ensure data quality are being deployed, both at the data collection level and at the processing level. Control schemas are being created that help define what the data layers should look like and provide regular alerts and data checks.

“After the past year of our cooperation, I would like to thank both teams for the great cooperation, expertise and support in running the whole project. Together we faced challenges and achieved results that exceeded our expectations. This project not only took our analytics to a new level, but also laid solid foundations for our future cooperation. I look forward to what further cooperation will bring, and I am convinced that the best is yet to come.”

Tomáš Hradecký