Read how T-Mobile is building a valuable data culture in collaboration with Function
T-Mobile is a major telecommunications provider, well recognisable by its magenta capital T. Together with its other brands Tele2, Ben and Simpel the organization serves millions of users in Europe. The market-leading brand is active in both the business-2-consumer and business-2-business market. Function and T-Mobile started their collaboration mid 2022.
T-Mobile is invested in its data-driven marketing capabilities, and establishing a future-proof architecture. As a result of market developments over the last years, such as browser tracking preventions - and new technical opportunities to collect qualitative online data, such as server-side tracking, the current standard of collecting online data did not fit T-Mobiles ambitions and strategy anymore. Therefore, mid 2022 T-Mobile took the leap to be one of the early adapters in the market to innovate its data collection towards an infrastructure that is fit for the coming years (or buzz term: ‘cookieless era’). Realisation of this infrastructure is needed to maintain the highly valued trust in data quality, so teams can confidently serve business decisions that grow digital business operations.
Our objective is to migrate all media data pipelines (the data that is sent from T-Mobile to platforms like Google Ads, Facebook and Snapchat) for T-Mobile and Tele2 from a client-side implementation towards a first-party, server-side implementation. Working with media related platforms means the involvement of a lot of internal and external stakeholders, and involves critical GDPR consent components.
What we're doing
The technical engine
- Established the data fundament for real-time ROAS optimisation: In order to estimate how successful a paid campaign is and how its bidding should be tweaked, we have to be able to uncover what the campaign contributes in terms of revenue. We built an infrastructure that has the needed real-time capability and reliable data quality that makes attribution models accurate and meaningful.
- Removed any future dependency on third-party cookies in the infrastructure: Third-party tags have been replaced or enriched by various conversion API’s.
- Data collection using client-side GTM and server-side Tealium products: We moved from 100+ third-party tags to 12 first party tags, meaning simplified governance and better website performance.
- Simplified data collection: Structured events with similar definitions towards single – well defined – events, validated the implementation and accuracy, and properly documented the setup for all stakeholders to reference.
The organisational engine
As in all data projects, there were some unknowns before our agile team started. The exact stakeholders, possible impediments, current status of the infrastructure and concrete potential value were unknown to an extent. Therefore, we apply our self-developed framework for marketing data initiatives that allows us to onboard and clarify the current status fast, deliver results early, respond to impediments quickly and involve all different stakeholders. You can find our framework elsewhere on this website, but in short it consists of 4 iterative steps, performed simultaneously during ongoing agile sprints:
To elaborate on how this leads to a quicker time to value, we zoom-in on a real situation:
- During sprint 2 we discovered an impediment on an external technical capability. Since there is a high amount of technical requests, choices have to be made in terms of priority, and our requests have a time period before they can be picked up. In case of this specific request, that period was expected to be approximately 5 months.
- During sprint 3 we performed a research spike to find a workaround. In our sprint review we proposed our findings with our stakeholders.
- During sprint 4 we implemented the temporary workaround.
- 4 months later the external team realised the user story, with minimal delay impact on the aimed objective.
Currently, this agile team is responsible for collecting and delivering clean data, prioritising and delivering requests from stakeholders, still using the same agile best practices.
To us, this is what building a data culture means. We believe that everything you give attention to, tends to grow. So if an organisation’s process towards working with data is given the attention needed; facilitating people with the right approach and tech, data makes work easier and decisions better: it becomes functional.