At the "Digital Analytics: Adapt or Regress" breakfast, organized by Data on Duty, a panel of experts and professionals gathered to decipher the meteoric evolutions shaking up digital analytics. Moderated by Éric Dumain, co-founder of Data on Duty, the meeting underlined an essential point: in an environment where uses and technologies are evolving at unprecedented speed, adapting is no longer an option, but a necessity.
Around the table, leading voices in digital analytics - including Julien Coquet (consultant and trainer), Jean-Marc Vandenabeele (representing a major luxury goods group) and Éric Bricier (formerly of Boardriders) - shared their experiences and their vision of the business in the face of technological and regulatory change.
A crisis of confidence in data
From the outset, our findings were clear: trust in data is eroding. According to studies by Gartner and Forrester, many decision-makers today doubt the quality of their information. Data considered 80% reliable may turn out to be unusable... if you don't know which 20% is wrong.
The Data on Duty platform, which specializes in tracking compliance analysis, often detects discrepancies of up to 80% between active tracking and the tagging plan. These discrepancies render a large proportion of the data collected useless, a phenomenon described as "disconnected data".
There are several reasons for this loss of reliability:
- Constantly evolving regulations (RGPD, CCPA, etc.), which are tending to unify. For fear of sanctions, some companies excessively limit data collection - sometimes even the data they are entitled to use - which deprives marketing of essential information.
- Inadequate tracking management, especially after years of using Tag Management Systems (TMS). While these tools are not bad, their rules-based approach is showing its limitations in the face of today's demands.
A new paradigm: doing more with less
The speakers described a new context marked by :
- The regular arrival of new European rules and the frequent updating of existing ones.
- The growing impact of Artificial Intelligence on tracking models.
- An ever-increasing workload for digital analysts.
- Time and budget constraints force teams to do more with the same or even fewer resources.
In this context, one question dominates: how can you do more with less without the right tools? Unlike Business Intelligence or CRM, where dedicated platforms are the norm, Digital Analytics still relies on Excel or Google Sheets, or on rudimentary or unsuitable solutions.
Top ten challenges facing digital analytics teams
The experts have identified ten major "pain points":
- Increasing complexity and diversity of the business.
- Lack of management involvement (or very high demands once they realize the strategic impact of analytics).
- Move from simple e-commerce reporting to multi-level insight production.
- Increased workload (tracking, governance, privacy, data analysis, CDP).
- Stable or declining human resources.
- Mixed governance: group strategy and local implementation.
- Non-compliance with or misapplication of governance rules.
- Desynchronization between marking plan evolution and release rhythm.
- Multiplication of marking tools, complicating deployment and acceptance.
- Maintenance of high-performance, compliant tracking, albeit "invisible" to internal teams.
Participants also spoke of the sometimes negative impact of AI: over-ambitious promises can generate unrealistic expectations, driving some companies to downsize. Others pointed out the gap between the importance of tracking and the investment actually made. Finally, many insisted on the need for time to produce knowledge and not just graphs.
AI: between marketing promise and reality
Julien Coquet clarified an essential point: generative AI (often used to create content) represents only a small part of what AI actually covers. In digital analytics, 95% of usage is in Machine Learning, which automates data processing and analysis. AI should therefore be seen as a tool for automation, detection and alert, and not as an autonomous intelligence.
Several participants shared disappointing AI experiences, where results didn't match promises, often because solutions weren't tailored to the specific needs of analytics. Some feared that AI would reduce the role of analysts to mere "tracking experts", whereas their primary mission is to create value. Ethical concerns were also raised, notably the energy footprint of AI.
Data on Duty's integrated approach
Unlike pure marketing approaches, Data on Duty has integrated AI from its conception in 2020 as a business tool. Its platform uses AI to :
- Generate object-oriented tagging plans, capable of intelligently reconstructing a site's tagging.
- Validate and monitor data quality, recreate failed visitor journeys and suggest relevant alternative paths.
- Detect regressions and propose automated alerts, based on success and failure patterns.
A new brick, Enforcement Manager (scheduled for late 2024), will even automatically correct erroneous tags. The aim is clear: to free analysts from repetitive tasks, so that they can concentrate on analysis and value creation.
Impacts already visible
Jean-Marc Vandenabeele testified to the positive effects of this evolution: increased business maturity, strengthened governance and improved data quality thanks to an omnichannel approach. For his part, Éric Bicier recalled that the COVID crisis had elevated digital analytics to a strategic position, arousing the interest of senior management but also putting considerable pressure on teams with limited resources. His response: automate reporting and train operational teams (web merchandisers, marketers, SEOs) to become more autonomous.
The debate on the siloing of roles (tracking expert, CRO expert, analyst) has revealed varying practices across companies. However, all agree on one point: avoid reducing analysts to "data plumbers". The high turnover (less than 18 months in post on average) clearly illustrates the gap between the stated mission and day-to-day reality.
Data Layer and industrialization: strategic levers
The Data Layer is seen as a major asset, especially for multi-brand groups. Jean-Marc shared his approach: 80 to 90% of the framework is centralized, with the possibility for each brand to add its own specific indicators under quality control.
Industrialization (doing it once and applying it everywhere) andautomation (making the whole process more fluid) have emerged as essential levers for saving time, accelerating time-to-market and reducing the burden on teams. Without them, companies remain in a reactive mode, discovering tracking flaws too late.
Thanks to Data on Duty, it is now possible to generate thousands of verification rules in a matter of seconds, and to process complex tagging plans in a matter of hours to days, instead of weeks/months. This efficiency transforms unitary control into intelligent, regular monitoring, capable of detecting deviations before they become critical.
Change management: impossible without the right tools
All the experts have stressed that, in such a demanding environment, change management cannot succeed without dedicated tools. Data on Duty is positioned as a shared collaborative space, offering a single repository (tagging, data layer, privacy) and access adapted to the various contributors, from administrators to simple viewers.
Ten key actions to adapt and progress
The session concluded with a series of priority actions:
- Involve C-levels in the analytics strategy.
- Set up cross-functional teams whenever possible.
- Deploy a centralized repository for tracking and compliance.
- Industrialize processes to improve speed and efficiency.
- Free analysts from technical monitoring tasks to refocus on analysis.
- Standardize implementations.
- Automate tests and set up intelligent monitoring.
- Integrate data quality into corporate culture.
- Track and communicate time savings achieved through automation.
- Complement online tracking with a CDP to cross-reference data.
Discussions also highlighted the importance of involving IT teams and developers at an early stage - particularly for mobile applications - by integrating tracking into CI/CD processes. One of the panellists from the B2B energy sector illustrated how a customer-centric approach and the leveraging of data have helped to prove the performance of B2B marketing and to obtain an increase in digital budgets.
Finally, digital analytics is undergoing a profound transformation. In this complex environment, industrialization and automation are essential. Platforms such as Data on Duty enable analysts to refocus on the essentials: analyzing, recommending and generating value for the company, based on reliable, high-quality data.
