fifty-five helps brands to make optimal use of data and technology to develop more relevant, innovative and effective marketing strategies. Their team includes specialist digital consultants, and data engineers, as well as experts in tracking, advertising technology and media – a synergy that enables the data company to guide the marketing teams of major brands in a variety of sectors throughout the world.
Jérémie Le Tarnec is fifty-five’s resident expert in Digital Analytics and AT Internet. Prior to fifty-five, he worked in the digital banking sector (BforBank) on acquisition, e-merchandising and analytics missions on web perimeters and mobile applications – putting his expertise at the service of advertisers. Today, he trains and coaches around 25 fifty-five consultants, mainly on Analytics and CRO (conversion rate optimisation) projects for some fifteen clients in the banking, media, retail, tourism and automotive sectors. His role as a key advisor leads him to regularly exchange with editors of customers’ analytical solutions, especially AT Internet, to follow product roadmaps and pass on this knowledge to consultants. In this interview, he reveals his favourite features in the Analytics Suite!
Which AT Internet functionality is essential in your opinion?
I’d say Explorer, because that’s the entry point for operational data management. It is the first tool used by fifty-five experts when they start an assignment with a client using AT Internet. With Explorer, they have quick and easy access to key metrics of digital business and a macro view of web data and mobile applications. It is a first level of granularity on the global audience, content, and traffic sources, while providing contextualisation of the data.
The other interesting aspect of Explorer is that it responds to very specific issues. In the tourism sector, for example, monitoring cross-sell performance through self-promotion of a product is very simple. It is easy to obtain metrics such as clicks, impressions and cross-sell conversions. To give another example – the analysis of the internal search engine makes it possible to easily find the keywords typed by visitors on a website and quickly identify possible flaws and areas for improvement in a product or service.
The Navigation module is also very relevant for advertisers – who are often looking to follow the navigational routes and analyse all the steps in the visitor’s journey following the home page entry point. Now it is possible to accurately and easily identify throughput rates and leakage points.
Some of our customers ask for user retention analyses, especially those whose business is very app centric. For example, you can find out how many people use an application after it has been downloaded from an app store. Thanks to Explorer‘s user-centric analyses, it is possible to analyse this with retention models. AT Internet allows them to reconcile visitor-specific paths between app and web with a comprehensive view of the cross-device and cross-assets part (based on identified visitors). I’m thinking in particular of the Venn diagram visualisation, which is very useful for our teams, who can easily import them into presentations.
Another major plus for Explorer is the speed and efficiency of running reports and analyses in an interface that gives almost immediate access to data, without getting lost in too much depth of analysis, a problem you can encounter with other analytics providers.
Which tool is particularly suited to advanced data mining?
We mainly use Data Query, the complementary component of Explorer. Its use requires some training, but it is the tool that, in my opinion, makes AT Internet’s Analytics Suite a standout solution! It is reserved for certified users who are familiar with the world of AT Internet and fully understand the nomenclature of sites, dimensions and metrics of the product.
For example, for a client in the media industry, we used Data Query to measure the performance of video content through A/B testing to optimise the user experience. To quantify our tests, we make direct use the AT Internet solution where we recover all the performance of the variants using native connectors (reference to Kameleoon). Once in the query tool, the test data can be cross-referenced or associated with a wide range of custom indicators (e.g. conversion rate, click rate, read rate) and thus obtain a true multidimensional analysis to determine the winning variant.
To take another example from the banking sector – Data Query is used for the analysis of complex paths. For issues relating to the subscription of 100% online banking products, the tool enables all points of contact to be reconciled, even when there has been a break in the process: application to open an account on the site, return of supporting documents… In other words, many steps that are suspended and performed at different times. The coherence of this type of analysis requires of course a reliable collection, with the presence of a unique identifier (order id or id of an identified visitor).
Data Query is also used as a query tool for dashboarding. In practice, different datasets can be created in the tool and exported via APIs to dashboards or data visualisation tools.
How does AT Internet’s solution meet very specific business needs?
Analytics has become more democratic in recent years. We can see this at fifty-five because we are receiving an increasing number of requests for training in analytics from cross-functional professions such as product owners, product managers, e-commerce managers, acquisition managers in the media sector and even decision-makers (C-Level). All sectors are in demand, whether they are pure players or traditional players whose main activity is offline.
To address this, we use AT Internet’s dashboard tool, whose dataset editor was redesigned and improved last year. It offers the various data user businesses a type of visual and graphic information that is well suited to the need for simple reading and understanding of the data. This data visualisation tool allows them to easily retrieve analyses without spending too much time on it. Once dashboards are created, our experts can contextualise multi-site data, structure and prioritise information and share it easily. The degree of customisation is also very advanced, there are a multitude of representations and themes that produce their effect when the dashboards are delivered to the customers. For example, a Product Owner will be able to consult the KPIs of the throughput or click rates of their insurance simulators every morning, to prioritise the features to be developed or optimised in their product, without having to connect to the AT Internet interfaces. To sum up, I would say that it is the ideal training tool for non-analytical profiles who want to follow specific KPIs adapted to the complexity of their business on user-friendly dashboards
Among the latest Data Science-oriented innovations, which one stands out for you?
At the end of October, I had the opportunity to test the AXON insights prediction module on anomaly detection and contribution analysis. I was very impressed by the power and value provided by this feature based on machine learning algorithms.
Simply put, the module makes it possible in almost a single click to see where a decrease or increase in traffic considered as an anomaly comes from. The system automatically finds and identifies the exact page affected by the data variation. The time savings are enormous because it cuts out tedious investigative tasks. Customers, even without analytical expertise, are also able to find and explain possible problems themselves. At a media client analysing video content, for example, we had identified a significant drop of 15,000 visits. With a few clicks, the main page affected, and its provider were identified. Having a global vision of the market for analytical solutions, I can say that it is the most complete in the field of machine learning. And that’s a view shared by the other employees of the agency.
Anything to add?
Last year, the AT Internet team came to present us with the product roadmap, and I must say that their commitments have been kept and even exceeded in terms of development and respect of the schedule, notably thanks to the latest releases [Data Query 3 and AXON]. Our teams are satisfied with the product, and it’s really nice to see that a French solution offers such a complete analytical offer. Our expectations are now all the greater as the user experience has become more complex over the years. And following the recent announcements made at the Digital Analytics Forum, we are looking forward to the next developments around segmentation and custom metrics!
More about fifty-five
fifty-five helps brands make optimal use of data and technology to develop more relevant, innovative and effective marketing strategies. Strategy consulting, data architecture, media consulting, customer experience… The complementary nature of its expertise enables it to guide more than 200 prestigious brands in a multitude of sectors. Founded in Paris in 2010 by former Google Europe executives, the company has quickly established itself as a pioneer in data consulting, and today has more than 250 employees in Paris, London, Shanghai, New York, Hong Kong, Geneva, Shenzhen and Taipei. fifty-five is part of You & Mr Jones the world’s largest Brandtech™ group.