Product Management in the Era of Data and AI
David Spire joined Converteo in September 2023 as a partner to lead a new practice focused on product management. Jérémie Lévy, currently Managing Director at Converteo, has previously overseen product teams for several years in major companies, particularly in the telecommunications sector. David and Jérémie discuss the reasons behind Converteo’s decision to establish this new product management consulting practice and outline the vision the firm aims to promote in the market.
Product management is sometimes seen as a “startup thing.” How do we educate people about this expertise today?
David – “First, we should recognize, without idealizing, the enormous contribution of startups and scale-ups over the past fifteen years. These companies have brought innovative products to market at sometimes astonishing rates, creating significant methodological and cultural shifts by refocusing on user problems, demands, and behaviors. This movement has made conventional companies aware of the urgency to reorganize in order to produce superior features, service levels, and experiences. Successful startups have all excelled in product management, achieving much of their growth through success in user adoption, usage, and retention because their products were outstanding!”
Jérémie – “To think that product management only emerged with startups is obviously incorrect. Looking at just the past twenty years, with the digitization of many services, it’s been primarily very large companies that have designed and deployed digital products adopted by millions of users. Your bank’s smartphone app, your telecom provider’s app, your energy supplier’s customer portal – these are areas where companies, alongside startups, have dealt with continuous user feedback collection and an explosion of usage data, two key factors of modern product management. They simply did it with a historical context – often referred to as legacy – which can sometimes represent a real technical, organizational, and cultural debt.”
But it would be a mistake to pit startups against large corporations: it is undeniable that new players have introduced products that bring radical innovations in both features and user experience. The time-to-market has been faster, and product evolution has been quicker.
The numerous successes of digital products from these startups have generated significant interest and recognition for their Chief Product Officers, their methods, their tools, and the organizational frameworks established within these companies.
At the heart of this revolution in product management is data. What is your view on the importance of data issues in product management?
JL: “In ten years, one of the most striking changes for me has been the decline of the so-called ‘waterfall’ methodologies and the rise of the agile era. Although agile methods sometimes face a form of ‘dissolution’ when interacting with still too rigid corporate cultures, fundamentally, agility has won the cultural battle. It is now the most desirable framework for product management, and rightly so.
Our analysis at Converteo is that data has been the main driver of the advancement of agility. What do you ultimately gain by moving from waterfall to agile if not data? You bring smaller product changes to users more quickly, which allows you to learn in real-time and at market scale. And this data can, in turn, be the foundation for better decisions on future development roadmaps.”
DS : “In theory, product management departments today have understood this paradigm. In practice, deriving all the lessons remains extremely complex. Product managers and product owners must handle data from highly diverse sources: internal product usage data, A/B tests, customer interactions with customer service (CS), advertising performance data, web comments, competitive data…”
Upon receiving all these data streams, it can be overwhelming! Teams often lack either time, expertise, or both to sort through, structure what is important, identify gaps, and make the best use of data in relation to the company’s business objectives.
Data science and data engineering, data analysis, and data visualization: these skills are now at the core of product team needs, but they are rare, expensive due to high demand, and sometimes challenging to integrate into existing product teams without the right cultural and organizational adapters.
The demand for data in product management also stems from a recent refocus, post-COVID, on the business performance of the product. Is my product economically viable? How does it fit into the company’s operational ecosystem?
The centrality of the user has been absorbed by organizations, and a new phase opens up, dominated by issues of efficiency, performance, and frugality. At Converteo, we audit the data maturity of product teams in this perspective of alignment and integration with the business through a proprietary analysis framework. This approach helps our clients better understand their starting point and the efforts needed to drive their teams towards greater excellence.
Comment les entreprises peuvent se préparer à répondre à ces défis massifs, à la fois techniques, business et organisationnels voire culturels ?
JL: “We see similar needs among many of our clients. Firstly, there is a need for expertise around data profiles, supported or complemented by leaders who can reconcile an end-to-end vision of both data and product. But perhaps even more pressing, there is a desperate need to streamline interactions and the product vision among numerous stakeholders—business, tech, and business—whose interests are not always aligned. Creating environments that facilitate data flow between these parties is an absolutely strategic challenge for the success of product teams, and this often involves the right processes and tools.”
DS: “We are starting to concretely feel the considerable leverage that artificial intelligence can represent for a company wanting to continue investing in and focusing on product management, with new roles like AI Product Managers and Data Product Owners. As in other functions within the company, AI should generate significant productivity gains in performing lower-value tasks. I recommend, if you haven’t already, asking your favorite LLM to write you a test specification; you’ll see, there is a before and an after!”
Use cases are also numerous in all tasks related to data product management: querying a database in natural language, soon by voice, and receiving in return a graph, a data visualization, useful for decision-making, is now possible with the AI solutions from major providers. AI for “product ops,” in other words, we are there.
Even more exciting, I believe that a great source of value lies in AI’s ability to access and make sense of data that is almost otherwise inaccessible. Take a source of spontaneous customer feedback collected on the web: if you sell a consumer product, your corpus potentially consists of millions of customer comments scattered across numerous review platforms, forums, and social networks.
Analyzing such volumes, applying sentiment analysis, measuring the progression of certain conversational items over time, comparing with competitors… and sharing all this material in real-time within the fluid environments Jérémie mentions: we are now capable of doing this thanks to AI, with our proprietary semantic analysis technology at Converteo.
New data sources are becoming accessible, and above all, whether new or old, these sources are more easily exploitable, integrated into product management operations more quickly and simply. Well understood and mastered, these AI applications already provide significant efficiency gains for product management teams: we are there, and it is part of the key missions we are starting to undertake with our clients. It is our ambition and our commitment to support all stakeholders who wish to advance in this direction.
