Borrowing the Product Roadmap: What can HR teams learn from how product managers use AI to launch new products

An HR team spent months designing a new performance management framework. They consulted the relevant stakeholders, studied the industry practices and launched it across the organisation. 

But six months later, when HR looked at the statistics, the adoption rate was very poor. The managers were just adding scores to the system, but were not having improvement conversations with their team members, while the employees found the entire process too cumbersome. 

Sound familiar? This is one of HR's most persistent and most frustrating problems. They invest significant time and resources in designing people programmes, only to discover after launch that the design missed something important. The feedback that would have fixed the programme before it went live was available but was not collected in time, or in the right form, to change the outcome.

Product managers face an identical problem. They build products that miss the mark, launch features that users do not adopt, or invest in roadmap items that solve the wrong problem. But over the last couple of years, the best product teams in the world have used AI to close the gap between what they build and what their users actually need.

HR teams need to borrow this playbook. 

This blog examines exactly how product managers use AI to build better products, and what HR teams can take directly from that approach to design, test, and launch people programmes that actually work.

The product manager's problem is HR's problem

Product managers and HR leaders are solving the same problem in different domains.

Both design experiences for a specific group of people, make decisions based on incomplete information about what those people actually need, launch at scale, discover the gaps only after launch, a are under pressure to give instant results.

The difference is that many product managers have built a systematic capability for closing this gap using AI. They use it to understand their users, test assumptions before launching, and track adoption after launching.

But right now, HR teams do not use these processes. At least not consistently. The gap between how product teams and HR teams use data and AI to design their respective products is one of the most consequential inefficiencies in modern people management. But…it is entirely closable.

How product managers use AI: The playbook HR needs to borrow

Successful product managers have built a systematic, AI-powered approach to understanding their users, prioritising their roadmap, and validating their decisions before launching them at large scale. The playbook has four distinct stages. Each one has a direct HR equivalent.

• Stage one - Deep user research at scale: Before a product manager builds anything, they want to understand the problem they are solving. They use AI to analyse large volumes of user feedback, support tickets, behavioural data, and market signals to identify where users are experiencing the most friction, what they are trying to accomplish, and where existing solutions are falling short. This analysis would take a human researcher months, but AI does it in days.
  ‍
  The HR equivalent is understanding what employees actually need from a people programme before designing it. HR teams that use AI-powered analysis of existing data alongside stakeholder consultation are significantly better positioned to identify the real gaps before designing a solution. AI-powered analysis of existing employee feedback, pulse survey data, exit interview content, and engagement signals can give HR the same depth of user understanding that product teams use to drive roadmap decisions.

• Stage two - Assumption testing before building: Product managers do not build full products before testing their core assumptions. They build the small versions of the product, test it with a user group, and use AI to analyse the results before committing to the full build. This approach is designed specifically to catch the assumptions that would have negatively affected the adoption after a full launch.
  
  HR teams need to adopt a similar approach. For instance, an HR team designed an onboarding programme with the assumption that the primary gap in the process is information delivery. Instead of immediately adopting this process for all the new joiners, they tested it on a small cohort of new joiners. The AI analysis revealed that their assumption was incorrect and that the actual gap was in social integration. This saved the HR team from wasting months before discovering the issue.  

• Stage three - Behavioural data over stated preference: Product managers have learned, through hard experience, that what users say they want and what they actually do are frequently different things. They use AI to track user behaviour directly, measuring what people do inside the product rather than relying on what they report in surveys.
  
  HR teams rely almost entirely on engagement surveys, focus groups, and stakeholder consultations to capture what employees say about their experience. They do not capture what employees actually do: where they disengage from a process, which parts of a programme they skip, which tools they use versus which ones they were given. AI-powered behavioural analytics can give HR the same kind of direct usage data that product teams use to drive design decisions.

• Stage four - Continuous iteration based on adoption signals: Product managers track adoption signals continuously after launch and use AI to identify where users are struggling, where they are dropping off, and where the product is delivering its intended value. They draw conclusions based on what the data tells them, improving the product in the direction that real user behaviour indicates.
  
  HR launches a programme and moves on. This results in programmes like the annual performance review cycle that managers find burdensome and employees find meaningless, yet it continues year after year, because the measurement infrastructure to reveal its failure to deliver value has never been built.

Applying the product manager's approach to HR programme design

Borrowing the product manager's AI playbook does not require HR teams to adopt a different relationship with data, with testing, and with the concept of iteration. Here is what that looks like in practice.

• Treat every HR programme as a product with a user: Product managers define their user before they define their product. HR teams must do the same. A performance management framework has two primary users: the manager who conducts the conversation and the employee who receives it. A new onboarding programme has one primary user: the new joiner. Designing from the user's perspective is the first and most important shift.
• Define the problem before designing the solution: Before designing a new programme, HR teams must use AI-powered analysis of existing data to define the exact problem. This involves asking questions like: What do the last two years of pulse survey data actually reveal about where employees are experiencing the most friction? What patterns emerge in exit interview data that point to specific gaps in the employee experience? What do the behavioural signals in collaboration and communication data tell HR about where engagement is breaking down? This analysis defines the problem that the programme needs to solve. Without it, HR is designing solutions to problems it has assumed rather than diagnosed.
• Pilot before you scale: Product managers test before they build. HR must test before it launches. A new performance framework should be piloted with two or three teams before it is released across the organisation. The pilot is not a soft launch. It is a structured test of core assumptions, with AI-powered analysis of the results built into the design before the pilot begins.
• Build measurement into the design: Product managers do not decide how to measure success after launching the product. They define their success metrics before building, and they build the measurement infrastructure into the product from the start. HR must do the same. Before a programme launches, HR must define specifically what adoption looks like, what behaviour change the programme is designed to produce, and how AI-powered analytics will track those outcomes in real time after launch.

Where AI fits in the HR roadmap: Specific applications

The product manager's AI playbook has specific applications in the HR context that HR teams can begin building toward immediately.

• AI-powered needs analysis: Before designing any new HR programme, HR teams should use AI to analyse the full body of existing employee feedback data. Identify the specific friction points, the recurring themes in negative sentiment, and the gaps between what employees say they need and what the current programme architecture provides. This analysis replaces the stakeholder consultation process that currently drives most HR programme design.
• Predictive adoption modelling: Before launching a programme across the entire organisation, use AI to model predicted adoption rates across different employee segments. This will help answer questions like: Which teams are most likely to engage with the programme as designed? Which are most likely to resist it, and for what structural reasons? This predictive modelling allows HR to design targeted implementation support for the segments most at risk of poor adoption, before the launch, rather than after.
• Continuous feedback synthesis: Build an AI-powered feedback synthesis capability that continuously processes employee feedback across all available channels, formal surveys, informal communication, support requests, and exit data, and flags the signals most relevant to current programme performance. This gives HR the equivalent of the continuous user feedback loop that product teams use to drive ongoing roadmap decisions.

The product manager's AI playbook works because it treats the gap between design intent and user behaviour as a data problem. HR teams that apply the same logic to people programme design will close a gap that has been costing organisations money, morale, and talent for decades

Key Takeaways

• HR teams and product managers are solving the same problem in different domains: designing experiences for people, making decisions based on incomplete information, and discovering the gaps only after launch. The difference is that product managers have built a systematic, AI-powered approach to closing this gap. HR teams have not, and the cost shows up in every programme that misses its adoption targets.
• Before building anything, product managers use AI to analyse large volumes of user feedback and behavioural data to define the exact problem they are solving. HR must do the same. AI-powered analysis of pulse survey data, exit interview content, and engagement signals can replace the stakeholder consultations and brief surveys that currently drive most HR programme design.
• Product managers test core assumptions before committing to a full build. HR must pilot before it scales. A structured pilot with two or three teams, with AI-powered analysis of the results built in from the start, is designed to catch the assumptions that would have damaged adoption after a full launch. 
• What employees say they want and what they actually do are often different things. HR relies almost entirely on surveys and focus groups. AI-powered behavioural analytics can tell HR where employees disengage from a process, which parts of a programme they skip, and which tools they actually use versus which ones they were given. This is the data that drives better design decisions.
• Most HR teams launch a programme and move on. Product managers track adoption signals continuously and draw conclusions based on what the data reveals. HR must build the measurement infrastructure before launch, defining exactly what adoption looks like, what behaviour change the programme is designed to produce, and how those outcomes will be tracked in real time.
• Three specific AI applications HR can build toward immediately: AI-powered needs analysis that processes existing employee feedback before any new programme is designed, predictive adoption modelling that identifies which employee segments are most at risk of poor adoption before launch, and continuous feedback synthesis that processes signals across all channels to drive ongoing programme improvement.