Now, next, later, and nothing else
Maintaining the backlog and the roadmap are common responsibilities for a Product manager. When I arrived at Orus in august 2022 as the first product manager, my main order of priority was to organize the already long list of “things to do”. Six months later, after many iterations, we adopted a system that suits us and the stakeholders too: the famous “Now, Next, later” roadmap. With a few additional rules, we made it our own.
The “Three columns roadmap”
The idea behind the “Now-next-later” (NNL) roadmap is simple: work is organized into 3 time-horizons “now”, “next” and “later. It is usually materialized in a board with three columns:
“Now” is for the “in progress” work, clearly defined and broken down into granular tasks. Scope is stable (with common sense adjustments possible).
“Next” is for the work that will happen once the “now” work is done. The work is still being refined and may vary based on discovery or a critical event to maximise ROI.
“Later” is for everything else. We maintain a list of 3 to 5 candidates to “Next” that evolves with a lot of flexibility, without creating specs to avoid creating stock that deprecates quickly.
Timeline roadmaps do not work
Before adopting the NNL roadmap, we had classic frustrations between the product/tech team and the rest of the company: business team wanted deadlines but tech team did not want to waste time planing something that we would do anyway, multiple teams requested updates on topics that were not even started and stakeholders had a hard time prioritizing their own requests beyond one top priority, etc.
Anyone that worked in programming or any kind of tech project management capacity knows that deadlines are at best unreliable, at worst a driver of bad decisions that compound to create a mess of legacy and debt. But it is critical to be able to articulate what is being worked on and what will come next so that the work can be prepared and the rest of the company can adapt accordingly.
Providentially, one of our business angels suggested we give NNL roadmaps a try. Adopting the NNL roadmap was the solution because it appeared to fix exactly the kind of problems we re facing: timeline roadmaps do not work.
The Now-Next-Later roadmap exists because timeline roadmaps aren’t effective, simple as that. Timeline roadmaps insist on deadlines, which means product teams are deprived of the flexibility that allows them to do their best work. Janna Bastow (ProdPad CEO)
Benefits: Clarity, focus and (embraced) uncertainty
NNL roadmaps have multiple benefits and many people have written extensively around it (here). For us, what did it was three main benefits:
Clarifying expectations: Anyone can understand the 3 columns. it is simple and that’s why it works. Also, by clearly defining how we would be doing in the future, we were able to dissipate many areas of ambiguity that triggered many frustrations between teams. This alone was worth doing the change. It is also easier to share status updates to the rest of the company, less parameters means I can do it more frequently.
Focusing on what matters most: Limited spots means better choices. A functional organization can only dedicate a limited bandwidth to the “Now” and the “Next”. Everything else is in the indiscriminate “Later”. The product and tech team can therefor focus on their work and be more efficient because of it.
Embracing uncertainty: Given that “Later” is a moving target and what is in it changes continuously, it gives enough flexibility to adapt to changing conditions. By having “uncertainty” as the default hypothesis for anything beyond “Next”, it reconnects the rest of the company with the reality of development and product management.
Never half-ass two things, whole-ass one thing
We made NNL roadmaps our own by adding a few rules, with some being commitments to the rest of the organization to adjust the trade-off:
“Now = ETA”: As soon as a topic goes “Now” it must have an Estimated Time of Arrival. This helps the rest of the company organize themselves. We may marginally adjust the scope to match it. And when we pass it anyway, it is an opportunity for learning not finger pointing.
“One Piece Flow”: Each tech/product squad only works on one project at a time. That means only one topic in “Now” and one topic in “Next” and that is it. Less parallel work means less context switching. Overall that means better efficiency and quality. This rule is definitely inspired by the Toyota Production System, that I discovered at my previous job at Qonto.
“Quick wins are the exception”: If an improvement provides significant value (”Win”) and can be done in a matter of a few hours (”Quick”), we may pass them as “Quick Wins” between two tasks related to the “Now” project. It does not happen often because it is rare that a task is small and valuable enough to be considered a quick win, but this gives us the flexibility to otherwise only work on one topic at a time.
“No sprints”: The time spend trying to make multiple things in an arbitrary time-box is a waste. However, the bigger the project the higher the uncertainty. To counter that we defined a maximum length a project can have (right now it is two weeks because that is what makes sense for our size). If it is less than that we are good and we go to the “Next” thing after. If it is more than that we review the scope and split it, potentially pushing the “Next” thing back into “Later”.
What’s next?
We have implemented that system for a few month now and the results have been good. The rest of the company has a clear view of what is being worked on and when it will be done. We see a lot less messages requesting updates or expressed frustration on shifting ETAs.
For now we only have one squad, so it is easy to work like that. With time we will have to scale that approach so we may have to adapt our views. But industry experience tells us that this system scales well.
Now that we have a clear view of what we work on and a clear system to share it with the rest of the company, our next challenge will be clearly evaluating the success of what we do. If you have strong opinions on that topic, don’t hesitate to reach out on twitter (@florianhcq).
