The Now-Next-Later Roadmap
Replace the Gantt timeline with three confidence buckets so plans get vaguer as they get further out.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 95%
Janna Bastow's alternative to the timeline/Gantt roadmap: instead of an x-axis of dates and a y-axis of features, you lay out the problems you intend to solve in three columns — Now, Next, Later. Granularity decreases as you move right, matching the cone of uncertainty. Dates are allowed on the specific items that genuinely have hard dates (regulatory, seasonal), but are not forced onto everything by the format itself.
Origin
Invented by Janna Bastow and her ProdPad co-founder Simon. The first version of ProdPad was a digitised timeline roadmap (drag-and-drop Gantt). Customers kept asking for a multi-select drag so they could shift everything right by a month; a five-whys dig revealed nobody was actually delivering roadmaps in the timeframes they published. Rather than build the multi-select, they killed the timeline. The uncertainty gradient borrows from the cone of uncertainty.
Core principles
- 01The format of the artefact dictates the behaviour — a timeline forces every item to have a due date and a duration whether or not it has one.
- 02The further out you plan, the more you are making it up; the roadmap should show that honestly.
- 03The unit on the roadmap is a problem to solve, not a feature to ship.
- 04Dates are an exception you opt into, not a default the template imposes.
How to run it
- 1
Delete the timeline axis
Remove the date x-axis from the roadmap entirely. The moment time is an axis, everything below it inherits a due date and a duration by implication of the format alone.
Pro tip You do not need software. Bastow says the Now-Next-Later roadmap can be done with Post-it notes on a wall.
- 2
Create three columns: Now, Next, Later
Now = what the team is actively working on (weeks away, not months). Next = the near term. Later = the future, deliberately coarse. Originally framed as current / near-term / future.
- 3
Populate with problems, not features
Each card states a problem you believe you have and want to solve. Order them. This makes the roadmap a set of assumptions you can be challenged on rather than a delivery contract.
Pro tip Force each card to answer: what problem does this solve, why solve it, what outcome do we expect, and did we measure it afterwards.
- 4
Reduce granularity as you move right
Items in Later should be broad and few. Items in Now can be specific. Resist the urge to detail the Later column — that detail is fiction and will be re-litigated.
Watch out Do not treat vagueness in Later as an excuse for having no strategy — the columns still express a considered problem sequence.
- 5
Add dates only where a date is real
Where an external forcing function exists (GDPR, a school year, the Christmas rush), attach a date to that one item and plan it properly: more buffer, earlier soft launch, iteration time before the hard deadline.
Watch out If you plan every item to that standard you either cut quality by shipping last-minute, or you move so slowly that a tiny team out-delivers your large one.
- 6
Pair it with regular communication rituals
Because the roadmap no longer broadcasts dates, replace them with cadence: keep stakeholders looking at the Now column, and run launch-readiness meetings so people know what is in testing and about to appear.
In the wild
About a month after shipping the drag-and-drop timeline roadmap, early users started asking to multi-select and shift everything right by a month. Instead of building it, Bastow and Simon ran a five-whys on the request and discovered that none of these product managers — including ones they considered better than themselves — were delivering their roadmaps on time. The request was a symptom of a broken artefact, not a missing feature.
→ They shipped the three-column Now-Next-Later roadmap instead, which produced the first significant bump in ProdPad usage.
When the concept first launched, Bastow says people called it crazy talk and insisted you could not work that way. ProdPad kept publishing the position ('no timelines, come for the now next later') on the site and the community kept practising it.
→ Years later she reports the reaction has flipped to 'of course this is the way that it works, why would you work any other way'; roughly 70% of teams coming to ProdPad now self-select for Now-Next-Later.
Common mistakes
Assuming missed roadmaps are a personal delivery failing
Bastow spent her early PM years believing she just needed better buffers and better expectation-setting. The discovery that essentially nobody hits a timeline roadmap reframes it as a defect of the artefact, not the operator.
Reading it as 'no deadlines ever'
It is fewer deadlines, not zero. Regulatory and seasonal dates are real and get dates. The point is to stop penalising the team by stamping a date on every single item.
Keeping the timeline and just renaming the columns
If time stays on an axis, the Gantt behaviour returns — everything gets an implied due date and duration regardless of the labels.
Is it for you?
Best for
Product managers and product leaders whose roadmap is a colourful Gantt chart that is never delivered as published, and who want to shift stakeholder conversations from dates to problems.
Not ideal for
Delivery-driven work where the entire portfolio is genuinely date-bound (regulatory programmes, hardware ship dates, contractual milestone delivery), or organisations where leadership will not sanction any change to how commitments are made.
From the transcript
“we found out that no one was actually delivering the road map in the time frame that they were saying they were”
“the further out you plan the more you're making it up we know this and so we all wanted more flexibility”
“it's taking from the concept of the cone of uncertainty taking from the idea that things get less certain as they get further away”
“it just means not penalizing ourselves by having a date on everything on the roadmap”
From the episode
Building better product roadmaps
Janna Bastow (Mind the Product, ProdPad)