PM Roadmapping (Now/Next/Later, Outcome-Driven): A Working Guide (2026)

Now/Next/Later: a real worked artefact

Worked example for a growth PM team owning new-user activation on a 12M-MAU consumer subscription product. Each row is an outcome with the underlying hypothesis and 1–2 candidate solutions; the candidate solutions are visible but not committed.

Now (in flight, this 4-week cycle)

• Outcome: Lift step-4 (Code Entered) conversion from 41% → 50% across all sources.
  Hypothesis: The 6-digit-code entry pattern fails on mobile-app-first traffic.
  Solution in flight: Magic-link-instead-of-code A/B test (50/50 split, 4-week run).
  Eng partner: Identity team. Status: Day 9 of 28.
• Outcome: Reduce onboarding-completion drop between session-1 and session-2 from 38% to 30%.
  Hypothesis: Users abandon when day-2 isn't valuable enough to return.
  Solution in flight: Personalized day-2 content recommendations.
  Eng partner: Recommendations team. Status: Day 5 of 28.

Next (most likely 6–8 weeks ahead)

• Outcome: Lift week-2 retention by 6pp on the social-graph-active cohort.
  Hypotheses (un-prioritized): (a) Friend-discovery surface; (b) shared-content nudges; (c) social-proof in onboarding.
  Decision pending: Customer-discovery interviews due by week 4 to pick hypothesis.
• Outcome: Reduce trial-to-paid friction at upgrade prompt.
  Hypothesis: Pricing-page copy and placement.
  Decision pending: Pricing experiment design with finance partner.

Later (directional intent, 3–6 months out)

• Activation experiments on geo-expansion markets (Q3).
• Habit-loop intervention on day-7-active cohort (Q4).
• Re-engagement playbook for 28+ day dormant users (Q4).

The artefact lives in Notion or Linear (Linear's roadmap surface supports outcome-grouped initiatives). It's published to the broader org, reviewed weekly with eng/design partners, and updated as outcomes are hit or hypotheses are killed. The Now section is the only commitment; Next and Later are working assumptions.

Outcome vs feature framing

Two roadmap rows, same intent:

The outcome framing is durable because there are usually 3+ candidate solutions for the same outcome. If the first hypothesis fails (the magic-link experiment doesn't move step-4), the team still owns the outcome and tries a different hypothesis. The feature framing collapses on first contact: 'we shipped onboarding-redesign v2; conversion didn't move; what now?'

Marty Cagan's 'outcomes over outputs' framing is the canonical reference for this shift; SVPG's writing on outcome-driven roadmaps documents the pattern at scale across coached companies.¹

Real published roadmap examples

Two public examples worth studying:

• Linear's company roadmap. linear.app/method documents the outcome-driven cycle-based shipping cadence; specific public roadmap surfaces (e.g., Linear changelog) reveal the actual shipping rhythm.
• GitLab's public product handbook. handbook.gitlab.com/product publishes the company's product strategy, OKRs, and roadmap publicly. The level of detail is rare; reading it demystifies how a 1,000+ person product org actually shapes a roadmap.²

What both have in common: outcomes named at the top of each roadmap section; hypotheses listed but not committed; date commitments are coarse (this quarter, next quarter, later) rather than fine (March 14).

Stakeholder pressure patterns

The three patterns that break prioritization (drive-by adds, customer escalations, HiPPO overrides) also break roadmaps. Roadmap-specific defenses:

1. Drive-by add: 'Can we add X to the Now section?' Move it to Later by default. Promotion to Next requires a customer-discovery cycle; promotion to Now requires displacing a current Now item. Make displacement explicit.
2. Customer escalation: 'Big customer wants Y by Q2.' Pipeline the request through the customer-input → roadmap process. If Y is genuinely strategic, it goes through Next; the pipeline-shape forces the trade-off conversation.
3. HiPPO override: 'I want Z on the roadmap.' Get the override in writing; force it to displace an existing commitment, named explicitly. The HiPPO can override; what they can't do is override silently while the PM owns the missed commitment.

What hiring managers screen for

Senior+ PM interviews on roadmapping converge on three signals:

• Outcome-vs-feature judgment. 'Walk me through your last roadmap' — strong PMs lead with outcomes; junior PMs lead with feature lists. The choice is the screen.
• Naming the trade-off. 'What did you cut to make room for X?' If the answer is 'we found capacity,' the screen tightens. Real roadmaps require visible cuts.
• Stakeholder-follow-up scenarios. 'Tell me about a time you held the roadmap against a senior leader's drive-by add.' Specifics matter. The PM who can name the conversation, the trade-off documented, and the outcome of the override decision converts better than the PM with a clean roadmap and no friction.