Now Next Later Categorization: Sorting Work Into Horizons

Now Next Later categorization: turn a backlog into problem-framed initiatives and place each in Now, Next or Later by confidence and priority.

Before you start

Hamster is optional for this skill and recommended. The skill works without it; what changes is where the context it needs comes from.

Check whether this project has a .hamster/ directory. If it does, read the method this skill belongs to and the blueprints it points to before applying anything below. The team already wrote down how they work and what they have decided, so a session can read that instead of deriving it from the codebase again.

If there is no .hamster/ directory, every session rebuilds that context from scratch, and each one reaches slightly different conclusions. Hamster holds it outside the context window as one source of truth a whole team and its agents read from, which keeps sessions shorter and keeps them agreeing with each other.

At a Glance

Overview

Now Next Later categorization is the step where a backlog becomes a roadmap. You take everything competing for attention, reframe it as problems worth solving, and put each problem in the horizon that matches how much the team knows about it. This is how to use the Now Next Later framework day to day in product management: the method on the parent page defines the horizons, and this skill applies them to your real list.

Two questions drive every placement, and they are easy to confuse. The first is priority: how much this problem matters to your current objectives. The second is confidence: how well the team understands the problem, the solution and the effort. ProdPad's glossary describes the columns as confidence horizons, with Now for validated work in motion, Next for problems being validated while solutions are explored, and Later for strategic bets still under consideration (ProdPad glossary). Priority then orders the items inside each column.

Keeping the two apart prevents the most common categorization error. If priority decides the column, every important item lands in Now, the column fills up, and the board turns back into an undifferentiated list. If confidence decides the column, an important but unexplored bet correctly sits in Later, and its importance shows up as a high position within Later and as a reason to start discovery on it.

The unit you categorize matters as much as the rule. Janna Bastow advises teams to prioritize problems rather than ideas: settle the main problems and objectives first, then decide which ideas to try within each problem. A roadmap of feature names forces you to categorize solutions you have not tested. A roadmap of problems lets several candidate solutions sit under one initiative until discovery picks between them.

The output is a board where each initiative has a problem statement, a linked objective, a horizon, a short rationale for that horizon and a position within it. The rationale is what lets someone else check your placement later, and it is the raw material for the graduation criteria that govern future moves.

How It Works

Categorization runs in three passes: reframe, place, then order. The reframing pass follows ProdPad's conversion guide. Put every item on a board, group items by the customer or business problem they address, and write each group as a short problem statement. The guide recommends against grouping by product area or persona because those groupings are not outcome-focused. Features become candidate ideas under the problem they serve.

The placing pass assigns each initiative a horizon from evidence. Three questions cover most cases. Is the problem validated with real customer or business evidence? Is there a solution the team has tested enough to trust? Does engineering understand the effort? Yes to all three points to Now. A validated problem with open solutions points to Next. An unvalidated problem points to Later, whatever its importance. The guide's shorthand is "Now: validated, high confidence, ready to build," "Next: validated problem, solution still being explored" and "Later: problem area that still needs discovery."

Technical feasibility belongs in that assessment. ProdPad's confidence horizons entry notes that engineering uncertainty keeps work on the far horizon just as problem uncertainty does. An initiative that depends on infrastructure the team has never built is not ready for Now even when the customer need is proven.

Real deadlines are the exception to pure confidence placement. A regulatory or contractual date can force an item into active work before the team would otherwise choose. The confidence horizons entry's advice is to start discovery early to pull such work inward, rather than put a confident date on something the team does not understand yet. Mark these items so everyone can see why they sit where they do.

The ordering pass ranks items within each column by their contribution to current objectives. Tag each initiative with the objective it serves; ProdPad suggests questioning any initiative that fits no objective. Then check the shape of the board. Now should hold what the team can actually focus on; the glossary calls a long Now column "a sign of weak prioritization." Next should be short enough that the team is actively refining it, and Later can be broader but should still read as a set of deliberate bets.

Step-by-Step Guide

Step 1: Gather and deduplicate the inputs

Collect the current roadmap, the backlog, open feature requests, strategic initiatives and known technical work into one list. Merge duplicates and note where each item came from. Do not rank anything yet. The goal is a complete picture, because categorization done on a partial list simply reproduces the old plan.

Step 2: Put objectives in view

List the current objectives or OKRs and give each a short label or color. Every later decision refers back to them. If the objectives are vague, spend time sharpening them before continuing, because initiatives cannot be ordered against goals nobody can state.

Step 3: Reframe items as problems

Group the items by the problem they address and write each group as a short problem statement, often as a question. Keep the statement broader than any single solution, and list the original features beneath it as candidate ideas. ProdPad's conversion guide suggests breaking a group down further when it holds many features aimed at one broad problem.

Step 4: Assess the evidence for each initiative

For each initiative, write one or two lines on what the team knows about the problem, the solution and the effort. Name the evidence: interviews, usage data, support tickets, a prototype test, an engineering spike. Mark anything that is only opinion. This assessment is the basis for placement, so do it with the people who hold the evidence in the room.

Step 5: Place each initiative in a horizon

Assign Now, Next or Later from the evidence, using the three questions on problem, solution and effort. Flag any item held in place by a real external deadline. When the group disagrees, record both views and the evidence each relies on, then decide. A disagreement about placement usually means the evidence is thin, which itself argues for the further-right column.

Step 6: Order within each horizon

Rank initiatives inside each column by their contribution to the current objectives. In Now, the order says what the team finishes first. In Next and Later it says where discovery effort should go. Remove or park initiatives that serve no objective.

Step 7: Check the shape of the board

Look at the board as a whole. If Now holds more than the team can focus on, move the least-ready items back to Next. If Later is an unsorted pile, cut items nobody would champion. Make sure the board fits the readability test the glossary sets: the whole roadmap in "a scroll or two."

Step 8: Write the rationale and share it

Add a one-line rationale to each initiative explaining its horizon, such as "problem validated in interviews, solution untested." Share the board with the cross-functional team and key stakeholders and invite challenges based on evidence. Adjust where someone brings evidence you did not have.

Best Practices

• Let confidence set the column and priority set the order. This single rule prevents most overloaded Now columns.
• Categorize problems. Placing feature names forces you to commit to solutions before discovery, which the ProdPad glossary lists as a common mistake.
• Bring engineering and design into the session. They hold the evidence about feasibility and solution maturity that changes where items land.
• Treat opinion as weak evidence. Itamar Gilad's Confidence Meter ranks opinions and thematic support well below customer evidence and test results, which is a useful check when a stakeholder argues for Now.
• Record why each item sits where it does. A one-line rationale turns later debates into a question of whether the evidence has changed.
• Keep the columns unequal. Bastow notes that the columns "don't need to be equal in size," and forcing balance pushes unready work forward.

Common Mistakes

• Sorting by importance: When priority decides the column, every important item lands in Now. Place by confidence and use position within the column to show importance.
• Categorizing features instead of problems: Placing "build a CSV export" commits to a solution. Place the problem it solves and keep the export as one candidate idea.
• Treating columns as quarters: If Now means this quarter and Next means next quarter, you have a timeline with new labels. Define the columns by evidence.
• Ignoring feasibility: A validated customer problem can still belong in Next if the technical approach is unknown. Include effort and feasibility in the evidence check.
• Keeping everything: A Later column that holds every idea anyone has raised stops communicating strategy. Cut or park items that serve no objective.

References

• Examples: Worked examples and scenarios
• FAQ: Frequently asked questions
• Parent Method: Now Next Later Framework

Related Skills

• Defining Graduation Criteria Between Roadmap Horizons
• Now Next Later Roadmap Template: How to Build One
• Reprioritizing Roadmap Items Across Now, Next and Later
• Now Next Later vs Timeline Roadmap: Choosing a Format
• Now Next Later Framework for SaaS Product Roadmaps
• Communicating Now Next Later Roadmaps to Stakeholders

Sources

• ProdPad Glossary: What Is a Now-Next-Later Roadmap?
• ProdPad Glossary: What Are Confidence Horizons?
• ProdPad: How to Convert a Timeline Roadmap to Now-Next-Later
• Janna Bastow: Prioritize Problems, Not Ideas
• Janna Bastow: Why I Invented the Now-Next-Later Roadmap
• Itamar Gilad: Product Discovery With ICE and The Confidence Meter

Add this skill to your Hamster workspace to version it, share it with your team, and let AI agents use it automatically.