What to Build Next: Customer Demand for Your Gaps, Next to What Competitors Already Ship

Product teams are not short of requests. ProductPlan's 2023 State of Product Management report, built on more than 1,500 product managers' answers, found that reviewing customer feature requests is the number one source of actionable product ideas.[1] It also found that 54% of roadmaps are designed around outputs, with only 44% communicating outcomes, and that about 52% of product managers say their strategy is mainly shaped by requests from senior executives or direct customer feedback.

The hard part is choosing. In Craft.io's 2023 survey of 500 product professionals, capacity planning and prioritization tied as the top challenges, each named by 41%.[2] Among those whose biggest challenge was prioritization, 35% pointed to disagreement over the method, 35% to collecting the data to base the decision on, and 30% to the loudest or highest-paid person in the room winning.

Part of the problem is where the evidence sits. What customers ask for is spread across the CRM, call notes, and survey tools. What competitors ship sits in a battlecard deck or a feature spreadsheet. Neither half answers the question on its own. This post sets out a workflow that puts the two side by side, using parts of PYRAMYD that work today.

Why you need both halves

Economists who value product features make the point formally. Greg Allenby, Jeff Brazell, John Howell, and Peter Rossi define a feature's market value as the change in a firm's profits with and without it, once prices settle across the market. Computing that takes more than demand: the demand has to be supplemented by information on competing offerings and costs, and their profit-based measure gave very different answers from the standard willingness-to-pay calculation.[3]

The people who choose features say much the same. When Pilar Rodríguez, Cathy Urquhart, and Emilia Mendes interviewed 21 decision-makers at three software and software-intensive companies about how they select features, they found 47 value propositions in six dimensions. Customer value came up most: all 21 raised it, in 387 coded instances. Market competitiveness came second with 273 instances, and economic value came up in 120.[4]

A count of requests, read without the market, can mislead you about what a feature is worth. A parity list, read without demand, builds what rivals have and your customers never asked for. The four steps below keep both halves on the same screen.

Step 1: List where you trail

Start in Products › Gaps. Its Competitors view lists, feature by feature, the rivals that score ahead of you, your average score and theirs, and the widest gap. It computes this from your battlecards, so it needs at least one. In Competitors › Battlecards, create a card for your product against each rival you meet most often; APEX drafts it from reviews, news, the rival's website, and product features.

Know the limit before you rely on it: this view does not yet show the deals or revenue lost on each feature, or how often each feature is mentioned in calls, deal notes, and tickets. It tells you where you trail on scores, not what trailing costs.

To see the whole category rather than one rival at a time, open the Feature Taxonomy. It lays out each category's standard features in one vendor-neutral tree, and each feature's profile lists the vendors that implement it, its maturity, and stored scores across eight dimensions. A feature that most vendors in your category implement is table stakes; one that only a few implement is a choice.

Step 2: Count the customers behind each gap

Now switch Gaps to its Matrix. Gaps are grouped by module and capability, and the Customers column counts the distinct customers behind each one. Sort by it and the gaps the most customers raise come first. Choose a gap's Evidence count to see every linked item by kind, such as lost deals, interviews, surveys, and conversations from calls and emails, and choose any count to see how it was counted.

This is where a single loud request stops setting the agenda. Say nine different customers raised the same gap across calls and lost deals: it now sits visibly above a request from one large account, however forcefully that account made it. Know this limit too: the column counts only evidence that is linked to a feature and names a customer, and public reviews are not linked to features, so they never appear in it.

The column is only as full as the evidence linked to it: a deal or an interview counts toward a gap once it is linked to the feature the buyer named. Two habits keep it full. Sync your CRM's deals from Workspace › Library › Connectors; syncs run on demand, when you choose Sync now. And ask buyers directly in Accounts › Engagement: build a survey by hand, from a template, or as an AI draft; publish it by link; and read the results with NPS, segment cuts, and AI themes. For won and lost deals, the AI interviewer runs voice or text interviews through a public link, tied to the deal, and returns transcripts, summaries, and themes.

Step 3: Check what buyers say and what rivals ship

With a short list of gaps that many customers raise, test each one against the market. Products › Reviews shows a product's reviews over time, its score distribution and source mix, and its positive and negative themes, from reviews collected up to mid-June 2026. Its trends come mostly from reviews and are read one screen at a time; a single view across your tickets, calls, and reviews is not built yet. For a specific capability, ask APEX: it searches reviews by meaning rather than by keyword and answers with the quoted reviews and their sources.

For the rival's side, Win/Loss › Competitive reads any competitor's public reviews for its top features by review score, why buyers leave it, and what they consider alongside it. To put the products side by side, add the gap's standard features as rows in a Competitors › Comparisons matrix, or open the Capabilities tab of your battlecard, which shows who wins each shared feature on the stored scores and warns you where a product has no feature data.

Reading the two halves together gives four cases:

• Many customers ask, and rivals already ship it. The gap costs you in every comparison. Close it, and check how mature the rivals' versions are before you size the work.
• Many customers ask, and few rivals ship it. This is a chance to lead. Read how reviewers describe the need before you design the answer.
• Few customers ask, but rivals ship it. Watch it. Look for it in lost deals and in the rival's review strengths before you commit.
• Few customers ask, and few rivals ship it. Park it, and look again next planning cycle.

Step 4: Decide, and send it to the tool your engineers use

Once you decide, create the item in Products › Roadmaps, or ask APEX to add the roadmap item for you, and send it with Sync to External Tools to Jira, Linear, Asana, or Productboard on your own connection. If your team already plans in one of those tools, the roadmap can show that tool's items instead. The sync is one way: an item is sent, or the tool's items are shown, and edits made in both places are not merged.

Carry the reason with it. A PYRAMYD roadmap item does not hold the evidence behind it today, so write the case into the item's description: how many customers raised the gap, the kinds of evidence, the rivals that score ahead, and what reviewers say. Then follow progress on the roadmap's Analytics tab, which charts status, weekly completion, burndown, and velocity.

When it ships, go back to the customers who asked. A short survey to that group tells you whether the gap is closed. It is also the evidence leaders say they want: in a 2018 Harvard Business Review Analytic Services survey, 90% of respondents said more testing with, and feedback from, customers during development would raise their confidence in the return on their product investments.[5]

The same facts in the room

Deciding what to build next will always take judgment. The workflow's job is to make that judgment about the same facts: how many customers raise the gap, what rivals already ship, what buyers say about it, and where it sits on the plan. Run it once each planning cycle, start with the five gaps your customers raise most, and send each decision on with its reasons attached.