Concept 2 of 4

Prioritising with market data

5 questions test this

Build is the course where the market knowledge becomes an order of work. Its stated purpose for this activity is direct, to maximise impact and minimise infighting by using market data to rank requirements objectively.

The infighting half is not incidental. Most disagreements about what to build next are not disagreements about facts, they are two people with different private evidence and no shared way to compare it.

What counts as market data

Pervasiveness. How many people in the defined segment have this problem, expressed as a proportion of a named denominator.

Urgency. What it costs them per occurrence, and what they currently do instead.

Evidence of willingness to pay, or in an internal product, evidence that somebody would spend budget or effort.

Revenue at risk or revenue reachable, stated as a commercial input rather than smuggled in as a market fact.

Effort, estimated by the people who will do the work.

Notice which signals are absent. Volume of requests, seniority of requester and recency all describe the organisation rather than the market.

Comparing on the same basis

The mechanism matters less than the consistency. Cost of delay divided by duration, a weighted score, or a simple ranking against two criteria all work, provided every candidate goes through the same questions.

What the model produces is not a decision but a legible argument. When somebody disagrees they now have to say which input they think is wrong, which is a question with an answer, rather than asserting that their item is important, which is not.

Ranking against a fixed capacity

The ordering only means something when it is set against how much the team can actually do. A ranked list of forty items with no line drawn is still a wish list. Drawing the line, and saying out loud what falls below it this cycle, is what turns prioritisation into a decision.

Metrics for stakeholder alignment

Build pairs prioritisation with the use of metrics to raise predictability. The same principle applies. Delivery data that is published on a rhythm, before anyone asks, does more for stakeholder trust than an accurate forecast produced under pressure.

The pairing is deliberate. Stakeholders accept an order they disagree with far more easily when they can see progress against it and can predict when their item comes into view.

Keeping the evidence attached

A ranking is only re-usable if the reasoning survives. Recording the market evidence against each requirement, in the same table, means the order can be recalculated when something changes rather than relitigated from the start every quarter.

Common misconceptions

The number of support tickets is a good ranking signal.

Ticket volume measures who complains, not who has the problem. It is dominated by the customers with the closest relationship and the most time, and it is silent on the people who left rather than raised a ticket.

Ranking objectively means removing judgement.

Judgement stays. What market data removes is the requirement to settle disagreements by seniority, because both parties can look at the same evidence and argue about the evidence instead.

The largest customer's request should rank first.

Revenue concentration is a real input and it belongs in the argument explicitly rather than implicitly. Recording it as a commercial reason keeps it honest and stops it being disguised as market evidence.

5 questions test this concept

A product manager ranks requirements by the number of support tickets each has generated. What is the weakness of this signal?

  • ATicket counts are usually recorded inaccurately.
  • BTickets should be weighted by how long each took to resolve.
  • CThere is no weakness, since tickets come from real customers with real problems.
  • DTicket volume measures who complains rather than who has the problem. It is dominated by the customers with the closest relationship and the most time, and it is silent on the people who left rather than raising a ticket.
Check whether it stuck.

One per page, with a worked explanation.

Start the set
Related material
Book
Escaping the Build Trap, On why a queue of requests is not a strategy.
Book
Actionable Agile Metrics for Predictability, On flow data that supports a delivery conversation.
Book
Shape Up, On fixed time and variable scope as a prioritisation device.
Template
RICE scoring sheet, Reach, Impact, Confidence and Effort columns with the score formula built in, defined confidence bands, and a column recording the assumption behind each input.