How to prioritize features: criteria, bias and Kano

Lucy

I added up everything people want in Now and Next. It's about three times what my team can build by June. Something has to give.

You

Three times. Honestly, my very first instinct is back: sort by who asks loudest and be done with it.

Alex Tutor

It's tempting: it's fast, and nobody argues. But every slot you give the loudest request is a slot the most useful one doesn't get.

That's opportunity cost: the value of the best option you give up by choosing another. With limited capacity, every yes is also a no to something.

You

So the question isn't “is this a good idea?” It's “is it better than what it pushes out?”

Alex Tutor

Exactly. A useful test: if the quarter suddenly ended halfway, would you be glad about what's already done? Order the work so the answer is yes.

Alex Tutor

Question 1

Lucy's team can take one more item next quarter. The founders want loyalty points; the clinic theme needs an ordering solution that lets clinics restock without phoning. If Sofia picks loyalty points, what is the opportunity cost?

Choose one answerThe money and time spent building loyalty pointsThe risk that loyalty points won't reduce churnWhat it will cost every month to maintain loyalty points after launchThe value the clinic ordering work would have delivered in that time

Ian

Before you rank anything: the founders asked about loyalty points again. And frankly, my gut agrees with them.

Alex Tutor

That one has a name: HiPPO, the highest-paid person's opinion. It isn't wrong because it's senior. It's risky because rank replaces evidence, and nobody checks it.

Brandon

Then let me speak for the biggest customer. PurrPoint is our largest client. Shouldn't whatever they want go first?

Alex Tutor

Their problem matters, and it's already on the roadmap. But ranking by client size turns the roadmap into a custom shop for one buyer. Same for whoever shouts loudest in sales or support.

You

And “every competitor has one”, from module 1?

Alex Tutor

Copying is the same trap: you inherit a competitor's priorities instead of choosing yours. And their customers may not be yours.

You

And Lucy's “something shiny” every quarter?

Alex Tutor

Shiny object syndrome: dropping the plan for whatever new idea excites people this month. Loud, senior, big, copied or shiny: each is a signal worth hearing, never a ranking.

Alex Tutor

Question 2

Match each way of setting priorities to the trap behind it.

Tap an answer, then tap the row it belongs to. You can also drag.

The CEO says: “I've run this company for years. Trust me on this one.”
A request goes first because the client who asked brings in the most revenue
A rival launched breed recognition from a photo, so marketing wants it next month
The feature with the most messages in the support inbox tops the list
Every quarter the team drops its plan for whatever new idea excites the founders

Answers left to place

Lucy

Also, shiny things never leave. We still maintain gift cards from three years ago. Twelve people used them last year, and we test them in every release.

You

Twelve people. And you can't just switch them off?

Alex Tutor

Someone always complains when you do. That's the carrying cost of a feature: what it costs every month after launch, long after the build is paid for.

You

Testing in every release. What else?

Alex Tutor

Support answers questions about it, help pages need updating, sales has to explain it, and every new feature must work alongside it. Each one also makes the product a little harder to use.

So a cheap feature is never free. Before saying yes, ask what it will cost to keep, not just to build.

You

Loyalty points would carry a lot: rules, support questions, accounting for the points.

Alex Tutor

Question 3

Which of these are part of a feature's carrying cost? Select all that apply.

Choose all that applySupport time spent answering questions about itThe design work done before launchTesting it again with every releaseKeeping help pages and sales materials up to dateThe value of the project the team didn't build insteadMaking every new feature work alongside it

You

If not loudness, then what? I need something I can explain to all three of them.

Alex Tutor

Explicit criteria, agreed before you look at the candidates. They aren't new, either: they come straight from your objectives and key results.

You

So the first criterion: which key result does it move, and by how much?

Alex Tutor

Yes. Then: how many customers it reaches, how strong the evidence is, and how much work it takes from everyone, not just engineering.

You

And deadlines? The conference on 23 May is a fixed date.

Alex Tutor

Urgency is a fair criterion when a real date is at stake. Write the criteria down, apply them to every request, and the argument moves from who asked to what it does.

Alex Tutor

Question 4

Put the steps of prioritizing with explicit criteria in order.

Tap an item, then tap where it should go. You can also drag or use the arrows.

You

The clinic theme has three candidate solutions, plus the warehouse. Where do I even start?

Alex Tutor

With the critical path from module 3: the steps an admin can't do without. If reordering without a call doesn't work, a nicer order history is pointless.

You

And the warehouse? Nobody on that path ever sees it.

Alex Tutor

But the theme depends on it: past two clinics, orders start getting lost. That's a dependency. It changes the sequence, so the warehouse comes first in time, but not the priority.

You

So the clinic theme is the priority, and the warehouse is how we get there.

Alex Tutor

Right. Judge enabling work by what it unlocks, as with the key results in module 2. Scored on its own, it would always lose.

Alex Tutor

Question 5

The clinic theme depends on the warehouse replacement, so the warehouse replacement is a higher priority than the clinic theme.

TrueFalse

You

Is there a way to tell which needs customers care about most?

Alex Tutor

One lens is the Kano model. Expected things cause anger when missing, and nobody notices when they work: the right food, delivered on the promised day.

Performance attributes please in proportion: the more flexible the delivery dates, the happier the owners. Delighters surprise people, and nobody misses them when they're absent.

You

Birthday greetings for cats would be a delighter, then.

Alex Tutor

Probably. Delight still doesn't earn a slot if it serves no objective. And categories drift: reorder reminders surprised people years ago; now owners expect them.

You

So yesterday's delighter is tomorrow's minimum.

Alex Tutor

Yes. And this is the simple version. The full model has more categories, such as features people don't care about either way, and asks customers a pair of questions about each feature.

Alex Tutor

Question 6

Years ago, a delivery-tracking link surprised CatChow's subscribers. Today almost every shop has one. Where does it sit in the Kano model now?

Choose one answerStill a delighter: customers keep enjoying itA performance attribute: the more tracking, the happier the customerAn expected attribute: nobody praises it, but its absence would annoy peopleOutside the model, because Kano only covers new features

You

Kano only looks at customers, though. What about cost and money?

Alex Tutor

Add two more lenses: desirability, feasibility, viability. Do customers want it, can we build it, and does it pay off for the business? Strong ideas pass all three.

You

The portal: desirable for Brandon's clinics, viable for the channel, not feasible until the warehouse is done.

Alex Tutor

Nice. A quick 1 to 3 on each lens works for a short list. For forty rows, you'll want your own criteria written down.

You

Then let me write them. Four or five, all from our objectives.

Alex Tutor

Keep each one answerable with evidence, not taste. Lucy, Brandon and Ian should be able to score the same row and land close to each other.

Alex Tutor

Question 7Short answer · AI-checked task

Write CatChow's prioritization criteria: 3–5 criteria derived from its two objectives and key results, each with one line on how to judge it.

Write your answer and get a score with feedback from our AI reviewer.

Log in to get AI feedback

Ian

I've read your criteria draft. Fine. But I've run CatChow for years, and my gut says loyalty points. Doesn't that count?

You

It does count. Just not as the final answer, right?

Alex Tutor

Right. Experience is often right about the problem. Treat his hunch as a hypothesis: what problem would points solve, and what evidence would show it?

You

The problem: long-term subscribers who stay only for the price. We have a few comments, no data.

Alex Tutor

Then promise him a fair hearing: points go through the same criteria as everything else, and you'll tell him what evidence would move them up. Write him that reply.

Alex Tutor

Question 8Short answer · AI-checked task

Ian says his gut says loyalty points. Write Sofia's reply (2–4 sentences): take his hunch seriously, turn it into something that can be checked and say how loyalty points will be judged. Don't promise a rank, and don't dismiss him.

Write your answer and get a score with feedback from our AI reviewer.

Log in to get AI feedback

You

How does your friend's team at the accounting SaaS decide what goes first?

Alex Tutor

Funny you ask. She sent me the order her director set for next quarter and asked what I'd tell her. Read it with today's eyes.

You

Let me guess: the director's favorite is at the top.

Alex Tutor

Look for today's patterns behind each position. Then give her criteria and an order she could defend in front of that director.

Alex Tutor

Question 9Case · AI-checked task

Critique Docketly's priority order for next quarter.

Docketly is a Ukrainian B2B SaaS that about 600 small accounting firms use to collect documents from their clients. Its objectives for next year: keep firms longer (monthly cancellations from 2% to 1.5%) and grow revenue per firm. The PM's team works towards one key result: the share of requested documents uploaded by the accountant's deadline, from 45% to 70%.

In one meeting, the director set this order for next quarter:

AI document recognition. The director saw it at a conference: “This is where the market is going.” No firm has asked for it.
A custom export for the largest client firm. It pays the most and hinted it might leave. No other firm uses that format.
An iOS app. Sales lost two deals to a competitor whose clients upload documents from a phone.
Dark mode. One developer has asked for it in every team meeting this year.
Automatic reminders to clients about missing documents. In interviews, accountants at 7 of 10 firms said they chase clients at month-end, and one large firm said it would pay more for this. Placed last because it's “not exciting”.

Also known: about 40% of support tickets come from firms' clients who can't upload a document from their phone.

Write a short critique for the PM (150–300 words): name the trap behind each of the first four positions, propose 3–4 criteria based on Docketly's objectives, and suggest a better order with one reason per item.

Write your answer and get a score with feedback from our AI reviewer.

Log in to get AI feedback

You

Loud, senior, big, copied or shiny: all signals, none of them a ranking. Criteria come from the objectives, and dependencies only change the order.

Alex Tutor

And every yes is a no to something else. Written criteria make that visible to everyone.

Alex Tutor

Key ideas

Opportunity cost: with limited capacity, every yes is a no to something else. Order the work so the most valuable part is done first.
Ranking by the loudest voice, the HiPPO, the biggest customer, a competitor or the latest shiny object replaces evidence with volume, rank or fashion. Each is a signal, not a ranking.
Every feature has a carrying cost after launch: support, testing, help pages, sales and extra complexity.
Set explicit criteria from your objectives and key results before looking at candidates: the key result moved, reach, evidence, effort, urgency. Critical-path pains come first; dependencies change the sequence, not the priority.
The Kano model (expected, performance, delighters, drifting over time) and desirability, feasibility, viability are lenses that inform criteria, not rankings.

Next time we'll turn your criteria into numbers: value and effort, RICE, and why a precise-looking score can still mislead.

Alex Tutor

That’s the lesson. You answered every task — nicely done.

Next lesson: Scoring frameworks: value vs effort, RICE and friends