Skip to main content

Saying No to Early Feature Requests

Focus · 10 min read ·

Early users ask for everything. How a small team decides what to build, what to park and how to say no kindly without losing the people who asked.

Illustration: A two-colour riso collage of a paper funnel with many torn slips going in and three coming out the bottom, labelled Requests and Do now, in teal and mustard

The first people who use your product are generous. They give you time, feedback and patience. They are also full of ideas. "Could it do this?" "It would be perfect if it also did that." "My friend needs something like this, but with a calendar."

It feels wonderful to be asked. It is also dangerous. A small product that says yes to everyone becomes a large, confusing product that nobody loves. Learning to say no kindly is one of the most useful skills a solo maker can have.

Why no is so hard

You want to please people. You are grateful that anyone is using the product. A request feels like a compliment, and refusing it feels rude. There is also a nagging fear: "What if this was the thing that would have made it work?"

Keep a few facts in mind.

  • Every feature has a cost. It takes time to build, then time to test, document, explain, fix and keep working as other things change.
  • Time is your scarcest resource. Each hour on a feature is an hour not spent on something else. Economists call this the opportunity cost: the value of the best alternative you gave up.
  • More is not better. Extra features often make a product harder to understand and use.
  • A request is not a requirement. It is data about what someone thinks they want.

The danger of scope creep

Project management writing has a name for requirements that grow beyond the original plan: scope creep. It usually arrives in small, reasonable steps, each of which sounds harmless. "Just one more option." "A quick export." "A tiny setting." Months later, you have built a different, bigger, slower product, and the original job is no longer done well.

For a small team, the defence is a clear statement of what the product is for, and a habit of testing every request against it.

Write down what the product is for

On a single card, write:

  • Who it is for. The person you are serving.
  • The job it does. The one thing it should do better than anything else.
  • What it will not do. At least three things.

Pin it where you see it. When a request arrives, hold it against the card.

The idea of a minimum viable product says that you build just enough to be usable by early customers, so that they can give feedback. It also implies a discipline: do not add what the first users do not need to get value.

Sort requests honestly

When a request comes, do not decide on the spot. Write it down and sort it later.

Ask:

  1. Who asked? Is this someone in your target audience, or someone outside it?
  2. How many have asked? One voice is a story. Three or four independent voices is a pattern.
  3. What problem are they really having? The feature they ask for may not be the best fix. Ask "what are you trying to do?"
  4. Does it serve the core job?
  5. How big is it? A small change that many people want is different from a large one that one person wants.
  6. What would I not do if I did this?

Frameworks exist to help sort work. One well-known method, MoSCoW, divides items into things you must have, should have, could have and will not have this time. Whatever you use, the fourth bucket, "not this time", is the one small teams forget.

The three answers

Most requests fit one of three replies.

Yes, soon. The request fits, many people want it and it is a manageable size. Say when, honestly, and tell them when it ships.

Not now. It is a reasonable idea but not a priority. Put it on a visible "later" list. You are not refusing it forever, you are deferring it.

No. It does not fit the product. Say so clearly and kindly, and where possible point to something else that might help.

Avoid "maybe" as a default. It sounds polite and creates a debt you will have to pay.

How to say no kindly

A good no has four parts.

  1. Thank them. They took the time to help.
  2. Show you understood. Say back what they need.
  3. Explain your focus. Without a lecture: "We are keeping the product small so that it does one thing well."
  4. Offer an alternative if you can: another tool, a workaround or a promise to keep listening.

Example: "Thank you, that is a really clear request. I can see you want to see all your invoices by client. For now I am keeping the product focused on weekly reports, so I will not build that. Many people export to a spreadsheet for that job and I have written a short guide on how. If other people ask for the same, I will take another look."

It is warm, honest and clear. Most people accept a reasoned no more easily than a vague promise.

Keep a public list

A simple page that shows what you are working on, what is planned, what is under consideration and what you have decided not to do builds trust. People see that you listen, and they can see where their idea sits.

Be honest about status. Do not promise dates you cannot keep. "Considering" and "not planned" are acceptable words.

Update it when you ship. It costs a few minutes and turns "no" into "I can see it, and I understand".

Mind who is asking

Not every request deserves equal weight.

  • Target users. Pay most attention to those, especially the ones who use the product most.
  • Heavy users know what is missing in the core, but their needs may drift towards complexity.
  • Light users tell you where it is confusing.
  • Non-users rarely tell you what would make them start. Be sceptical.
  • Paying customers deserve more weight than free ones, but not unlimited.

Beware the loudest voice. One enthusiastic person can create the impression of demand.

The paid exception

Sometimes a customer offers to pay for a custom feature. It can be tempting, especially when money is tight.

Consider carefully.

  • Would you build this for others too?
  • Does it fit the direction of the product?
  • Is the payment enough to cover the work, including maintenance?
  • Will it distract you from the core?
  • Who owns what is built, and what are the terms?

If the answer to most is yes, it can be a good deal. If not, say no, kindly. Custom work for one customer can swallow a small team.

Listen for the problem behind the request

Often the request is a guess at a solution. Dig for the problem.

"Could it send reminders?" might mean "I forget to check it." The fix may be a clearer first screen, not a reminder system.

Ask: "What are you trying to get done when you wish it could do that?" Write the answer down. You may find a simpler way to meet the need, or discover that several different requests are the same problem in disguise.

Reviewing your list once a month

Set a monthly slot to look at the requests list.

  • Merge duplicates.
  • Mark ones that have repeated.
  • Remove ones that no longer matter.
  • Choose one or two to work on.
  • Write the "not now" ones into the public list with a short reason.

Do not let the list grow into a guilt pile. A list that is pruned stays useful.

When saying no feels wrong

Occasionally you will say no and later wish you had said yes. That is part of the job. If a request reappears from several people, change your mind and say so: "I said no before, and I was wrong." People respect a maker who updates their view in the open.

The reverse is also true. If you say yes and regret it, finish the work, learn from it and tighten your rules for next time.

Mistakes to avoid

  • Saying yes to be liked.
  • Saying maybe to avoid conflict.
  • Building for one loud voice.
  • Adding settings instead of deciding.
  • Letting the list become a promise.
  • Never explaining the focus.
  • Forgetting to ship what you said yes to.

On this site

The blog collects more guides for small teams, the submit page starts a listing and the launches page shows how focused small products describe themselves in a sentence. The founders page tells you who is behind Zivaro Launch, and the contact page is where to send a question.

A short worked example

A founder of a weekly-report tool receives five requests in one fortnight: a dark theme, a calendar view, an export to a spreadsheet, a mobile app and a way to share reports with clients. She logs each with who asked. The spreadsheet export has come from three different people in her target group and takes an afternoon, so she says yes, soon. The client sharing request comes from two people and touches the core job, so it goes on the visible list as under consideration. The dark theme is nice but not now. The calendar view and the mobile app do not fit and are small teams' traps, so she says no, with the kind note and the workaround.

Within a month, the three people who asked for export see it shipped, and everyone else can see on her public list where their idea went and why. Nobody left. Several said thanks for the clarity.

A note on saying no to yourself

The hardest requests to refuse are your own. At two in the morning, a clever feature seems irresistible. Apply the same test: who is it for, does it serve the core job and what would I not do if I did this? Write the idea in your notebook instead of in the code, and look at it again after a night's sleep. Most of them look different by breakfast.

The short version

Say no kindly and often. Write down who the product is for and what it will not do, sort requests by who asked and how many, look for the problem behind the request, reply with a clear yes soon, not now or no, keep a public list and review it monthly. A focused small product beats a sprawling one.

Questions and answers

Why should I say no to feature requests?
Because a small team has limited time, and every feature adds work to build, explain, test and maintain. Focus lets you do a few things well.
How do I decide what to build?
Look for requests that repeat, that come from your target users and that relate to the core job of the product.
How do I say no without upsetting people?
Thank them, explain your focus, say what you are doing instead and keep a visible list so they know you listened.
What is scope creep?
The gradual growth of a project's requirements beyond its original plan, usually without a matching increase in time or resources.
Should I ever build one-off features for a customer?
Occasionally, if it is paid for and fits your direction. Otherwise it can pull the product away from its core.

Sources

Help centre