Saying No Well
A no that keeps the relationship is a skill, and it is mostly about what you say instead.
You are in a roadmap review. The VP of Sales asks for a feature that would help close one specific logo. You know it is wrong for the product and you know the quarter is full. You say no. The temperature in the room drops, and for the next two months you find out about things after they are decided. The no was correct. The way it landed cost you more than the feature would have.
Most PM advice about saying no stops at 'say no more'. That is the easy half. The hard half is that every no is a small withdrawal from a relationship you need, and the people asking are not being unreasonable. They have a target, they have information you do not, and they came to you because they think you can help. A good no treats all of that as true.
A no that keeps the relationship
Why bad nos go wrong
- The no is about capacity. 'We do not have room' invites a negotiation about capacity, which you will lose, because someone senior can always create room.
- The no is about priority without showing the ranking. 'It is not a priority' sounds like 'you are not a priority', and there is no way to respond to it except escalation.
- The no arrives late. Three weeks of silence and then a refusal reads as avoidance, and they have already told the customer it is coming.
- The no offers nothing. A refusal with no alternative leaves them exactly where they started, except now they are also annoyed with you.
The common thread is that a bad no is about you: your capacity, your priorities, your process. A good no is about the decision: what is being traded, who owns the trade, and what happens next for the person in front of you.
A no that costs you
- 'We do not have bandwidth this quarter.'
- 'That is not a priority right now.'
- 'That does not fit our vision.'
- Silence, then a refusal three weeks later.
A no that holds
- 'Doing that means dropping the billing work, which two accounts are renewing on. You pick.'
- 'Here is the ranked list and where this lands. If it should be higher, tell me what moves down.'
- 'Here is the problem we think it solves, and here is a smaller thing we could do in two weeks. Would that unblock the deal?'
- 'No, and here is why, within two days of you asking.'
The four part no
- Show you understood
- Repeat the situation back, not the request. 'You need something to show security reviewers at enterprise accounts before contract stage.' If you can state their problem better than they did, the rest of the conversation is easier.
- Name the trade, not the capacity
- Say what specifically would not happen. Vague fullness invites argument. A named casualty with a name attached to it turns a refusal into a choice.
- Offer the cheaper path
- Almost always there is a partial: a manual process for one account, a report you run by hand each month, an existing workaround nobody showed them. It may be ugly. Ugly and this week often beats elegant and never.
- Say what would change your mind
- 'If two more enterprise deals stall at the same step this quarter, this moves to the top.' This converts a no into a not yet with a condition, and it gives them something to do other than escalate.
Worked example
Parcelbird, a shipment tracking tool
Sales wanted a white label tracking page for one retail prospect, scoped at five weeks. The PM's reply, sent the next morning: the prospect needs customers to see tracking without leaving their brand, that is a real gap and it will come up again. Building it this quarter means dropping the carrier exception alerts, which three renewing accounts named in their QBRs. In the meantime, the tracking page already supports a custom logo and colour, and a customer domain via CNAME takes an afternoon. Would that get the deal to signature? It did, for that deal. The full white label version was funded the next quarter when two more prospects hit the same wall, which was the condition the PM had written down.
The PM did not win that exchange by being firm. They won it by doing the work: knowing what the trade actually was, knowing what the product could already do, and answering fast enough to be useful.
The part people skip
Write down your nos and revisit them. A list of what you declined, with the date and the reason, does three things. It shows you patterns, and three nos to the same underlying situation in a quarter is a yes you are missing. It protects you, because when someone says nobody ever asked for this, you have the record. And it keeps you honest, because you will find nos you got wrong, and finding them yourself is much better than having them found for you.
Last thing. Some nos are not yours to give. If the request comes from a founder with a strategic bet behind it, or a regulator, or a contract already signed, the honest move is to say so rather than to perform a decision you do not own. Saying 'this is not my call, here is what it costs, let us go ask' keeps your credibility for the nos that are yours.
Quick check
Why is 'we do not have bandwidth this quarter' a weak way to decline a request?
The takeaway
A good no names the specific trade, offers a cheaper path, and states the condition that would change the answer.
Try this tomorrow
Start a running doc of requests you decline, with the date, the reason and the trade you named. Review it at the end of the month and look for any underlying situation you have now declined three times.
Answer the check above, then bank the day.
Where this comes from
- Rework, Jason Fried and David Heinemeier Hansson
- Empowered, Marty Cagan and Chris Jones
- Escaping the Build Trap, Melissa Perri
Product Thinking is one of six tracks. These lessons summarise and build on the work above, they do not reproduce it. Buy the books, they are better.