Three Filters I Use Before Saying Yes to a Project
I said no to a project recently. Two years ago I would have said yes without thinking. Every yes spends time and attention I’ve already promised to someone else, so I now run each offer through three filters before I answer.
The yes trap
Early in my consulting career I said yes to everything. Every project. Every speaking gig. Every “quick favour.”
The result:
- 60-hour weeks
- Half-finished projects
- Mediocre work across the board
- Burnout within 18 months
The problem wasn’t effort; I never chose what to drop. Saying yes to everything means saying no to doing anything well, and the things that suffer are usually the ones that matter most.
The trap is that every individual yes looks reasonable. The project is interesting, the client is nice, the money is good, and it’s “only a few hours a week.” The cost only shows up in aggregate, weeks later, when four “few hours a week” commitments collide in the same fortnight.
Three filters before I say yes
Now every opportunity goes through three filters. It has to pass all three.
Filter 1: Does it align?
Does this project fit my focus areas? Right now that’s Azure OpenAI, Microsoft Fabric, and AI consulting.
- A React frontend project? Probably no, even if the money is good.
- A Power BI dashboard? Maybe, if it connects to Fabric.
- An AI chatbot on Azure? That’s my lane.
The filter doesn’t have to be all or nothing. Plenty of requests are mixed: a Fabric data platform with a React frontend bolted on. You can decline part of a project, not only all of it. “I’ll take the Fabric part and refer the React frontend” is a perfectly good answer, and in my experience most clients would rather hear that than have a consultant quietly do the half they’re weaker at.
This filter matters more than it looks. Every off-focus project costs more than its hours. It pulls attention away from the area where I’m building depth, and it trains people to come to me for things I don’t want to be known for.
Filter 2: Can I do it well?
Not “can I technically do it.” The question is whether I can deliver excellent work given what I’ve already committed to.
If I’m already at 80% capacity, a new project means one of two things: something slips, or everything degrades a little. Neither is visible on the day I say yes. Both are visible to the client a month later.
Capacity is easy to overestimate, so check it rather than guess. Add up the hours you’ve actually committed per week across current work and compare that to a ceiling you set in advance, not the number you could survive on. My rule of thumb is to keep committed hours at or below about 80% of a normal week, leaving room for the slippage every project has. Then look at your deliverables: if anything you’ve already promised is running late, you’re over capacity regardless of what the hours say.
I’d rather be known for a few excellent projects than a long list of mediocre ones. In consulting, the next piece of work usually comes from how well the last one went.
Filter 3: Does it energise me?
If the thought of starting one makes me dread Monday, it’s a no.
Energy also affects quality. Motivated work is better work: you notice the edge cases and you push back when something is wrong. Work I resent gets the minimum. Clients can tell the difference even when they can’t name it.
The catch is telling real dread from ordinary nerves. Almost every new project carries some start-of-project resistance: a new client, an unfamiliar codebase, a kickoff you’d rather skip. My test is to picture week three, after the kickoff and the unfamiliar codebase are behind me. If I’d still dread opening the laptop, it’s the work, not the nerves, and that’s a no. If the dread is all front-loaded, it’s the friction of starting, and that alone isn’t a reason to decline.
How to say no
Be honest, and be prompt.
| Approach | What it looks like | Why it fails or works |
|---|---|---|
| Ghosting | No reply at all | Unprofessional, and it burns the bridge |
| Stalling | ”Let me think about it” for weeks | Wastes everyone’s time; they can’t plan |
| Clear decline with a referral | ”Thanks for thinking of me. This doesn’t align with my current focus, but I’d recommend [name]. They’d be great for this.” | Respects their timeline and keeps the door open |
Refer whenever you can. It builds goodwill with the person you recommend, and it helps the client solve their actual problem. A good referral often does more for the relationship than a reluctant yes would have.
A fast no is also kinder than a slow one. The client has a deadline. Every week you sit on the decision is a week they can’t spend finding someone who wants the work.
The hardest part
Fear of missing out. What if this was the project? What if they never call again?
They might not. That’s okay.
The projects you turn down make room for better ones you can’t see yet. If your calendar is full of work that only half fits, there’s nowhere to put the right project when it shows up.
What changed after saying no more often
Quality went up. Fewer projects means more attention per project. Clients noticed.
Revenue didn’t drop. Focused work led to better referrals and more repeat clients. My read is that people find it easier to refer someone whose lane is obvious.
Stress went down. Not juggling six projects means actually sleeping.
Reputation improved. People now know what to bring me, so the offers that arrive fit better in the first place.
This lines up with my January retro: I turned down two consulting gigs that month, and taking weekends off was the change I was proudest of. Both came from the same decision: protect capacity instead of spending it by default.
When to say yes anyway
This advice has limits. Selectivity is much easier once referrals are flowing. If you’re early in your career or early in building a practice, a project that only half fits can still be the right call, because you need the experience, the client and the reference more than you need a clean focus.
The same goes when cash flow is tight or the pipeline is thin. Turning down paid work to protect your focus only makes sense if you can afford the gap. Two other exceptions I’d make: a stretch project you take deliberately to build a new focus area, and adjacent work for an existing client whose relationship is worth protecting. In each case, say yes knowingly, with a clear reason, rather than by default.
The rule
Derek Sivers’ “hell yeah or no” rule (from his 2009 post “No more yes. It’s either HELL YEAH! or no.”), in my words: if it’s not a clear yes, it’s a no.
It applies to projects, to meetings, and to most decisions. When an offer fails one of the three filters and I find myself building arguments for why it’s fine anyway, that’s the signal to decline.