Not selling? How to tell a P/MF problem from a go-to-market problem
Short answer: If the few customers you do win stay, use the product and tell others about it, you most likely have a sales and marketing problem. If they churn, stall after onboarding or need constant pushing, it's a P/MF problem, and more marketing will only make it more expensive. Look at what your best customers do, not at your funnel.
Every founder I meet who's stuck asks some version of this question.
And almost every one of them is hoping for the same answer: "It's marketing." Because marketing is fixable. Hire a growth person, change the ads, done.
Sometimes that's true. Often it isn't. In my experience, marketing problems and product problems usually come from the same place: a value proposition that doesn't fit the customer.
Here's how I investigate it.
The core idea: P/MF lives upstream of your funnel
Product/Market Fit is four variables, solved in order:
- ICP (who it's for)
- Value Proposition (why they should care)
- Product (how you deliver on the promise)
- Business Model (how value flows back to you)
Sales and marketing sit downstream of all four. They carry your value proposition to your ICP. If the first two variables are wrong, no amount of sales skill fixes it. You're just delivering the wrong promise to the wrong people, faster.
So the question isn't really "sales or P/MF?" It's "which variable is broken?"
Five clues to check
Clue 1: What happens after the sale?
This is the strongest clue, so start here.
- Customers stay, expand, renew. The product delivers. Your problem is getting more of these customers. That's sales and marketing.
- Customers churn within a few months, or never really activate. The promise you sold isn't the value they got. That's P/MF.
A leaky bucket doesn't need a bigger hose.
Clue 2: Is anyone pulling?
Look for movement you didn't cause:
- Referrals and word of mouth
- Inbound requests from people you never contacted
- Users asking "when can I get this?" or complaining loudly when something breaks
Complaints are a good sign, by the way. People only complain about things they rely on.
If every single customer came from you pushing, and none came from customers pulling, that's a P/MF clue.
One quick test: ask a happy customer who else they'd recommend you to. "I know a bunch of people with the same problem" means there's a market out there. "I can't think of anyone" means your niche may be too tight to build on. It's a market-size probe disguised as a sales ask.
Clue 3: Do your best customers look alike?
Take your five best customers. Not the biggest. The ones who get the most value with the least effort from you.
- They share clear traits (same trigger, same pain, same role or situation). Good. You have an ICP. Your marketing is probably talking to everyone else.
- They have nothing in common. You don't have an ICP yet. That's variable one, and it's a P/MF problem.
Then ask one more question: are your happy customers typical, or special?
- Typical: they represent a larger segment you just can't reach or activate fast enough. That's distribution.
- Special: each one turns out to be a one-of-a-kind edge case doing something very specific. The market is too small. That's P/MF.
Clue 4: Where does the deal die?
Map the stage where prospects drop off:
| Where they drop | What it usually means |
|---|---|
| They never respond | Wrong channel, or a message that doesn't call out their pain. Likely marketing. |
| Great first call, then silence | The pain isn't severe enough to act on. Likely P/MF (ICP or value proposition). |
| They try it and don't come back | The product doesn't reach the Aha! Moment. P/MF (product). |
| They love it but won't pay | Business model, or a "nice to have". P/MF. |
Clue 5: Run the "Hold and Listen" test
Turn off your paid acquisition for a few weeks (or simply look at the organic slice of your traffic) and watch what remains.
If usage and sign-ups collapse to zero, your growth was bought, not earned. An aggressive marketing budget can fake traction for a while. It can't fake retention.
A quick verdict table
| Signal | Sales and marketing problem | P/MF problem |
|---|---|---|
| Retention of won customers | High | Low |
| Inbound and referrals | Some, growing | None |
| Best customers | Look alike | Random |
| Drop-off point | Top of funnel | After the first call or after onboarding |
| Without paid ads | Slows down | Stops |
Two or more marks in the right column? Stop optimising the funnel. Go back to your ICP and value proposition.
A real case: RallyUp
When Alec, RallyUp's founder, came to me, the problem looked like activation.
RallyUp sold employee-generated content to B2B companies: help your employees post on LinkedIn, and their reach brings the company pipeline. Companies signed up. Employees didn't post.
That's Clue 4: they signed up, tried it, and didn't come back.
The obvious move was to treat it as an engagement problem: better onboarding, nudges, rewards. We didn't. We ran a handful of customer interviews first, and they were enough.
The clue: most employees have no reason to post. There's no intrinsic motivation. Rewards and prizes can lift engagement for a while, but it fades the moment you stop paying for it. In my experience, missing intrinsic motivation is one of the biggest red flags there is. Worse, their network is often the wrong audience. A DevOps engineer at an e-commerce platform is connected to other engineers, not to the online sellers the company wants to reach.
No sales or marketing fix solves that. It was a product problem.
We laid out the options: run experiments to rescue activation, or pivot. We chose to pivot.
To decide what to change, we checked the four variables in order:
- ICP: B2B executive teams who want pipeline from LinkedIn. Still right.
- Value Proposition: attention for the brand, inbound, meetings booked. Still right. Nobody ever wanted "employees who post". They wanted what the posting was supposed to bring.
- Product: wrong. Getting employees to post was the wrong mechanism for delivering the promise.
- Business Model: had to change with the product. Change a variable and everything downstream of it moves.
We kept the ICP and the promise, and changed the two variables below them. The product went from software that activates employees to a service that builds the CEO's own audience on LinkedIn. The business model went from a SaaS subscription to an agency retainer. Same customer, same promise, a different way to deliver value and a different way to capture it.
Inbound came in. Conversions followed. MRR has grown every month since, and RallyUp is now closing in on $1M ARR.
The pivot wasn't a new market or a new pitch. It was a new answer to variables three and four.
Common mistakes
- Blaming the salesperson. Founders replace their first sales hire and the second one fails too. When two good people fail at the same thing, look at the thing.
- Reading compliments as evidence. "Great demo!" is not a signal. A signed pilot, a card on file or a referral is. Look for feedback, not compliments.
- Changing the message when the ICP is wrong. Rewriting copy for an audience that doesn't have the pain is rearranging deck chairs.
- Diagnosing from averages. Your overall churn might look fine while one segment retains beautifully and another bleeds. Split by segment before you decide.
What to do next
- Pull retention and activation for your won customers, split by segment.
- Write down the shared traits of your five best customers.
- Mark where deals die.
- Score yourself on the verdict table.
If the answer is P/MF, the fix starts with variable one: your ICP. Here's how to define your ICP with signals, not labels.
Still not sure which variable is broken? That's exactly what I look for in a P/MF Sanity Check. Tell me where you're stuck.