Intro.
#Why a plan with only a problem definition loses persuasive power
It comes down to one sentence: a problem definition is, in the end, still a claim. "This kind of customer has this kind of friction" — no matter how plausible it sounds, it's a declaration made by the founder. And a declaration always carries two possibilities mixed together.
- A fact that actually exists — someone is genuinely experiencing it, and it's verifiable
- An assumption the founder imagined — a problem plausibly assembled inside their own head
Looking at the problem definition alone, these two look identical. The sentences don't tell them apart. Only evidence can. So a problem definition with no evidence attached leaves a reviewer with no way to determine whether it's fact or assumption.
What matters is which way a reviewer defaults when they can't tell. The short answer: absent evidence, they tend to read it as 'assumption.' Not because reviewers are being harsh — it's because they can't premise an investment or a grant decision on an unverified claim treated as fact. With no evidence, the safer reading — the one that treats it as an unconfirmed assumption — is the only reasonable default. The problem definition is the first domino in your plan, so if it's standing on an assumption, the credibility of everything stacked on top of it — your solution, market size, revenue model — drops along with it.
02
#Why reviewers catch this so fast
The fact that a problem definition has no evidence behind it shows more than you'd think. The reason is simple.
Anyone can imagine a problem. Sit at a desk, stitch together your own experience with a few news articles, and you can write a perfectly plausible description of a problem you've never actually experienced. Which means fluency in describing a problem isn't a signal of fact. If anything, when the writing is fluent but the evidence is absent, reviewers tend to read it as a signal that "this might have been imagined, not lived."
By contrast, the trace of having actually met real customers is hard to fake. A specific customer's actual words, an unexpected detail, a reaction that contradicts the founder's own hypothesis — these are hard to invent from behind a desk. Reviewers see enough business plans that they become attuned to this difference, and they tend to sense whether evidence is present within just a few pages.
In other words, a reviewer isn't grading how polished your problem statement sounds — they're checking whether a real customer is standing behind it. If no one is, then no matter how good the next sections are — market size, solution, team — the foundation of trust is already shaky.
03
#The common signals of an unvalidated business plan — a self-diagnostic checklist
This is the heart of the article. Go through the items below one at a time against the problem-definition section of your own business plan, open right now. The more that apply, the higher the risk it reads as an 'imagined problem.'
Signal A — the subject is 'many people'
- [ ] You describe the problem using a vague, blanket subject like "many people," "most users," or "everyone"
- [ ] There's no sentence that narrows the people experiencing this problem down to a specific group or specific situation
- [ ] You substitute background narrative — "given current trends," "in this day and age" — for an actual description of the problem
Diagnosis: the broader the subject, the more likely you never actually validated it. "Everyone," which is really no one in particular, is impossible to verify by definition — whereas a customer you actually met always has a specific face.
Signal B — the customer's voice is nowhere in the document
- [ ] Not a single line quotes something a customer actually said, or references an actual conversation
- [ ] There's no language pointing to a trace of having actually met people — interviews, surveys, a waitlist, pre-signups
- [ ] Every piece of evidence for the problem is secondary — articles, statistics, industry reports — and nothing else
Diagnosis: secondary sources tell you the market is 'big,' but they can't tell you that your customer actually experiences this problem. "The market is big" and "the specific customer I've defined experiences this specific problem" are claims at two different levels, and it's the latter a reviewer wants to see.
Signal C — severity and frequency are 'guessed'
- [ ] You use intensity words like "extremely inconvenient" or "a major burden" with nothing backing them up
- [ ] There's no confirmed information about how often this problem actually occurs
- [ ] The only thing backing up the severity is the founder's own guess or personal anecdote
Diagnosis: severity and frequency are what determine the actual 'size' of the problem, and if you fill that in with guesswork, the entire size of the problem ends up resting on a guess.
Signal D — you leap from the problem straight to 'willingness to pay' with no evidence
- [ ] You skip straight from "it's inconvenient" to "so of course they'll pay for it" with no validation in between
- [ ] There's no evidence that customers are currently spending money or time trying to solve this problem
- [ ] There's no trace confirming willingness to pay or purchase intent — no paid beta, no pre-payment, no letter of intent (LOI)
Diagnosis: inconvenience and willingness to pay are two different stories. People are inconvenienced far more often than they're willing to open their wallet over it. Make this leap with no evidence, and the entire premise of your business becomes an assumption, top to bottom.
TIP
A one-line self-test: in your plan's problem-definition paragraph, look for who experiences this problem, when, and how often. If you can't state all three with evidence behind them, that problem may still be an 'imagined' one.
04
#Imagined problem vs. validated problem — what's actually different
The same idea can read completely differently depending on whether the problem definition is backed up or not. The two descriptions look similar on the surface, but they leave a reviewer with opposite impressions.
| Dimension | Imagined problem (red flag) | Problem with evidence attached (persuasive) |
|---|
| Subject of the problem | "Many office workers feel friction with ~" | "Office workers we spoke with directly said, repeatedly, that ~" |
| Form of evidence | Founder's inference, general statistics | Interview quotes, pre-signups, surveys, paid beta, LOI |
| Severity/frequency | "It probably happens often" (a guess) | "In this situation, it happens this often" (an observation) |
| Willingness to pay | "They need it, so they'll pay" | "They told us what they'd pay / they actually paid" |
| Falsifiability | No data exists to contradict it (because it's an assumption) | Even the mismatched reactions are documented (because it's reality) |
| How it reads | The founder's assumption | A hypothesis backed by evidence |
| Impression left on the reviewer | "This feels invented, not lived" | "This person actually knows their customer" |
The important thing to notice is that the number in the right column doesn't have to be large. Even a small number, if it's a genuine trace of having met real people, beats an imagined scale of 'many.' A sentence built on a handful of real interviews carries far more trust with a reviewer than a single line claiming "countless people suffer from this." The power of validation comes from being real, not from being big. Existence comes before scale.
05
#So what evidence actually closes the gap (briefly)
This article is meant as a diagnosis, so it doesn't dig into methodology. But it's worth flagging what actually counts as evidence. Attach any one of the following behind your problem definition, and it becomes far more persuasive than a claim left stuck at the level of imagination.
- Customer interview quotes — an actual customer's words, presented unedited
- Pre-signups / waitlist — people who raised their hand and said "I'm interested" before the product even existed
- Survey responses — results obtained by directly asking people who've experienced the problem about its frequency and severity
- Paid beta / pre-payment — a trace that shows willingness to pay through action, not words
- Letter of intent (LOI) — an especially powerful piece of paper evidence from the other party, particularly for B2B
To emphasize it again: what matters at this stage isn't inflating the numbers to look bigger — it's attaching evidence that actually came from real contact with customers. Even a small sample, as long as it has its feet on real ground, elevates everything built on top of it — market size, solution, growth scenario — from 'assumption' to 'hypothesis.' That said, how you actually generate this evidence — how you recruit customers, what questions you ask, how you check for bias-free results — is outside the scope of this article. We'd point you to this series' articles on customer interview methodology and early-stage market validation to continue from here. This article stops at helping you notice whether that gap exists in your own document.
06
#How to spot this gap yourself before you submit
The evidence gap in a problem definition is, ironically, hard for the very person who wrote it to see. A founder has already spent a long time thinking about this problem, so it's easy to confuse the conviction in their head with evidence on the page. "I know this is true — why don't they believe me?" is exactly the trap this mistake sets. Your own conviction is evidence only to you. To a reviewer, it still isn't evidence at all.
Which is why, before you submit, you need to go back through it once from a reviewer's perspective. For every claim in your problem definition, coldly check: "is there customer validation evidence behind this sentence?" Ask yourself, before you hit submit: "is this a fact I confirmed, or an assumption I believed would turn out to be true?" Any sentence you can't answer that question for yourself is exactly the spot a reviewer will flag first.
CTA
If it's hard to check this yourself, one option is to have that reviewer's question put to you in advance, before you submit. Find the empty spots where evidence is missing first, and you can fill them in before you send it out.
Summary.
#Frequently Asked Questions (FAQ)
Q. We're pre-product. How do we include customer validation evidence at this stage?
You can generate evidence even without a product. Pre-signups, a waitlist, interviews, and surveys are all obtainable before launch. In fact, "we validated the problem before we ever built the product" is a stronger narrative than "we couldn't validate because we don't have a product yet."
Q. Doesn't citing market-size statistics or news articles count as validation?
"The market is big" and "my specific customer actually experiences this problem" are claims at different levels. Secondary sources explain the market's backdrop, but they're not evidence you actually met 'your' customer. You need both, and one can't substitute for the other.
Q. I only interviewed a handful of people and I'm embarrassed to even mention it. Doesn't a small sample actually lose points?
If your sample is small, it's better to state that limitation openly, and be specific about who you talked to, why, and what you confirmed. Expanding your validation is a task for the next stage — at the starting point, what matters is leaving a concrete record of how you met people and what you learned.
Q. We're B2B, and meeting customers is especially hard.
The more B2B you are, the more weight a written record like an LOI or documented pilot discussions carries. Even if you've only met a handful of companies, one piece of paper from the other side can be the decisive evidence separating imagination from fact.
Q. How exactly do I run customer interviews or market validation?
This article focused on diagnosing whether validation is missing, so it didn't cover methodology. For how to actually meet customers and confirm a problem, see this series' article on customer interview methods; for validating the market at the early-idea stage, see the article on early-stage market validation.
This article is one entry in OpenSeed Discovery's "Business Plan Mistake Log" series. Rather than any specific review outcome or statistic, it describes a pattern we've observed repeatedly, from a diagnostic angle. A problem definition is a claim too, and claims need evidence — the next entry in this series covers another point-losing pattern.
Check the 'Evidence Gap' in Your Problem Definition Before You Submit
Before you submit your business plan, OpenSeed's AI analysis checks, from a reviewer's perspective, whether your problem definition has customer validation evidence attached. See where the line falls between an imagined problem and a validated one before you ever walk into a real review.
🔒 Free during beta · your submission isn't saved
Start Free AI Feedback →