What goes in
The quick diagnostic works from one message: a startup description of 50 to 4,000 characters sent to our Telegram bot after the /score command. You do not need a deck for it. Four facts are usually enough: who you help, what problem you solve, what already works, and how you earn.
The review reads only that text. It does not search the internet about you, open your website or check your claims against outside sources. Anything the text does not contain is not guessed — it is listed as missing data. That is why a short, vague description gets a low-confidence result: not because the idea is weak, but because there is little to check.
What we check
The review reads a startup description the way an early-stage investor reads a first message: not as slide design, but as a story that either holds together or breaks under the first question.
| Area | The question behind it |
|---|---|
| Problem and customer | Is there a specific person with a specific, repeating pain? |
| Product and maturity | Is it a concept, a working MVP, pilots, or paying users? |
| Traction | What has actually been confirmed, and by what numbers? |
| Market and model | Who pays, how much, and does the logic of earning add up? |
| Team | Is there domain expertise for this particular problem? |
| Arithmetic | Do the numbers in the description reconcile with each other? |
| Dates | Are the milestones in the past or the future relative to today? |
How the 0–100 score is set
The score is a triage signal: it helps you decide what to do next, not whether the company is worth funding. The bands are wide on purpose, so that two different projects do not end up with nearly the same number.
| Score | What it means |
|---|---|
| 75–90 | A strong signal with first confirmations |
| 60–74 | A live idea, part of it already confirmed |
| 45–59 | There is something real here, but data is thin or there are visible gaps |
| 25–44 | A weak signal or major problems |
| Below 25 | Very little that can work in its current form |
Three rules shape where a project lands inside its band.
Maturity moves the score
A working MVP, first users and revenue move a project up within its band; a pilot that is only planned keeps it lower. Two projects with a similar idea but different maturity should not get the same score.
Arithmetic is recalculated, not repeated
If the description contains a calculation — price times number of customers, a percentage, unit economics — the review recalculates it instead of copying your number. A mismatch is shown with the actual figures, for example "5 × 1,900 is 9,500, not 7,200". It is a flag that lowers trust in the economics by a few points; it does not wipe out confirmed strengths such as live pilots or an expert on the team. You can see a real case of this in our full review example.
Dates are compared with today
If a milestone in your description is already in the past — a launch, a pilot start — the review does not say "not launched yet". It says the milestone was planned for that date and asks you to confirm the current status.
What you get back
The result usually arrives in three to seven minutes and has the same structure every time:
- a score from 0 to 100 and a short triage label, such as "wait for pilot metrics" or "look further";
- one headline sentence about your project and its main open question;
- what already works — up to five signals from your text;
- what to verify — the three most important points, phrased as checks rather than accusations;
- what is still missing — up to three facts that would change the picture;
- five questions an investor is likely to ask;
- one next step;
- a confidence level with the reason for it, for example "low because pilot metrics are not confirmed yet".
If the main gap is missing facts, you can send one free clarification the same day: reply with two to five concrete facts, and the score is updated. When the triage label says the project needs first numbers or pilot metrics, we tell you directly that adding facts will help more than a paid review right now.
Quick diagnostic and deep review
The quick diagnostic is free, once per person per UTC day; a run that fails does not use up the slot. It works from text only. The deep review is a separate paid service at $19: it works from your PDF or PPTX deck and returns a slide-by-slide fix map. To order it, message @hi_support_team.
What the method does not do. It is not investment advice, due diligence or a valuation, and it does not predict or guarantee a fundraising outcome. The text is generated by a language model following a fixed script, with no human review. It does not judge slide design, and it does not assess personality, health or any personal traits of the founders — only business signals that are visible in what you sent.
What happens to your text. We do not publish it, do not pass it to investors and do not use it to train models. You can delete your description and result with /delete inside the bot. Details are in our privacy policy.
Which FounderScore this is
This page describes FounderScore at founderscore.pro: a pitch review that works in Telegram through @FounderScore_bot, in English and Russian. We are not affiliated with founderscore.ai or with other products that use a similar name.
See how your own pitch scores
Open the bot and send /score followed by 50–4,000 characters about your startup. You will get the score, the reasons behind it and five likely investor questions.