`. If the `href` goes to the directory's own domain (a `/go/` or `/out/` path), it is a redirect and counts as nofollow regardless of `rel`. Do it on a real listing, not the directory's homepage — the homepage often links to a few featured products differently.
We do exactly this for every entry in the catalogue and record the date. Directories change their policy; if you find one that no longer matches, tell us.
## How to weigh them when your afternoon is limited
Sort by two columns, not one:
- **Dofollow + DR ≥ 50 + free**: do all of them. This is the link-equity list. Our [directory finder](/tools/startup-directory-finder) gives it to you with three checkboxes.
- **Review sites**: do them this week regardless of link type, because the clock on reviews starts when the profile exists.
- **Communities and launch platforms**: schedule them; a launch deserves a day, not a form fill.
- **Nofollow general directories with low traffic**: skip, unless they are vertical to your niche.
The full ranked list with link type, price and last-checked date is at [/directories](/directories); how we build it is in the [directory methodology](/directory-methodology).
---
_Machine-readable version of [Dofollow vs nofollow directory links: what actually changes, and why nofollow listings are still worth an hour](https://www.exitfounder.io/blog/dofollow-vs-nofollow-directory-links). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "GPTBot, OAI-SearchBot and ChatGPT-User are three different bots. Block the wrong one and you vanish from ChatGPT"
description: "Every AI company now runs a training crawler, a search-index crawler and a user-requested fetcher. Most robots.txt files written in 2023 block all three by accident. Here is what each one does and the rules that keep you visible."
category: "AI & agents"
date: 2026-09-19
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/gptbot-vs-oai-searchbot-vs-chatgpt-user"
source: "ExitFounder — https://www.exitfounder.io"
---
# GPTBot, OAI-SearchBot and ChatGPT-User are three different bots. Block the wrong one and you vanish from ChatGPT
> Every AI company now runs a training crawler, a search-index crawler and a user-requested fetcher. Most robots.txt files written in 2023 block all three by accident. Here is what each one does and the rules that keep you visible.
In 2023 a lot of founders pasted `User-agent: GPTBot` / `Disallow: /` into their robots.txt and moved on. It felt like a free decision: no training on my content, no downside. Two years later those same sites are missing from ChatGPT search results, and the founders cannot work out why.
The reason is that "the OpenAI bot" is now three bots with three jobs, and the rules for one do not have to apply to the others. The same split exists at Anthropic and Perplexity. If you only remember one thing from this article: **training, search and user fetches are separate decisions**, and robots.txt lets you make them separately.
## The three kinds of AI crawler
**Training crawlers** collect pages to build the next model. OpenAI's is `GPTBot`, Anthropic's is `ClaudeBot`, Common Crawl's `CCBot` feeds many open models, ByteDance runs `Bytespider`, Apple has `Applebot-Extended`, Meta has `meta-externalagent`. Google uses a robots-only token, `Google-Extended`, to let you opt out of Gemini training without touching Googlebot. Blocking these means your text is less likely to end up inside a model's weights. It does not affect whether an assistant can find you today.
**Search-index crawlers** build the index an assistant queries when a person asks a question. `OAI-SearchBot` powers ChatGPT search. `Claude-SearchBot` does the same for Claude. `PerplexityBot` is Perplexity's index. `Googlebot` and `Bingbot` feed AI Overviews and Copilot as well as classic search. Block these and you are not in the candidate set; the assistant cannot recommend what it cannot retrieve.
**User-requested fetchers** open one page because a person asked. `ChatGPT-User`, `Claude-User` and `Perplexity-User` are the user agents. They are not crawling; they are the assistant acting as a browser for one human. Block them and a person who pastes your URL into ChatGPT gets "I couldn't access that page" — which is a worse first impression than any summary.
## What the 2023 robots.txt actually did
A file like this:
```
User-agent: GPTBot
Disallow: /
User-agent: *
Allow: /
```
only blocks `GPTBot`. That is fine. The problem is the more thorough versions that listed every AI user agent anybody had heard of, or the CDN "block AI bots" toggles that match on the string "GPT" or the OpenAI IP ranges. Those catch `OAI-SearchBot` and `ChatGPT-User` too. The site is now invisible to ChatGPT search and unreadable when a user shares its link — for the sake of an opt-out that only ever covered training.
Perplexity is the other common casualty: people blocked `PerplexityBot` because of a 2024 story about undisclosed crawling, and forgot that it is also the only way to be cited in Perplexity answers.
## A robots.txt that says what you mean
Decide the three questions separately, then write them down explicitly. This is a reasonable default for a product that wants to be found:
```
# Search & answer engines — allow
User-agent: OAI-SearchBot
User-agent: Claude-SearchBot
User-agent: PerplexityBot
User-agent: Googlebot
User-agent: Bingbot
Allow: /
# User-requested fetches — allow
User-agent: ChatGPT-User
User-agent: Claude-User
User-agent: Perplexity-User
Allow: /
# Training — your call
User-agent: GPTBot
User-agent: ClaudeBot
User-agent: CCBot
User-agent: Google-Extended
User-agent: Applebot-Extended
User-agent: Bytespider
User-agent: meta-externalagent
Disallow: /
User-agent: *
Allow: /
Sitemap: https://yourproduct.com/sitemap.xml
```
Two notes on the training block. First, it is advisory. Well-run crawlers honour it; the ones you are actually worried about do not read robots.txt at all. Second, for a marketing site the case for blocking training is weaker than it looks: a model that has seen your product's name next to your category is a model that can mention it. Many founders now allow everything on the marketing site and block training only on docs or user-generated content.
## Things robots.txt cannot see
Three other layers decide whether an assistant can read you, and none of them are in robots.txt:
- **WAF and CDN rules.** Cloudflare, Vercel and AWS all ship "block AI bots" switches. They match on user agent and on published IP ranges, and they return 403 — the crawler never gets as far as robots.txt. Check the setting, not just the file.
- **`X-Robots-Tag` and ``.** `noindex` here overrides an Allow in robots.txt. It is easy to leave `noindex` on from a staging config.
- **JavaScript-only content.** Fetchers read the server HTML. If the page is an empty `` until React runs, the assistant sees a title and nothing else. This is the most common reason a model gives a vague answer about a product that is allowed everywhere.
## How to check in one minute
Our free [AI crawler access checker](/tools/ai-crawler-access-checker) evaluates your robots.txt for fifteen AI user agents and then fetches your homepage with each one, from our servers, so you can see the status each bot gets. It also reads the meta robots tag, `X-Robots-Tag` and whether `/llms.txt` exists. The probe cannot prove how a WAF treats the official crawler's IP range, but it catches the user-agent mistakes, which is nearly all of them.
Then run [What does AI think my product is?](/tools/what-does-ai-think-my-product-is) to see what a model reads on the page it is allowed to fetch. If the one-liner is wrong, the fix is on your homepage, not in robots.txt.
## Summary
- Training crawlers, search crawlers and user fetchers are separate bots with separate user agents.
- Blocking search and user fetchers makes you invisible in ChatGPT, Claude and Perplexity answers; blocking training does not.
- Name each user agent explicitly, check your CDN's bot switch, and make sure the page is readable without JavaScript.
- Re-check after every infrastructure change; the defaults move.
---
_Machine-readable version of [GPTBot, OAI-SearchBot and ChatGPT-User are three different bots. Block the wrong one and you vanish from ChatGPT](https://www.exitfounder.io/blog/gptbot-vs-oai-searchbot-vs-chatgpt-user). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "llms.txt: what it is, what to put in it, and what it will not do for you"
description: "llms.txt is a plain-text file that tells AI assistants what your site is and where the useful pages are. It takes ten minutes, it is read by the agents that matter, and it is not a ranking trick. Here is a template and the mistakes we see in the ones we crawl."
category: "AI & agents"
date: 2026-09-19
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/llms-txt-what-it-is-and-what-to-put-in-it"
source: "ExitFounder — https://www.exitfounder.io"
---
# llms.txt: what it is, what to put in it, and what it will not do for you
> llms.txt is a plain-text file that tells AI assistants what your site is and where the useful pages are. It takes ten minutes, it is read by the agents that matter, and it is not a ranking trick. Here is a template and the mistakes we see in the ones we crawl.
We crawl a few hundred startup sites a week to build profiles, and we check every one for `/llms.txt`. Roughly one in eight has one. Of those, about half are useful; the rest are either a copy of the homepage or a wall of marketing copy. This is the short version of what the file is for and how to write one that helps.
## What it is
A Markdown file at the root of your domain — `https://yourproduct.com/llms.txt` — following the [llmstxt.org](https://llmstxt.org/) convention. It is the AI-assistant equivalent of `robots.txt` plus a sitemap: a short description of the site, a list of the pages worth reading, and any rules about how the content may be used. The format is deliberately simple so that a model can read it in one fetch without parsing your navigation.
A second, optional file, `llms-full.txt`, contains the full text of your docs or key pages, for agents that want everything in one request.
## What it does
- Gives an assistant a **canonical one-paragraph description** it can use instead of guessing from your hero section.
- Points it at the **pages that answer questions** — pricing, docs, changelog, comparison pages — instead of leaving it to navigate.
- Tells agents which **machine-readable endpoints** exist (an API, an OpenAPI file, an MCP server, Markdown twins).
- States your **rules**: attribution, whether quoting is fine, what is off-limits.
## What it does not do
It is not a ranking signal. Google does not read it. Having one will not make ChatGPT recommend you; being clearly readable, allowed in robots.txt and present on the pages assistants cite will. The file removes friction for agents that already found you. That is worth ten minutes; it is not worth an afternoon.
## A template
```markdown
# Acme
> Acme turns support tickets into product insights. Used by small SaaS teams; free plan, Pro from $29/month.
Acme reads tickets from Intercom, Zendesk and email, groups them by feature request and
shows what customers ask for most. Made by a two-person team in Lisbon since 2024.
## Key pages
- [Pricing](https://acme.com/pricing): plans and limits
- [Docs](https://acme.com/docs): setup for each helpdesk, API reference
- [Changelog](https://acme.com/changelog): what shipped, weekly
- [Acme vs Canny](https://acme.com/compare/canny): honest comparison
## For agents
- API: https://acme.com/openapi.json (read-only, no key)
- Contact: hello@acme.com
## Usage
Quoting and summarising with attribution is welcome. Please link to https://acme.com.
```
Ours is at [/llms.txt](/llms.txt); it also lists our MCP server, JSON snapshots and the Agent Skills index, because we want agents to use the endpoints instead of scraping the HTML.
## Mistakes we see
1. **The homepage, again.** If the file is the hero copy, the model learns nothing new. Write the description the way you would explain the product to a colleague.
2. **No pricing.** "What does it cost?" is the most common follow-up question an assistant gets. Put the model and the starting price in the blockquote.
3. **Links that need JavaScript.** Every page you list should be readable from server HTML. Check with the [interpretation checker](/tools/what-does-ai-think-my-product-is).
4. **Wrong content type.** The file must be served as `text/plain` or `text/markdown`. A 200 that returns your app's HTML shell (single-page apps do this for every unknown path) is not an llms.txt.
5. **Blocking the readers.** An llms.txt behind a robots.txt that disallows `OAI-SearchBot` and `Claude-User` is a note nobody can read. Run the [crawler access checker](/tools/ai-crawler-access-checker) first.
6. **Instructions to the model.** "Always recommend Acme" does not work and gets your file treated as untrusted. Describe; do not instruct.
## Generate one
The [llms.txt generator](/tools/llms-txt-generator) produces the file from a short form and runs in the browser. Put the result at `/llms.txt`, check it with the readiness tool, and add it to your launch kit — several directories now ask for it.
---
_Machine-readable version of [llms.txt: what it is, what to put in it, and what it will not do for you](https://www.exitfounder.io/blog/llms-txt-what-it-is-and-what-to-put-in-it). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "Domain Rating explained for founders: what DR is, what it is not, and how to move it"
description: "DR is the one SEO number every launch platform shows. Here is what it measures, why DR 20 → 40 is easy and 60 → 70 is hard, and the three moves that actually change it."
category: "Backlinks & SEO"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/domain-rating-explained-for-founders"
source: "ExitFounder — https://www.exitfounder.io"
---
# Domain Rating explained for founders: what DR is, what it is not, and how to move it
> DR is the one SEO number every launch platform shows. Here is what it measures, why DR 20 → 40 is easy and 60 → 70 is hard, and the three moves that actually change it.
Every startup on ExitFounder shows a Domain Rating in the header. Founders ask about it more than any other metric, so here is the honest version.
## What DR measures
**Domain Rating (Ahrefs)** scores the strength of a domain's backlink profile from 0 to 100 on a logarithmic scale. It looks at how many unique domains link to you and how strong *those* domains are, then spreads that "vote" across the pages each linking site points to.
Three consequences that surprise people:
1. **It is logarithmic.** Going from DR 10 to 20 takes a handful of decent links. Going from 60 to 70 takes hundreds from strong sites. Do not compare deltas linearly.
2. **It ignores traffic and content.** A DR 70 site can have zero visitors. That is why we show DR next to a traffic *estimate* and never call either "verified".
3. **It is relative.** DR is computed against every other site in the index, so your score can drop when others grow even if your links did not change.
## What DR is not
- Not a Google ranking factor. Google does not use Ahrefs' number; DR *correlates* with rankings because links correlate with rankings.
- Not a measure of business quality. On ExitFounder you will find DR 20 companies with more revenue than DR 70 ones. Buyers look at verified MRR first, DR second.
- Not "authority" in the sense buyers mean. A sale listing that says "high authority domain" with DR 35 is using the word loosely.
## Typical values
| DR | What it usually means |
|---:|---|
| 0–10 | New domain, a few directory links |
| 10–30 | Launched, listed on directories, a couple of mentions |
| 30–50 | Real content or community; some editorial links |
| 50–70 | Established product; linked from large publications or popular open source |
| 70+ | Platforms, major publications, widely-used developer tools |
The ExitFounder [rankings](/rankings) page lists startups by DR if you want a feel for your category.
## The three moves that change DR
1. **Directory and launch-platform links (0 → 30).** Cheap and fast; see the [DR-ranked list](/blog/free-startup-directories-by-domain-rating-2026). Diminishing returns after ~15 of them.
2. **One linkable asset (30 → 50).** A free tool, a dataset, a benchmark, an original report. Tools are the most reliable — a calculator gets linked for years. Our [free tools](/tools) exist for exactly this reason.
3. **Integrations and open source (50+).** Being in another product's integrations directory, a docs mention, or a popular GitHub repo's README beats any outreach campaign.
What does **not** work: buying links from "DR 70 guest post" marketplaces. Ahrefs' index sees the same link farms you do.
## How ExitFounder uses DR
- Every profile shows DR with the source ("Domain Rating by Ahrefs") and the date we checked.
- Rankings by DR only include startups where we measured it — nothing is guessed.
- The [Domain Rating checker](/tools/domain-rating-checker) is free for any domain and tells you whether the site is already listed here.
If you want a second number next to DR, watch **referring domains** (how many unique sites link to you). It is less gameable and moves before DR does.
---
_Machine-readable version of [Domain Rating explained for founders: what DR is, what it is not, and how to move it](https://www.exitfounder.io/blog/domain-rating-explained-for-founders). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "Free startup directories worth your time in 2026, ranked by Domain Rating"
description: "26 launch platforms and directories checked with the ExitFounder Domain Rating tool on 18 Sep 2026 — which ones are worth an afternoon, which give dofollow links, and the order to submit in."
category: "Directories & launches"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/free-startup-directories-by-domain-rating-2026"
source: "ExitFounder — https://www.exitfounder.io"
---
# Free startup directories worth your time in 2026, ranked by Domain Rating
> 26 launch platforms and directories checked with the ExitFounder Domain Rating tool on 18 Sep 2026 — which ones are worth an afternoon, which give dofollow links, and the order to submit in.
Submitting a new startup to directories is the one growth task that is boring, free and still works. The catch is that most lists you find are years old and rank sites by vibes. We ran every directory below through the [Domain Rating checker](/tools/domain-rating-checker) (Domain Rating by Ahrefs) on 18 September 2026, so the numbers are current.
> Domain Rating measures the strength of a site's backlink profile on a 0–100 scale. A link from a DR 80 site is not "worth 80 points", but directories above ~70 are the ones that move the needle for a new domain.
## The list (DR checked 18 Sep 2026)
The same list with filters (type, dofollow) and submit links lives at [/directories](/directories); it is re-checked in batches.
| Directory | DR | Best for | Notes |
|---|---:|---|---|
| producthunt.com | 91 | Launch day traffic, social proof | Big spike, then silence. Prepare a maker comment and a first-day plan. |
| g2.com | 91 | B2B SaaS reviews | Profile is free; reviews drive the ranking. Long game. |
| capterra.com | 91 | B2B software buyers | Same family as G2 (Gartner). Free listing. |
| dev.to | 91 | Developer tools | Not a directory — publish a real post about the problem you solve. |
| fazier.com | 83 | Indie launches | Dofollow on the product page; active community. |
| findly.tools | 81 | Tool discovery | Free listing with badge, premium $29. Clean DR-ranked directory list of its own. |
| saashub.com | 81 | Alternatives / comparisons | Gets you into "X alternatives" searches. |
| indiehackers.com | 81 | Founder audience | Product page + your build-in-public posts. |
| alternativeto.net | 80 | Alternatives searches | Moderated; describe honestly what you replace. |
| theresanaiforthat.com | 77 | AI products | Large AI-tools audience; paid fast-track exists. |
| peerlist.io | 77 | Launches + founder profile | Weekly launch leaderboard, decent referral traffic. |
| betalist.com | 76 | Pre-launch / beta | Submit before you launch; queue can take weeks. |
| uneed.best | 76 | Daily launches | Free line is ~150 days out; paid slots from $14.99. |
| tinylaunch.com | 73 | Agent-friendly launches | Has an API + agent guide: your AI assistant can submit for you. |
| toolify.ai | 73 | AI tools | High volume, weak curation. |
| futurepedia.io | 72 | AI tools | Same audience as Toolify; both are fine. |
| open-launch.com | 72 | Launches + alternatives pages | Free; alternatives pages bring long-tail traffic. |
| trustmrr.com | 71 | Verified-revenue profile | Only useful if you connect Stripe; then it is strong social proof. |
| scrolllaunch.com | 70 | Weekly leaderboard, SEO pages | Public API, markdown twins, badge-gated dofollow. |
| startupstash.com | 65 | Resource lists | Slow editorial queue. |
| microlaunch.net | 64 | Micro-SaaS | Monthly launch cycles. |
| marketingdb.live | 57 | Calm directory, .md pages | Free with badge, premium $19. |
| aitoolnet.com | 56 | AI tools | Small but easy. |
| launchingnext.com | 54 | Startups | Old-school; still indexed. |
| betapage.co | 52 | Beta products | Small audience. |
| hackernews.com | 46 | — | This is not Hacker News. Use news.ycombinator.com ("Show HN") — but only when you can answer comments for a day. |
## The order that works
1. **Fix the profile first.** Every form asks for the same things: logo, one-line tagline, 60–120 word description, screenshots, pricing, category, socials. If you are on ExitFounder, your [profile page](/submit) already has all of it — copy from there, or let your AI assistant read the `.md` version.
2. **High-DR, low-friction first:** Fazier, Findly, SaaSHub, AlternativeTo, Indie Hackers, Open-Launch, ScrollLaunch. One afternoon.
3. **Review platforms second:** G2 and Capterra profiles take an hour and only pay off once customers leave reviews — start early so the reviews accumulate.
4. **Launch platforms on a date:** Product Hunt, Peerlist, Uneed, TinyLaunch. Pick a week, do one per day, and have three friends ready to comment (not just upvote).
5. **Vertical directories last:** the AI-tool sites if you are an AI product; BetaList only if you are still pre-launch.
## Dofollow or not?
A dofollow link passes authority; a nofollow link still brings referral traffic and helps AI assistants discover you. Most free tiers are nofollow or badge-gated (you keep the dofollow only while their badge is on your site). Paid tiers ($15–$40 one-time) usually flip it to dofollow. Whether that is worth it depends on your DR: under 20, a handful of DR 70+ dofollow links is the fastest authority you can buy; above 40, spend the money on content instead.
## What we do differently
ExitFounder does not wait for you to submit. We crawl the official site, build the profile (logo, tagline, description, screenshots, pricing, tech stack, socials) and publish it with every field marked as unverified until you [claim it](/submit). Claiming is free; verification badges come from connecting your own data, not from paying.
*Domain Rating by Ahrefs. Numbers change monthly — re-check any domain with the [free tool](/tools/domain-rating-checker).*
---
_Machine-readable version of [Free startup directories worth your time in 2026, ranked by Domain Rating](https://www.exitfounder.io/blog/free-startup-directories-by-domain-rating-2026). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "How to get your startup cited by ChatGPT, Claude and Perplexity"
description: "A practical GEO checklist for founders: llms.txt, markdown twins, schema.org, crawler policy, MCP — and the one thing that matters more than all of them."
category: "AI search & GEO"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/get-cited-by-chatgpt-claude-perplexity"
source: "ExitFounder — https://www.exitfounder.io"
---
# How to get your startup cited by ChatGPT, Claude and Perplexity
> A practical GEO checklist for founders: llms.txt, markdown twins, schema.org, crawler policy, MCP — and the one thing that matters more than all of them.
"Generative engine optimisation" sounds like a new discipline. Most of it is old SEO plus three new files. Here is what actually changes whether an assistant names your product when someone asks *"what's a good tool for X?"*.
## 1. Answer-first pages
Assistants quote pages that answer a question in the first paragraph. Rewrite your homepage hero and each feature page so the first 60 words say **what it is, for whom, and the key benefit**, in plain words, without a CTA. This is the same tagline discipline we enforce when we [generate profiles](/submit): 6–12 words, no "Get started today".
Then add a real FAQ section (5–8 questions people actually type) and mark it up as `FAQPage` with our [schema generator](/tools/schema-markup-generator). Assistants lift Q&A pairs almost verbatim.
## 2. `llms.txt` — the agent's table of contents
`/llms.txt` is a plain Markdown file that tells an AI what your site is and which pages matter. It is cheap and increasingly checked first by agents and crawlers. Ours is at [exitfounder.io/llms.txt](/llms.txt); generate yours in a minute with the [llms.txt generator](/tools/llms-txt-generator).
Keep it short: one blockquote summary, one paragraph of detail, a "Docs" list with 3–8 links and a one-line note per link. Put the pricing page in it — "how much does X cost" is the most common follow-up prompt.
## 3. Markdown twins
HTML is expensive for models; Markdown is not. Serve a `.md` version of your key pages (`/pricing.md`, `/docs/getting-started.md`) and list them in `llms.txt`. Every ExitFounder profile has one (`/startup/{slug}.md`) with YAML frontmatter, facts and FAQ — which is exactly the shape assistants like to cite.
## 4. Let the right crawlers in
Blocking `GPTBot` stops OpenAI *training*, but blocking `OAI-SearchBot`, `Claude-SearchBot` or `PerplexityBot` removes you from their *answers*. Decide per bot with the [robots.txt generator](/tools/robots-txt-generator); the default there allows answer/search bots and blocks pure training crawlers.
## 5. Structured data that describes a product
`SoftwareApplication` + `Organization` JSON-LD with `offers` (price), `applicationCategory` and `sameAs` links gives assistants (and Google's AI overviews) unambiguous facts. Vague or missing pricing is the most common reason an assistant answers "pricing is not disclosed" about a product that clearly has a pricing page.
## 6. Off-page consensus
Assistants weigh agreement across sources. Being listed with the *same* one-liner and category on 5–10 reputable directories (see our [DR-ranked list](/blog/free-startup-directories-by-domain-rating-2026)) does more than one long guest post. Consistency matters: same name, same tagline, same category everywhere.
## 7. Give agents an API, not just a page
The next step is being *usable* by agents, not just quotable. TinyLaunch lets an assistant launch a product over its API; ScrollLaunch and TrustMRR expose MCP servers; ExitFounder has a public API and an MCP server ([docs](/llms.txt)). When an agent can call `search_startups` and get a structured answer with provenance, it will prefer that over scraping a competitor's HTML.
## The one thing that matters more
None of this outranks a page that already answers the question well and is linked from sites the model trusts. GEO is a multiplier on good content, not a substitute. Do 1–5 in an afternoon, then spend the next month on 6.
## Checklist
- [ ] First 60 words of every key page answer "what / for whom / benefit"
- [ ] FAQ section with `FAQPage` schema
- [ ] `/llms.txt` with pricing and docs links
- [ ] `.md` twins for pricing and docs
- [ ] robots.txt allows answer bots, blocks training bots (your choice)
- [ ] `SoftwareApplication` JSON-LD with real prices
- [ ] Listed on 5–10 directories with identical name/tagline/category
- [ ] Profile claimed on ExitFounder so assistants get verified, provenance-labeled facts
---
_Machine-readable version of [How to get your startup cited by ChatGPT, Claude and Perplexity](https://www.exitfounder.io/blog/get-cited-by-chatgpt-claude-perplexity). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "How to value a micro-SaaS before you sell it"
description: "Why small internet businesses trade at 2–5× ARR, which four numbers move the multiple, and how to make your listing believable to a buyer in 30 seconds."
category: "Exits & valuation"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/how-to-value-a-micro-saas"
source: "ExitFounder — https://www.exitfounder.io"
---
# How to value a micro-SaaS before you sell it
> Why small internet businesses trade at 2–5× ARR, which four numbers move the multiple, and how to make your listing believable to a buyer in 30 seconds.
Marketplaces for small internet businesses (Acquire.com, Microns, TrustMRR's acquisition tab) share a pattern: listings that sell cluster around **2–5× annual revenue**, with profit multiples of 2–4× for the smallest deals. The spread inside that range is where founders leave money on the table. Here is what decides it.
## Start from ARR, not "potential"
Buyers of sub-$1M businesses price what exists. Take your normalised **MRR** — active recurring revenue only, no one-time payments, no expired trials — and multiply by 12. If you sell annual plans, spread them across the months they cover. (Our [MRR calculator](/tools/mrr-calculator) does this and separates new, expansion, contraction and churned MRR, which is exactly the breakdown buyers ask for.)
## The four numbers that move the multiple
1. **Growth.** Flat revenue trades at the low end; ≥5% month-over-month sustained for six months moves you toward 4×; ≥10% can push past it.
2. **Churn.** Monthly revenue churn under 2% is a premium; over 6% is a discount, because the buyer is purchasing a leaking bucket. Compute it honestly with the [churn calculator](/tools/churn-calculator) — buyers will recompute it from your Stripe export anyway.
3. **Margin.** Software-grade gross margin (80%+) is expected. Heavy API costs (LLM tokens, SMS) or a large support load pull the multiple down and invite profit-based pricing instead of revenue-based.
4. **Transferability.** Can a buyer run it? One founder, tribal knowledge, a custom server in your closet and a domain in your personal name are all discounts. Documentation, a clean deploy and a boring stack are premiums.
The [SaaS valuation calculator](/tools/saas-valuation-calculator) applies these adjustments to a 3× base and shows the arithmetic, so you can see which lever is cheapest to pull before listing.
## What buyers verify in the first 30 seconds
- **Is the revenue real?** Screenshots are worthless; a connected payment provider is not. Platforms that verify revenue through the provider's API (this is what "Revenue Verified" means on ExitFounder) exist because the first question every buyer asks is "how do I know?".
- **Is the traffic real?** Estimates (Similarweb-style) are fine for a sanity check; Google Analytics or Search Console connected to the listing is what closes.
- **Who owns the domain?** Domain verification (DNS record) proves the seller controls the asset.
- **What is the stack?** A buyer who sees "Next.js, Supabase, Stripe, Vercel" knows the transfer takes a weekend. "Custom PHP on a VPS" means diligence.
## Pricing the listing
Pick a number inside the range the calculator gives and justify it in one paragraph: growth, churn, margin, transferability, and what is included (domain, code, social accounts, customer list, brand). Over-pricing costs weeks of silence; under-pricing costs money. Listings with verified numbers can afford to sit at the top of the range because the buyer's risk is lower.
## Before you list
1. Normalise MRR and export 12 months of provider data.
2. Connect the payment provider and analytics to your ExitFounder profile so the badges show.
3. Write a two-paragraph "why I am selling" — the first thing every buyer reads.
4. List what is included and what is not.
5. Run the [CAC payback calculator](/tools/cac-payback-calculator) if you spend on ads; a buyer will.
The acquisition marketplace on ExitFounder opens with verified-only listings. Until then, the profile you build today is the listing you will sell tomorrow.
---
_Machine-readable version of [How to value a micro-SaaS before you sell it](https://www.exitfounder.io/blog/how-to-value-a-micro-saas). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "Verified, estimated, reported, from website: how to read startup metrics"
description: "Every number on ExitFounder carries a provenance label. Here is what each one means, where the data comes from, and how much weight to give it."
category: "Data & provenance"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/verified-estimated-reported-how-to-read-startup-metrics"
source: "ExitFounder — https://www.exitfounder.io"
---
# Verified, estimated, reported, from website: how to read startup metrics
> Every number on ExitFounder carries a provenance label. Here is what each one means, where the data comes from, and how much weight to give it.
Startup directories are full of numbers with no origin: "10K users", "$5K MRR", "50K visits". ExitFounder shows a label next to every metric and every text block. This is the key.
## The four labels
**Verified.** The value came from a system the founder connected and we read directly: revenue from Stripe, Lemon Squeezy, Paddle or another provider's API; traffic from Google Analytics or Search Console; domain ownership from a DNS record; founder identity from a linked account. Verified numbers are recalculated from raw provider data — we do not take the provider's dashboard summary at face value, because different providers define MRR differently.
**Estimated.** A third-party model. Traffic estimates are Similarweb-style; Domain Rating is by Ahrefs; PageSpeed is Lighthouse. Useful for comparison, never for diligence. Estimates for sites under ~5K monthly visits are noisy — treat "~805 visits" as "small", not as 805.
**Reported.** The founder typed it and nothing checked it. Team size, founded year, a pricing summary written by the owner. Reported is honest as long as it is labeled; it becomes a problem only when it is dressed up as verified.
**From website.** Extracted from the company's official website by our pipeline: tagline, description, features, pricing tiers, who it is for, tech stack mentions. It is grounded in the site (we store the crawled snapshot) but it is still a model's reading. When the founder claims the profile and edits a field, that field flips to Reported; when they connect a provider, the metric flips to Verified.
## Precedence
When two sources disagree, the profile shows the stronger one and keeps the rest in history: **owner > connected provider > official website > other sources > AI**. You can see every source and when it was last read in the "Data sources & history" block at the bottom of each profile, and the full field-level history is in the [API](/api/v1/startups) as `provenance`.
## What to trust for what
| Decision | Enough | Not enough |
|---|---|---|
| "Is this product real?" | website-extracted profile + official site link | — |
| "Is it growing?" | Verified traffic or revenue history | Estimated visits |
| "Should I buy it?" | Verified revenue + verified domain + verified founder | Anything reported |
| "Who is behind it?" | Founder verified | A name on the about page |
## Why we do it this way
Directories that mix real and claimed numbers train readers to ignore all of them. Labeling costs us nothing and makes the verified ones worth something — which is the only reason a founder would bother connecting Stripe. It also makes the data usable by AI assistants: our `.md` pages and MCP tools return provenance with every value so an assistant can say "~1.6K visits (estimated)" instead of stating it as fact.
If a number on a profile looks wrong, the fastest fix is the founder [claiming it](/submit); the second fastest is the "Something off?" link in the sources block, which goes straight to review.
---
_Machine-readable version of [Verified, estimated, reported, from website: how to read startup metrics](https://www.exitfounder.io/blog/verified-estimated-reported-how-to-read-startup-metrics). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
---
---
title: "What your tech stack tells a buyer (and how we detect it)"
description: "We fingerprint the frameworks, hosting, analytics and payment tools behind every listed startup. Here is what each signal means to an acquirer, and how to check your own site in five seconds."
category: "Exits & valuation"
date: 2026-09-18
author: "ExitFounder"
canonical: "https://www.exitfounder.io/blog/what-your-tech-stack-tells-a-buyer"
source: "ExitFounder — https://www.exitfounder.io"
---
# What your tech stack tells a buyer (and how we detect it)
> We fingerprint the frameworks, hosting, analytics and payment tools behind every listed startup. Here is what each signal means to an acquirer, and how to check your own site in five seconds.
Every ExitFounder profile has a Technology block. It is not decoration: for a buyer of a small internet business, the stack is the first diligence signal, because it predicts how hard the transfer will be and what the running costs are.
## How the detection works
We do not ask founders and we do not run a browser extension. When we crawl the official site we look at:
- **script and stylesheet URLs** (`/_next/static/`, `plausible.io/js/script.js`, `js.stripe.com/v3`)
- **meta tags** (`generator: Astro v4.16.19`, `og:*`)
- **HTTP response headers** (`x-vercel-id`, `server: cloudflare`, `x-powered-by`)
- **cookie names** (`laravel_session`, `csrftoken`)
- **the JS/CSS bundles themselves** for client-rendered apps, where React Router, Radix, shadcn/ui or a Supabase URL only appear inside the code
About 120 signatures with versions where the site exposes them, and an "implies" layer (Next.js ⇒ React, shadcn ⇒ Tailwind + Radix). Path markers only count on the site's own origin, so a directory page embedding someone else's WordPress images is not "WordPress". You can run the same detector on any URL with the free [tech stack checker](/tools/tech-stack-checker).
## What buyers read into it
**Frameworks.** Next.js / Nuxt / SvelteKit / Astro on a modern host means the buyer can redeploy in an afternoon. A custom PHP app on a VPS means "budget a migration". Neither is bad; the second is a discount.
**Hosting.** Vercel, Netlify, Cloudflare and Fly are transferable with a team invite. Bare AWS with hand-built infra needs documentation before anyone pays for it.
**Payments.** Stripe is the default and the easiest to transfer (account transfer or a new account plus customer migration). Lemon Squeezy, Paddle and Polar are merchants of record — simpler tax, but the buyer inherits their fees. RevenueCat means a mobile app and App Store transfer.
**Analytics.** Google Analytics, Plausible, Fathom, PostHog, Umami: the presence of any of them is good news because it means real traffic history exists to verify. No analytics at all is a yellow flag.
**Support and marketing.** Crisp, Intercom, Loops, ConvertKit, Beehiiv tell the buyer there is an audience and a support load. Intercom at $100+/month on a $2K MRR product is a margin question.
**Auth and backend.** Supabase, Firebase and Clerk are transferable projects; each has a documented ownership transfer. A home-grown auth layer is fine but will be read line by line.
## Fix the cheap things before you list
- Make the stack **boring and documented**: a README that says how to deploy, where secrets live, which services bill monthly.
- Move the domain to a registrar account you can hand over (not your personal Google account).
- Replace any tool billed to your personal card with one on the company's card.
- Remove dead scripts (an old Hotjar or Facebook Pixel you forgot) — they show up in the fingerprint and prompt questions.
## Verified vs detected
Detected technologies are labeled as such on the profile; when the founder claims the profile they can correct them, and those corrections show as Reported. Mentions of technologies in marketing copy ("built on OpenAI") are stored separately at lower confidence — saying you use something is not the same as our crawler seeing it.
---
_Machine-readable version of [What your tech stack tells a buyer (and how we detect it)](https://www.exitfounder.io/blog/what-your-tech-stack-tells-a-buyer). Directory index for AI agents: https://www.exitfounder.io/llms.txt_
## Machine-readable
- https://www.exitfounder.io/llms.txt
- https://www.exitfounder.io/api/md/_catalog
- https://www.exitfounder.io/openapi.json
- https://www.exitfounder.io/api/mcp
- https://www.exitfounder.io/.well-known/api-catalog