One score, on every contact
A single 0 to 100 value on the contact record, visible in the list, the board and the conversation, so it is part of the workflow rather than a report nobody opens.
LEAD SCORING
Know which twenty contacts to call first
Every contact carries a lead score out of 100. AI can set it from what the person actually wrote, the browser extension can score a profile before you save it, and you can override any of it by hand. Then sort, filter and sequence on it.
- A score out of 100
- Set by AI from real signals
- Always overridable
app.crmsolid.com/contacts
Sorted by scoreContacts
Elena Ruiz
Asked for pricing twice · 6 seats
Tom Bauer
Opened the proposal, no reply yet
Mara Ilic
Downloaded the guide
Ken Yamada
Newsletter signup only
Elena Ruiz · 91
Why
- Named a team size and a deadline
- Replied within an hour, twice
- Asked about the annual price
You can change this score by hand at any time
A number that survives contact with a real pipeline
Scoring is easy to build and easy to ignore. It only gets used when the score is visible where the work happens, explainable, and overridable by the person who knows better.
AI scores from what was written
Email and conversation analysis reads the actual thread and scores lead quality from it: a stated budget, a deadline, a team size, a decision maker in the reply.
Scored before you even save it
On paid plans the browser extension scores a profile while you are looking at it and says in one line why the person is worth your time, before they become a contact at all.
A deeper read on X profiles
For X specifically there are two separate numbers, lead quality and reach, plus a fake-follower audit, because a large audience and a useful audience are not the same claim.
Override it by hand
A salesperson who has spoken to somebody knows more than any model. The score is editable on the contact, and your edit stands.
Sort and filter the pipeline by it
Order a list by score, filter a board to everything above a threshold, and stop deciding who to contact next by scrolling.
Feed it into sequences
Enrol the high scores in a hands-on sequence and the low ones in a slower nurture, so the effort matches the value rather than being spread evenly.
The AI agent sees it too
When an AI agent answers a conversation, the contact's lead score is part of what it knows, so a high-value thread is handled differently from a casual question.
Readable and writable over the API
The lead score is available through the public REST API and the MCP server, so an external model or a scoring job of your own can set it.
What lead scoring is here
Lead scoring puts a single number between 0 and 100 on a contact to express how worth pursuing they are. In CRM Solid that number can be set by AI reading what the person actually wrote, by the browser extension reading a profile before you save it, or by a human who has spoken to them and knows better. It is stored on the contact and used everywhere: sorting, filtering, sequences and the AI agent's own context.
The point is triage. Once a pipeline is bigger than a person can hold in their head, the first decision every morning is who to contact, and that decision is usually made by whoever is at the top of an alphabetical list. A score is the cheapest way to make it a decision about value instead.
Where the score comes from
Three sources, and you can use any of them alone.
- Conversation and email analysis: the AI reads the thread and scores what is in it. A named budget, a deadline, a team size or an explicit next step move the score in a way that a page view never should.
- Profile scoring at capture: the browser extension scores a profile before it becomes a contact and says why in one line, so a prospecting session is triaged as it happens.
- Human judgement: anybody can set the score by hand on the contact record, and it is also writable through the public API and the MCP server if you want your own model to set it.
Why an override matters more than the model
Every scoring system eventually produces a number a salesperson knows is wrong. The customer mentioned on a call that budget was approved. The enthusiastic replies came from an intern. None of that is in the data the model read.
A score you cannot override becomes a score people work around, and once a team is working around a number they stop trusting the whole system. Making the override first-class, on the record and visible, is what keeps the automatic part useful.
What it costs
Lead scoring is available from Pro at $24.99 a month, along with the AI features that populate it and the browser extension's AI enrichment. Deep X profile analysis is also a Pro feature. The Free plan keeps the contact database and Telegram, without the scoring layer.
Frequently asked questions
A single number between 0 and 100 on a contact, expressing how worth pursuing they are, so a team can sort a pipeline by value rather than by alphabet or by whoever shouted loudest.
Three ways, usable together: AI reading the actual conversation or email thread, the browser extension scoring a profile at the moment you capture it, and a human setting it by hand. It is also writable through the API and the MCP server.
Yes, on the contact record, at any time. A person who has spoken to the lead knows things the model cannot see, and a score nobody can correct is a score nobody trusts.
Yes. Sort a contact list by score, filter a pipeline board to everything above a threshold, and enrol high scores in a hands-on sequence while low scores get a slower nurture.
Yes. X profiles get two separate numbers, lead quality and reach, plus a fake-follower audit and an account valuation, because a big audience and a useful audience are different claims.
Pro at $24.99 a month, along with the AI analysis that sets it and the extension's AI enrichment. Free includes the contact database without the scoring layer.
What the score plugs into
Ready to ship
Stop working the pipeline alphabetically
Lead scoring and the AI analysis behind it are included from Pro.
Free forever plan · GDPR-ready · No credit card required