AI Lead Intelligence

What it does

Every submission captured by LeadToSheet is run through AI before it lands in your Google Sheet. The model scores the lead, writes a one-sentence summary, gauges how urgent it is, and enriches it with company and person context — then writes all of that into the sheet alongside the submission.

• Lead Score — Hot, Warm, or Cold based on intent, urgency, email type, and message content
• Lead Score Reason — a short phrase explaining the score (e.g. "pricing intent + business email")
• Lead Summary — one sentence (≤ 25 words) describing what the lead wants
• Urgency — High, Medium, or Low, based on deadline and timing language in the message
• Company enrichment — Company, Industry, Company Size, and Country, inferred from the lead's email domain and a snapshot of their company homepage (business emails only)
• Person enrichment — Job Title and LinkedIn profile, found by searching the web for the lead's name alongside their company
• Match Confidence — High, Medium, Low, or None, telling you how sure the model is about the person it matched

How scoring works

Scores are computed from a combination of signals that we send to the model along with the submission:

• Form field content — what the lead typed in the message and other free-text fields
• UTM source and campaign — how they got to your site
• Email domain — business vs. generic (gmail/outlook/yahoo)
• Page context — page URL and title where the form was submitted
• Urgency keywords — "ASAP", "this week", deadlines, "urgent"

Score meanings

• 🔥 Hot — explicit purchase or demo intent, business email, named company, or clear urgency
• ✨ Warm — real interest but generic email, vague timing, or only partial context
• 🧊 Cold — low signal: generic email + vague text, mismatched UTM, or filler content

How enrichment works

Alongside the score, the model enriches each lead in two ways — company context and person context.

Company enrichment

When a lead uses a business email address, LeadToSheet takes a snapshot of that company's homepage and uses it — together with the email domain — to infer the Company, Industry, Company Size, and Country. Leads from generic providers (gmail.com, outlook.com, yahoo.com, and the like) skip this step and leave the company columns blank.

Person enrichment

The model searches the web for the lead's name alongside their company or domain to find their Job Title and LinkedIn profile. It only fills these in when a result clearly ties back to the lead's company, email, or country — it never guesses. Because this is a web search rather than a domain lookup, it can find a profile even for a lead using a personal email address.

Every person match carries a Match Confidence — High, Medium, Low, or None. A confidence of None means the model couldn't find a profile it was sure about, so the Job Title and LinkedIn columns are left blank rather than filled with a guess.

Where you'll see it

• Google Sheet — Lead Score, Lead Score Reason, Lead Summary, Match Confidence, Company, Job Title, LinkedIn, Industry, Company Size, and Country columns are added automatically (the Lead Quality dropdown at column A stays human-edited — AI never writes to it)
• Leads page — filter by Hot / Warm / Cold, see the AI summary plus urgency and match-confidence badges inline, and view all enrichment in the lead detail panel
• Dashboard — "Hot leads this week" and "Warm leads this week" widgets at a glance
• Analytics — score distribution chart shows your lead-quality breakdown over the last 7/30/90 days
• Real-time email alert — the alert subject leads with the score ("🔥 Hot lead from Acme Corp") and the body includes the summary and reason

Privacy and disabling AI

AI processing runs server-side and the submission's field content is not retained by the model after scoring. Outputs are stored on the submission record so you have a permanent audit trail. Note that person enrichment works by running web searches for the lead's name and company — if that isn't appropriate for a particular form, turn AI off for it (see below).

Disable for the whole workspace

1. 1

   Open Settings → AI & Quality

   From any site in your workspace, go to Settings and pick the AI & Quality tab (also reachable from the sidebar under Pipeline → AI & Quality).

2. 2

   Toggle workspace AI processing off

   All future submissions will skip AI scoring and the AI columns will be left blank in your sheet.

Disable for a specific form

On the same Settings → AI & Quality page, every form has its own toggle. Use this for sensitive intake forms — medical, HR, legal — where you don't want any field content sent to a model.

Performance and reliability

• Speed — AI runs in the background right after a lead is captured, so it never slows down the form for your visitor. A full pass — including the web lookups for person enrichment — usually takes a few seconds; the lead appears in your dashboard immediately and the AI columns fill in shortly after.
• If something goes wrong — when the AI call errors or times out, the submission still syncs to your sheet with the AI columns left blank. AI is never allowed to block or delay a lead.
• One attempt per lead — the AI step runs once per submission and isn't retried; the Google Sheet sync continues regardless.
• Audit log — every AI call (and every skip / failure) is recorded in your workspace audit trail.

Related