Skip to content

Qualify and route every inbound lead in minutes, not hours

Starts when a form, chat tool or CRM sends a new lead. Ends with a scored lead, a tier, an owner, a response deadline and a reply draft.

Bring the step you are stuck on: 30 minutes, a straight answer and the next step. If you would rather have it built, a reply within 5 working days.

The route

Sales inbox

6 steps: 3 rule steps, 1 AI step and no person step

Copy, share, discuss

Copy

Available once this workflow passes its checks.

Share

todkett.com/automations/sales/inbound-lead-qualification/

Discuss

Opens when the first workflow passes its checks. No GitHub account? Ask about this by form.

How the library, the tracks and the talk work: the community page.

At a glance

Starts when
A form, chat tool or CRM posts new leads to one webhook.
Input
New leads arrive at the webhook, up to 50 per call.
Output
A scored lead, a tier, an owner, a response deadline and a reply draft.
AI
Uses AIOne step reads the lead's message and website text and returns fit and intent scores, a company size band and a reply draft.
Tools
Webhook from any form or CRM, Public company websites, OpenRouter, Google Sheets
Last stress test
No recorded run yet

01Steps

The steps

6 steps: 3 rule steps, 1 AI step and no person step. Leads arrive while you are on a call and wait until tonight. Someone has to check the company, decide if it is a real buyer and pick an owner, so hot leads go cold and spam eats time.

  1. 01Trigger

    Lead arrives

    New leads arrive at the webhook, up to 50 per call.

  2. 02Rule step

    Tidy and cap the fields

    Tidy the fields and cap each message at 3,000 characters.

  3. 03Rule step

    Read the company homepage

    Read the company homepage title and description; free email domains are skipped.

  4. 04AI step

    AI scores and drafts a reply

    AI scores fit and buying intent and drafts a first reply.

  5. 05Rule step

    Score, tier, owner, deadline

    Code computes the score and sets the tier, the owner by region and segment, and a response deadline.

  6. 06Output

    Routed lead and reply draft

    Routed leads with their reply drafts.

    Sales inbox

02Where AI is used

Where AI is used, and what stays with you

AI reads, judges or drafts. Fixed rules check, decide and route. A person approves anything that goes out.

AI step

Step 04

One step reads the lead's message and website text and returns fit and intent scores, a company size band and a reply draft.

Buying intent in free text is hard to judge with rules. The model only scores; code decides the tier, the owner and the deadline.

What the model reads

  • The lead's message, capped at 3,000 characters and passed as data, not instructions.
  • The company homepage title and description, read before the AI step.

What it never sees or decides

  • The score, the tier, the owner or the deadline: code sets all four.
  • A guess: an unreadable answer sends the lead to needs review.
  • The send button: the reply stays a draft for a person.

Model calls: 1 model call per lead.

Fixed rules

They check, decide and route. 3 rule steps.

  • 02Tidy and cap the fields
  • 03Read the company homepage
  • 05Score, tier, owner, deadline

A person

No step waits for a person. Human sign-off (ST7) is not needed: returns a result only: writes and sends nothing.

03Checks

Checks around the AI

What runs before the model, and what checks its answer after, in plain words. Build these into your version too.

  1. 01At most 50 leads per request; each message capped at 3,000 characters.
  2. 02The homepage is read before AI, so judgement rests on facts; failed loads are recorded, never invented.
  3. 03Free email domains are detected and the score lowered by 10 points.
  4. 04Score is 55% fit plus 45% intent, computed in code; tier cut-offs at 70 and 45.
  5. 05Spam, vendor pitches and job seekers are disqualified by rule.
  6. 06An unreadable AI answer goes to needs review instead of a guess.
  7. 07The lead text is passed to the model as data, not instructions.

04The 8 checks

Stress test

The 8 checks we run on every workflow, and you can run too. A check marked Not needed says why.

Status
Not yet0 of 6 that apply have a recorded pass
Last run
No recorded run yet
Model review
Still open

Checks marked Not yet are still open. One of them was seen in the workflow design on 23 Oct 2026; a check shows Passed only once its test run is recorded.

  1. ST1Test set

    Status: Not yet

    No recorded test yet.

  2. ST2Prompt attack

    Status: Not yet

    No recorded test yet.

  3. ST3Run twice

    Status: Not needed

    Returns a result only: writes and sends nothing

  4. ST4Failure recovery

    Status: Not yet

    No recorded test yet.

  5. ST5Load and limits

    Status: Not yet

    No recorded test yet.

  6. ST6Failure alert

    Status: Not yet

    No recorded test yet.

  7. ST7Human sign-off

    Status: Not needed

    Returns a result only: writes and sends nothing

  8. ST8Rules around the AI

    Status: Not yet

    Seen in the workflow design on 23 Oct 2026. The recorded test run is still open.

0 passed · 2 not needed · 6 not yet

What we test

We show what we tested, never the test data. The example below is made up.

05Test replay

How to test it

Before you trust a workflow, try to break it. The sample record runs the route step by step. Each break test below gets its own replay once its recorded run exists.

Record
LEAD-0142
Data
Sample run, made-up data
  1. 01Lead arrives

    In at the webhook from ops@fernway.example

    Done
  2. 02Tidy and cap the fields

    Message within 3,000 characters

    Done
  3. 03Read the company homepage

    Homepage read: help desk software for clinics

    Done
  4. 04AI scores and drafts a reply

    Fit 80, intent 70; reply drafted

    Done
  5. 05Score, tier, owner, deadline

    Score 75.5 in code: tier A, owner and deadline set

    Done
  6. 06Routed lead and reply draft

    In the sales inbox; a person sends the reply

    Sales inbox
    Arrived

LEAD-0142 reached the last station: Sales inbox. In the sales inbox; a person sends the reply.

Sample run on made-up data. No stress-test run is recorded yet, so this plays the example, not a test. Nothing runs live.

The break tests for this workflow

  • ST3

    Same input sent twice

    Returns a result only: writes and sends nothing.

    Not needed
  • ST2

    Instructions hidden in the input

    No recorded run to replay yet.

    Not yet
  • ST4

    A connected service fails

    No recorded run to replay yet.

    Not yet
  • ST8

    The AI returns a broken answer

    Seen in the workflow design. No recorded run to replay yet.

    Not yet

06Example

The example, in and out

Made-up data, written to show the steps; no recorded run stands behind it yet. Fernway is a made-up small software company, and its address ends in .example, so it cannot belong to a real business.

InputExample

Example on made-up data: a lead from ops@fernway.example. Message: we answer about 300 support emails a week and want them sorted before the team starts. Can we talk this month? Homepage: help desk software for clinics.

OutputExample

Fit 80 and intent 70 from the AI step. Score 75.5, computed in code as 55% fit plus 45% intent: tier A. Owner set by region and segment, a response deadline set by the tier rule, and a reply draft waiting for a person.

07Limits

When not to automate this

  • If a few leads a week arrive, a person can qualify them faster than a build pays back.
  • It scores and routes. It does not send the reply: the draft waits for a person.

08What you need

What you need to run it

Each connected tool and the access it needs
ToolAccess
Your form, chat tool or CRMPost each new lead to one webhook
Public company websitesRead the homepage title and description
OpenRouterOne API key with a credit limit, for the scoring step
Google SheetsOptional: write a test log
Where it runs
In n8n, self-hosted or n8n cloud.
Model calls
1 model call per lead.
Running costs
Yours, paid to each service directly. Scope in writing first. A quote before any build. No prices on the site yet.

How to adapt it

Changes we have made or tested on this workflow.

  1. 01Swap the AI score for fixed scoring rules and an outreach tool's enrichment. That rule-only version is built but not live.

What changed

  1. ContentFirst version, written from the 23 Oct 2026 catalog.

09FAQ

Questions people ask about it

Does AI decide which leads matter?

No. AI returns fit and intent scores and a reply draft. Code computes the score as 55% fit plus 45% intent, sets the tier at cut-offs of 70 and 45, and picks the owner and the deadline.

What happens to spam and job seekers?

Fixed rules disqualify spam, vendor pitches and job seekers, so they never reach a sales rep's list.

What if the model returns something unreadable?

The lead goes to needs review instead of a guess, and a person decides.

Does it send the first reply?

No. It drafts the reply. A person reads it and sends it.

10Related

Stuck on your version of this workflow?

Bring the step that does not work. In 30 minutes you get a straight answer and the next step. If you would rather have it built, tell us your tools and the rules you already follow. Scope in writing first. A quote before any build. No prices on the site yet.