The Personalized Lead Nurturing Stack
A no-code build that qualifies every inbound lead, sorts them in your CRM, and generates a personalized report plus a 7-day email follow-up sequence.
Section One
What You're Building
Most intake forms do one thing: dump every submission into the CRM, identical. The lead who wants to start this week gets treated exactly like the tire-kicker who is "just exploring." Your attention spreads evenly across people worth wildly different amounts, and the good leads go cold while you catch up on manual follow-up.
That's the potential-lead trap: leads get captured but never moved forward, because qualifying them and following up by hand takes time you don't have. This system replaces "everyone into the CRM the same" with a brain that reads each lead, decides who's a fit, and does the follow-up work for you.
Fig. 1 — system flow
Capture
Webhook
Form submission arrives, payload cleaned
Qualify
AI classifier
Reads the lead, sorts qualified vs unqualified
Save
CRM
Every lead saved — nothing thrown away
Nurture
Report + 7-day emails
Qualified leads get a PDF report and email sequence
The rule that makes this work
Saving every lead, qualified or not, is the design, not an afterthought. An unqualified lead today can be a perfect fit next quarter, and filtering by status costs nothing later. Nothing captured here ever gets thrown away.
Section Two
What You Need
One workflow, six moving pieces — and every piece but n8n itself is swappable for whatever you already run.
| Piece | Role in the Build | Swappable For |
|---|---|---|
| Form / capture tool | Collects the lead's answers, sends them to a webhook | GoHighLevel, Typeform, ClickUp — any tool that can POST to a webhook |
| n8n | The brain: cleans data, routes, orchestrates every step | Self-hosted or cloud |
| AI model | Qualifies leads and writes the report and emails | Any capable chat model — standard, not reasoning |
| CRM / database | Stores every lead with a status and the generated assets | Airtable, n8n Data Tables, a spreadsheet |
| HTML-to-PDF service | Turns the report into a downloadable PDF | PDF.co or similar |
| Docs tool | Holds the email sequence as a shareable doc | Google Docs |
Qualified leads get the full treatment. Unqualified leads still get saved, never discarded — a "not right now" can become a "yes" in three months.
Section Three
Build Steps
01
Step One
Capture the lead
Add a Webhook node (method: POST) as the trigger. In your form tool, add an automation that fires on submission and sends the data to the webhook URL. Any capture tool works as long as it can POST to a listening webhook.
02
Step Two
Clean the payload
Use two Edit Fields (Set) nodes back to back: the first pulls only the fields you need out of the noisy payload, the second strips array brackets and stray quotes so ["ASAP"] becomes ASAP. Set it up once — every submission after that comes through clean.
Why this matters
Form platforms rarely send clean data — expect hidden line breaks in field keys and multi-select answers wrapped in arrays. Skip this step and every downstream node reads blanks.
03
Step Three
Qualify the lead with AI
Add a Text Classifier node with two categories, qualified and unqualified, judging against criteria you define — urgency, a real business, a process-based challenge. Set a fallback branch so leads the AI can't cleanly judge still land somewhere instead of vanishing.
Gotcha that will cost you an hour
Give the classifier a standard chat model, not a reasoning model. Reasoning models return output the classifier's parser can't read, and you get "Model output doesn't fit required format."
04
Step Four
Save to the CRM
Add two CRM create steps, one per branch: qualified leads get a Status of Qualified, unqualified leads get Unqualified.
Why this matters
Never throw a lead away. Saving costs one extra node, and an unqualified lead today may be a perfect fit next quarter. Filter by status to work each group differently.
05
Step Five
Generate the report outline
On the qualified branch, add an AI message node. Feed it the cleaned lead data and ask for a structured JSON outline: executive summary, current state, opportunity analysis, roadmap, recommendations, next steps. Point it at the cleaned questionnaire data, not the CRM record, so the report is built from the lead's actual answers.
Why two steps instead of one
Outline first, then expand. It produces a far more coherent report than asking for the whole thing in one shot — the outline locks the structure, the expansion fills it with depth.
06
Step Six
Expand the outline into a full report
Add a second AI message node. Feed it the Step 5 outline and ask it to expand each section into a complete report, addressed to the lead by name, tied to their specific tools and challenges. End the prompt with hard output rules: state today's date, no page-break markers, valid HTML only — starting with <html> and ending with </html>, no code fences.
Gotcha
Whatever field your AI node returns text in, reference it correctly downstream. Some nodes nest the text several levels deep — if a step gets an empty input, trace the previous node's output shape.
07
Step Seven
Convert the report to a PDF
Add an HTTP Request node pointed at your HTML-to-PDF service. Send the report HTML in the body and your API key in the header. You get back a PDF URL.
Good to know
Free tiers often expire the PDF link after about an hour. If you need it to last, save the PDF file itself into the CRM as an attachment, not just the link.
08
Step Eight
Write the report back to the CRM
Add a CRM update step. Match on the record you created in Step 4 and fill in the report PDF link, the PDF as an attachment, and the outline for your own reference.
09
Step Nine
Generate the 7-day email sequence
Branch off the report node into a third AI message node. Feed it the full report and ask for a 7-day sequence, one email per day, each hitting a different angle: overview, quick wins, the core challenge, tech stack, ROI, implementation, long-term vision. Every email's call to action is to book a call. Basing the emails on the report, not just the raw questions, keeps them specific.
10
Step Ten
Save the sequence and link it
Create a Google Doc, insert the generated emails, then update the CRM record with the doc link. Every qualified lead's record now holds their report, their PDF, and a ready-to-send follow-up sequence — all from one form fill.
Section Four
The Three Prompts
These are the reusable core of the build. Adapt the voice and criteria to your business.
Outline generator (system)
Act as an expert consultant. Receive a lead's questionnaire answers as JSON. Produce a personalized report outline as structured JSON with sections for executive summary, current state assessment, opportunity analysis, implementation roadmap, recommendations, and next steps. Draw every section from their actual answers. Be specific, never generic. Output only valid JSON.
Report generator (system)
Receive a report outline as JSON. Expand it into a complete, professional, personalized report addressed to the client by name. Tie every recommendation to their stated tools, goals, and challenges. End with the output rules: today's date, no page breaks, valid HTML only starting with <html> and ending with </html>, no code fences.
Email generator (system)
Receive the client's personalized report. Write a 7-day email follow-up sequence, one email per day, each targeting a different angle from the report. Personalize with the client's name, reference specific findings, and make booking a call the primary call to action in every email. Keep it professional but conversational.
Section Five
The Upgrade Path: Real Lead Scoring
The build above sorts leads into qualified and unqualified. The next level is a genuine score. Upgrade the classifier to output three things instead of a yes or no: a score (0 to 100), a tier (Hot, Warm, Cold), and a one-line reason so you can trust and debug the call.
Write all three to the CRM, then branch the nurture on tier: hot leads get a fast, direct "book now" track; warm leads get education first; cold leads sit in the CRM for a later check-in instead of burning a full email sequence.
The difference it makes
A yes/no classifier gives you a sorted list. A score, a tier, and a reason give you a scored pipeline — one you can route, prioritize, and trust without opening every record to check.
Section Six
The Result
One form fill now produces a qualified, sorted lead, a personalized report, and a week of follow-up — without you touching a thing. The good leads move toward a booked call while they're still warm. The rest are saved, tagged, and waiting. No more potential-lead trap.
“A yes/no classifier gives you a sorted list. A score, a tier, and a reason give you a scored pipeline.”
— The Personalized Lead Nurturing Stack, p. 8
§ Fin.
Rather have it built for you?
This guide is one system from the Soulwire Studio workflow library. We build AI automations for businesses that run on process: intake, scheduling, proposals, compliance-heavy operations. If you'd rather spend your time closing deals than debugging lead gen, book an audit and we'll find the automation hiding in your business.
