Most businesses do not have a lead generation problem. They have a lead handling problem — enquiries arrive and then sit. This is the blueprint we actually build: five stages, what each one does, and the arithmetic that decides whether automating it is worth the money.
The five stages
| Stage | What it does | What happens without it |
|---|---|---|
| 1. Capture | Every enquiry lands in one place, whatever channel it came from | Leads live in an inbox, a phone and someone's notebook |
| 2. Enrich | Adds the context you would otherwise google | Reps research instead of selling |
| 3. Score | Ranks by fit and intent against your rules | Everything is treated as equally urgent, so nothing is |
| 4. Respond | Replies and books within minutes, automatically | The fastest competitor wins the deal |
| 5. Follow up | Chases on a schedule until there is an answer | 80% of the pipeline quietly evaporates |
Stage 1 — Capture everything in one place
The first job is unglamorous: one destination for every enquiry, regardless of whether it came from a form, a phone call, an inbox, a DM or a marketplace. Until that exists, every downstream stage is guesswork because your data has holes in it.
The channel people forget is the phone. A missed call is a lead that never enters the system at all — it leaves no row to follow up. That is why an AI receptionist is often the highest-yield first build for local and service businesses: it converts an invisible loss into a captured record.
Stage 2 — Enrich before a human ever looks
By the time anyone opens the record it should already know the company size, industry, location, likely budget band and where the lead came from. This is pure grunt work — exactly the kind of lookup-and-append job that automation does perfectly and people do slowly and inconsistently.
Keep it honest: enrich with what you can actually verify. A confidently wrong company size will send a good lead down the wrong track, which is worse than no enrichment at all.
Stage 3 — Score against your rules, not a vibe
- Fit — do they match who you actually serve? Industry, size, geography, the things that make delivery work.
- Intent — did they ask for pricing, or download a guide? Those are not the same lead.
- Urgency — is there a deadline, a storm, a renewal, a broken process bleeding money right now?
- Disqualifiers — say them out loud. A pipeline that never rejects anything is a to-do list, not a funnel.
Write the scoring rules with the person who actually closes deals, then encode them. The value is not the model; it is that the rules become explicit, consistent and reviewable rather than living in one rep's instinct.
Stage 4 — Respond in minutes, because the maths is brutal
This is the stage that decides the return on the whole build. Published lead-response research is unusually consistent, and it is the same benchmark set our pricing meter and audit run on:
| You respond in | Qualified leads lost to a faster competitor |
|---|---|
| Under 5 minutes | ~5% |
| Under 1 hour | ~35% |
| Within 4 hours | ~78% |
| Same day | ~85% |
| Next day | ~94% |
Work an example. Say you get 40 enquiries a month, a customer is worth $2,000, and about 15% of the leads you actually reach turn into business. Reply next day and roughly 94% are gone before you start — you are competing for the remainder. Reply inside five minutes and almost all of them are still winnable. The gap between those two worlds is not a better sales pitch; it is a machine that answers while you are asleep.
Run your own numbers in the ROI calculator rather than trusting ours — it uses exactly these benchmarks and shows the payback against a real build price.
Stage 5 — Follow up until there is an answer
Most revenue lost after contact is lost to silence, not rejection. The follow-up sequence is the least clever and most profitable part of the pipeline: a defined cadence across email, SMS and voice, escalating to a human at the right moment, stopping instantly when they reply or ask you to.
Two rules keep this from becoming spam. Stop the moment they respond — nothing destroys trust faster than a nurture email arriving after a real conversation started. And cap it. "Forever" is not a cadence, it is harassment with a scheduler.
Where these pipelines break
- Nobody owns the exceptions. The automation handles 85% and the remainder silently piles up in a queue no one reads.
- Scoring never gets revisited. Rules written in January are wrong by June, and a stale score is worse than none because people trust it.
- The CRM becomes two sources of truth. If reps keep a private spreadsheet, the pipeline has already failed.
- Volume without qualification. Doubling enquiries while conversion stays flat just means you now waste twice as much time. Fix stage 3 before scaling stage 1.
What to build first
Not all five. Build stage 4 first — instant response — because it produces a measurable result within a fortnight and funds everything after it. Then stage 1 for whichever channel currently leaks most, usually the phone. Then follow-up. Enrichment and scoring come last: they make a working pipeline more efficient, but they cannot rescue one that answers a day late.
If you are not sure which stage is costing you most, the free 90-second audit will tell you, and it will say so plainly if the honest answer is that your volume does not yet justify building any of it.
