Walk through the whole loop on one fictional business — from the moment an agent spots it in the market to the moment its owner reads a free, accurate read of their own reputation. Every outbound agent is simply an inbound magnet run in reverse.
On a schedule, the reputation_finder agent scans local markets, pulls each business's real Google footprint, and flags the ones whose public reputation underperforms their neighbours. No guessing — it scores on verified numbers.
It scans, it doesn't spam. The agent works a market — here, plumbers around Maple Grove, MN — and benchmarks each listing against its local peers.
One business stands out. Northgate's rating and review volume both trail the pack. That's a real, fixable gap — and a real reason to reach out.
It scores and tags. The lead is written with a lead_score, a source_agent, the matching magnet, and the verified pain signal.
The agent upserts the prospect into HubSpot with structured metadata. Discovery never drafts the email; it stores the routing fields the shared sender will read at send time.
The pipeline fills itself. Every agent writes to the same contract, so one shared sender can serve all six magnets without forking logic.
The fields are the routing. source_agent picks the opening angle; recommended_lead_magnet picks which engine composes and, later, runs.
Nothing is sent yet. The record sits at not_contacted until the sender is good and ready — and only if the signal still checks out.
Before anything goes out, the sender re-fetches the site and re-runs the analysis fresh — sites change. Only if the gap still holds does it compose, pass a brand-voice check, wrap it for compliance, and send via Postmark.
Keyed to the angle. Because source_agent is reputation_finder, the email opens on the reputation gap — not a generic pitch.
Re-verified at send. The stored signal only guides; the numbers in the email are regenerated and re-checked. If the gap closed, the send is skipped — never a stale claim.
Compliant by construction. Sender identity, a physical mailing address, and a one-click unsubscribe ride on every message.
Hi there,
I was looking at plumbers around Maple Grove and pulled up Northgate. You're sitting at about a 4.2 across roughly 30 reviews, while the shops you come up next to are closer to 4.7 on a few hundred each.
That half-star — and the thinner count — is the first thing someone sees when they're comparing a few names on their phone. It makes a well-run business look smaller than it is.
I put together a short, no-cost read on exactly where your Google presence sits against your neighbours and what that gap is quietly costing you. Want me to send it over? Just reply "yes" and it's yours.
— Chris
PracticAI
chris@practicai.ca
A reply (or a click) routes the contact straight into the inbound magnet named on the record. The same recommended_lead_magnet field that picked the angle now picks the engine.
This is the seam. Outbound and inbound meet here. The prospect raised their hand; the system already knows which read to produce.
No re-keying. Their business name, URL, and city are already on the record from discovery — the magnet runs with what it needs.
The Reputation Read engine fetches Northgate's homepage for context and pulls its live Google listing (Serper /places): the real star rating, the real review count, and a benchmark of nearby peers. The read is anchored entirely on these numbers.
Same engine, forward direction. This is the exact analysis the agent ran in reverse to find Northgate — now run on the one business that asked.
Verified numbers only. Star rating, review count, and the peer benchmark come straight from Google's local results. The engine never invents a review, a complaint, or a number it can't see.
One plain-text email, sent 1:1 from Chris. It names the gap and what it costs — in their real numbers — and stops there. That honesty is the proof-of-value that pulls them toward a conversation.
The payoff. They came in skeptical of a cold email and leave with something genuinely useful about their own business.
Accuracy is the moat. Every figure is real and verifiable. One wrong claim would end the relationship — so the engine never makes one.
Hi there,
I pulled up your Google listing and put it next to the other plumbers showing up around Maple Grove. Your rating sits at about 4.2 across roughly 30 reviews, while the shops you come up alongside average closer to 4.7 on a few hundred each. The rating gap is real, but the bigger story is the volume — you've got a fraction of the reviews your neighbours carry.
Here's why that matters more than it looks. When someone's choosing a plumber, they're usually comparing three or four names side by side on a phone. A 4.2 next to a row of 4.7s reads as "the weaker option" before they've read a single word — and 30 reviews next to a few hundred makes a fifteen-year business look newer and smaller than it is.
The cost isn't dramatic, it's quiet: you're the name that gets skipped in the exact moment someone is deciding, by people who'd have been thrilled with your work. The reputation you've actually earned and the reputation that shows up in Google are two different things — and the second one is the only one a new customer ever sees.
If it'd be useful, just reply — happy to spend 20 minutes talking through what closing that gap would actually look like for you. No pitch.
— Chris
PracticAI
chris@practicai.ca
There aren't two systems here — there's one. The same code that produced Northgate's free Reputation Read is the code that found Northgate in the first place. Reverse the input and the engine flips from "show this business its gap" to "find businesses with the gap."
Given a market, the agent benchmarks every listing and surfaces the ones that underperform — scored and ready for outreach.
Given one business that raised its hand, the engine runs the same analysis and delivers an accurate, personalized read.
You just followed the Reputation Read end to end. Every other magnet works identically — each is one diagnostic engine that runs forward as a free read and in reverse as a discovery agent. Every one of them is live: pick the read that fits your business and it's yours, free.
Everything you just watched runs for real. Pick whichever of the six reads fits your business — it's free, it's accurate, and it lands in your inbox the same way Northgate's did.
Get your own read — pick any of the sixOr skip straight to the conversation: chris@practicai.ca