Inside CanvassIQ: how we map turf and track every conversation

CanvassIQ is in private beta with a small number of UK estate agencies, and we get asked roughly the same three questions every time we show it to a new branch manager. How does it actually divide up the patch? How does it remember conversations without making negotiators type for ten minutes after every door? What does the Monday morning look like for the manager? Here's the honest answer to all three, including the bits the product doesn't do yet.

Turf mapping: drawing zones that match how the branch already works

A branch's patch is almost never a neat rectangle. It's the bits of three postcodes the senior negotiator has worked for twelve years, with a couple of streets in a neighbouring postcode the manager doesn't want to give up to a rival agency. Software that tries to impose tidy postcode-shaped zones gets ignored within a fortnight.

CanvassIQ starts with the map and lets the manager draw. You sketch a zone — a polygon, a street, a cluster of streets — and give it a name. "North side of the park", "Old Hill terraces", "the new estate". Then you assign one or more negotiators to that zone. A negotiator can own a zone exclusively or share it; an address can sit in two zones if the branch genuinely covers it two ways.

Inside each zone, every individual address gets a row. The row carries:

• Property type (from the title register / EPC data where available)
• Last sale date and price
• HPI-adjusted current estimate
• Planning history flags
• Any conversation log entries the branch has ever made against that address
• A live priority score that updates as new signals come in

When a manager hands a negotiator their patch for the morning, what they're really doing is asking the system to print (or push to phone) the top-N highest-scoring addresses in that negotiator's zone, with the reason each one is high-priority shown next to it. The negotiator doesn't see "26 Acacia Avenue, priority 8". They see "26 Acacia Avenue — planning granted for loft 14 days ago, neighbour at 28 sold for £420k last month, previous conversation Aug 2024 noted family considering schools".

Conversation log: 20 seconds, not 10 minutes

The single most important design constraint we set ourselves was that logging a conversation could not take more than 20 seconds from the negotiator's phone. If it takes longer, it doesn't happen, and if it doesn't happen the whole tool collapses into an expensive map.

The phone log captures, in this order, with the minimum number of taps:

1. Outcome chip (no answer / brief chat / real conversation / valuation booked / not interested / do not contact). One tap.
2. Voice note up to 30 seconds. The negotiator just talks: "Mrs Patel, daughter starting at the grammar in September, husband works from home so they want a bigger kitchen, said come back in October." Auto-transcribed.
3. Next action and date. Single field, defaults to "+2 weeks" if the outcome was a real conversation, "+3 months" if it was a soft no.

That's the lot. The voice transcript is parsed in the background to extract anything that looks like a date, a name, a milestone, or a personal detail, and those become surfaced in the next interaction. When the negotiator goes back in October, the address card opens with "Mrs Patel — daughter at grammar Sept, kitchen, husband WFH" right at the top.

The conversation log isn't a CRM field set. It's a memory aid that fires at the exact moment the negotiator needs to remember something.

This is the bit we're proudest of and the bit that has taken the longest to get right. Negotiators don't fill in forms. They will, however, talk to their phone for 20 seconds in a doorway.

The Monday morning manager view

A branch manager opens CanvassIQ on a Monday at around 8.30am, before the morning meeting. The view they get is designed to answer four questions in under two minutes:

1. What follow-ups are due this week, by negotiator? Every address with a next-action date falling in the next seven days, grouped by who owns it. The manager can reassign, defer, or escalate.
2. What new signal has fired since Friday afternoon? New planning grants, new completed sales, new probate filings on the patch. Each one is automatically scored and added to the relevant zone's priority queue.
3. Where is the pipeline? Every address currently flagged as "warm but not yet instructed" — homeowners who've had a real conversation in the last six months and are flagged for return. This is the asset the branch should be most protective of, and it's usually invisible without a tool.
4. Where are the gaps? Which zones haven't been canvassed in over four weeks. Which negotiators haven't logged a real conversation recently. The unflattering numbers, deliberately surfaced.

That's the view. No dashboards full of vanity metrics, no charts of door counts. The four questions a manager actually needs to answer before the team meeting.

What it doesn't do yet

We're in private beta and we'd rather be honest than aspirational. Things CanvassIQ does not do today:

• Integrate directly with the major CRMs. It exports CSV cleanly and we have an API in preview, but a one-click Reapit or Alto sync isn't there yet. Coming.
• Run automated email or SMS follow-ups. The system tells you what to send and when; it doesn't send it for you. Deliberate for now — we want the negotiator's voice on the follow-up, not a templated nudge — but optional automation is on the roadmap.
• Cover lettings. The product is sales-only at the moment. Lettings canvassing has a different signal stack and we'd rather build it properly than half-do it. Our sister product, Landlord Lead Generator, covers some of that ground separately.
• Support multi-branch corporate hierarchies in full. It works for a single branch beautifully and a small group of branches well. Genuine multi-region corporate setups are next.

If any of those are dealbreakers for your branch, we'd rather you know now than three weeks in.

Where to look next

We're keeping the beta deliberately small while we get the conversation log behaviour right with real negotiators on real patches. If you'd like to see how CanvassIQ would work in your branch, the CanvassIQ overview page explains the product end-to-end and is the right place to start a conversation.