What your RealGreen cancel reasons are trying to tell you

Your cancel reasons are the cheapest market research you'll ever get. And most RealGreen databases waste them, with half the cancels coded "other" or left blank. Coded properly, cancels tell you whether you have a price problem, a results problem, or a communication problem, which ones you can save, and which lead sources are quietly selling you customers who leave. Here's how to read them.

Why does coding cancels properly matter?

Because an uncoded cancel is an unfixable cancel. If your office logs "customer requested cancel" for everything, you literally cannot tell the difference between a pricing problem and a weed problem, so you'll guess, and you'll probably fix the wrong thing. The discipline is boring but simple: a short list of 8–10 specific reason codes in RealGreen, "other" requires a typed note, and the office reviews the week's cancels every Monday. That last part is what keeps the codes honest; when nobody reads them, everything drifts back to "other." (Cancels-plus-reasons is one of the twelve numbers in our weekly KPI checklist for exactly this reason.)

What are the top cancel-reason patterns?

Every company's mix is different, but after enough RealGreen databases the same patterns repeat. Here's the honest read on each, including which are actually saveable and which aren't worth a discount:

Cancel reason	What it usually means	Saveable?	The play
Price / "too expensive"	Value wasn't demonstrated, or a competitor's postcard hit at renewal	Often	Downgrade offer (fewer rounds), not a discount on the same program
Moving	Exactly what it says, but it's two opportunities	Partly	Quote the new address; market to the new owner of the old one
Results / quality	Weeds still there, or expectations were never set	Yes	Service recovery: free re-treat + a manager call within 48 hours
Doing it myself	Often price in disguise; sometimes genuine	Sometimes	Hybrid offer: keep the hard rounds (pre-emergent, grubs), drop the rest
Unresolved complaint	A service call that never got closed out	Most saveable	Fix the complaint first, then ask to keep the program
Financial hardship	Real budget change	Rarely	Pause, don't push: a graceful exit earns the win-back later
One-and-done	They wanted one aeration, not a program; a sales/entry problem	No	Fix intake: stop entering one-time jobs as programs

The composition matters more than the headline churn number. Annual churn for residential programs is commonly cited somewhere in the 15–25% range, but 18% that's mostly moves is a healthy company in a mobile market, while 18% that's mostly results complaints is a production problem wearing a marketing costume.

Which cancels should you actually fight for?

Fight hardest for unresolved complaints and results cancels. These customers wanted it to work, and a fast, human service recovery saves a meaningful share of them. Fight smart on price: a downgrade to a lighter program keeps the customer and the route stop; a panic discount on the same program teaches customers that threatening to cancel is how you get 15% off. Let hardship and one-and-done go gracefully: a warm goodbye converts to win-backs later, and chasing them with offers just burns goodwill. Keep the worked math in view: saving a $550 program customer costs you a re-treat and a phone call; replacing them costs a new-customer acquisition, commonly $80–$250 in marketing cost plus the sales time, and the replacement starts over at day zero of early-churn risk.

How do you feed cancel data back into marketing?

Three loops, all running off the cancel record in RealGreen:

• The save sequence. Cancel entered → an automated, reason-matched response within hours: service-recovery outreach for quality codes, the downgrade offer for price codes. This is a standard entry in our RealGreen automation catalog, and it works because it's fast: a save attempt three weeks later is a courtesy call, not a save.
• The win-back campaign. Six to twelve months after a saveable-code cancel, an automated "we've missed you" offer. Spring pre-emergent timing works especially well. Win-back lists are pure gold because these people already bought from you once; even modest response rates beat cold acquisition. (This pairs naturally with the retention automations that reduce cancels in the first place.)
• Lead-source truth. Join cancels to original lead source and tenure. If customers from one source churn at twice the rate of your referrals, that source's real cost per kept customer is double what the ad platform reports. That should change how you spend, and it's the kind of thing you only see with reporting that connects marketing data to your customer database.

And back into ops?

Marketing loops get the attention, but the ops loop is where cancels actually shrink. Results-code cancels mapped by technician and route reveal training problems; price-code spikes right after a rate increase tell you the increase letter did a poor job selling value (see raising prices without losing customers); a cluster of unresolved-complaint cancels means service calls are falling into a hole between the office and the field. One pattern per month, one fix per month. That's the realistic pace, and it compounds.

How do you keep the codes honest over time?

Code quality decays without maintenance. The trend is always back toward "other." Three habits hold the line. First, make the cancel entry a conversation, not a checkbox: whoever takes the cancel call asks one open question ("what changed?") before picking a code, and the honest answer usually doesn't match the customer's opening line: "too expensive" often turns out to mean "I saw weeds in July and nobody called me back." Second, audit a sample monthly: pull ten coded cancels, read the account notes, and check whether the code matches the story. If three of ten don't, retrain. Third, keep the code list short. Every code you add past ten makes the office's choice slower and sloppier; when someone proposes a new code, retire an unused one. The goal isn't taxonomic perfection: it's a dataset trustworthy enough to bet marketing budget on.

The takeaway: enforce 8–10 real cancel codes, read them weekly, fight for the complaint and results cancels, downgrade the price cancels, release the rest gracefully, and feed the codes into save sequences, win-backs, and lead-source math.

Your cancel-data checklist

1. Rebuild your RealGreen cancel-code list: 8–10 specific reasons, "other" requires a note.
2. Retrain the office: every cancel gets a real code, reviewed Mondays.
3. Stand up the save sequence: reason-matched response within hours of cancel entry.
4. Build the 6–12 month win-back campaign for saveable codes.
5. Run cancels against lead source and tenure quarterly: reprice your sources accordingly.
6. Send one ops fix per month upstream: technician, route, or process.