Multi-location brands · the ask that misses the moment

Your best customer just left a five-star review. Your referral ask will arrive in three weeks anyway.

Advocacy software usually asks for a referral on a fixed schedule — thirty days after purchase, a quarterly campaign. The best moment is a signal, not a date: a five-star review just posted, a job that just closed cleanly, praise given by name.

The desire, and the ask that arrives on the wrong day

Nobody searching “customer advocacy software” wants a campaign tool. They want the happiest customers asked at the moment they are happiest, not on a date set when the account was created.

1. 1. The schedule — a fixed number of days. Every advocacy platform does this well; it is the well-known half, and it produces a consistent volume of asks regardless of how any individual customer currently feels.
2. 2. The signal — already captured elsewhere. A five-star review, a cleanly closed job, praise in a transcript. Reputation and field-service platforms already have these events; the advocacy tool does not read them.
3. 3. The close — ask on the signal, not the calendar. The referral ask fires within a day of the positive signal, while it is fresh, instead of waiting for a scheduled date that may land in the middle of an unrelated complaint.

This is the same finding /dispatch-software made about review requests, aimed at the ask this entire system exists to generate.

Which platforms close the gap

• Reputation management platforms, by scale — where the five-star-review signal already lives.
• Dispatch software: the event your reviews never see — the same job-closed timing, applied to a review request first.
• Reputation management software: one signal, two AI jobs — where the threshold logic behind all of this was first built.

Some links on the recommendation page are partner links: the vendor pays us if you sign up, your price does not change, and each button says which.

How to build it yourself, end to end

Five steps. Each names the vendor touchpoint from the picks above.

1. 1. Define the positive-signal events precisely. A four-or-five-star review, a job-closed event with no complaint inside a set window, explicit praise in a transcript — a short, specific list, not every positive-sounding interaction.
2. 2. Fire the ask within a day of the signal. Satisfaction that is a week old is not the same as satisfaction that just happened; the window matters as much as the trigger.
3. 3. Suppress the ask if a complaint follows the signal. A five-star review followed by an unrelated support ticket two days later should hold the ask, not send it into an already-souring relationship.
4. 4. Cap how often any one customer is asked. Multiple positive signals in a short span should not produce multiple asks — one per customer per reasonable window, regardless of how many triggers fired.
5. 5. Track referral rate against the old scheduled baseline. The scheduled approach is the number to beat; if the signal-triggered version does not outperform it, the extra complexity is not earning its keep.

Step 3 is the one worth building first — an ask that fires into a souring relationship is worse than no ask at all.

Or have it built and run

Three founding slots: a review-response agent built at a founding rate, per-location baseline included, in exchange for the right to publish the measured before-and-after.