Kinetric takes one decision your team makes by habit, rebuilds it as a score on your own history, and tests it against what you do today before anything in the list changes. Three depths: a week to find it, four weeks to test it, and a run to keep it tested.
Somewhere in your operation sits a number somebody set years ago. How deep the mailing goes. Which accounts get a call. Where the reorder point sits. Which renewals get a human. It feels like policy, and your own history already knows whether it holds up.
You can stop after any one and still have something whole.
About a week. We find the rules in your operation that decide something per record and put a money figure on each.
About four weeks. One rule rebuilt as a score on your history, judged once on records it never saw.
Per run, optional, after a Scored Decision. The decision stays tested instead of turning back into a habit.
On the public mailing sample, that was $6,366 more net on one drop than mailing everyone. It was also 28% to 66% ahead of every simpler cut a planner would have run.
The honest condition: it pays only if the decision actually changes. If the list pull ignores the score, the return is zero. Sometimes the proof says your current rule is fine, and then you have paid for a proof, not a change.
| Depth | Duration | Asks of your team | Fee |
|---|---|---|---|
| Rules Review | About a week | One conversation with the person who owns the operation, and read access to the history | A fixed fee for about a week |
| Scored Decision | About four weeks, in parallel with your desk. Nothing in the operation changes until the last week. | One owner, one call, one export, one read-only seat | Fixed once the decision and the data are named, at the scope call |
| Decision Run | Per run, or monthly for decisions that run often | The results file after each run | Fixed at the Scored Decision handover, once the cadence and the data are named. Decline it and the score still runs. |
This is not a new system and not a black box. Every input has a stated reason on record.
The expectation goes on record before your data is opened. The score is built on part of your history, frozen, and judged once on the rest. Every simpler way is tried first, and your current rule stays in every comparison. The limits are written down before anyone asks.
The one honest limit: the published proof is a public dataset, and no client engagement has run under the Kinetric name yet. We say that on this page and in the room.
A lapsed-donor file of 96,367 names, mailed in full by habit at $0.68 a piece. The score mailed 53.3% of the file and netted $16,926 against $10,560 for mailing everyone.
Read the case›One sentence: the decision, the rule that sets it today, and who owns it. If you can't name the rule, the Rules Review starts there.
A short call to fix the baseline and agree what counts as winning, before the data is opened. The fee is fixed on this call.
An export of past runs with their outcomes, and a read-only seat on the system the list comes from. One session with your admin.
We'll tell you what your own history can say about it, and which offering fits. If none does, we'll say that too.
Start a conversation