Kinetric brings AI to the work your team does with numbers and keeps the call with you. The check before a report goes out, the decision set by habit, the Monday readout, the workflow run by hand: each done faster with AI, tested on your own history, and handed back with the proof. Augmentation, not automation.
The shape matters more than the sector. A decision made over and over, record by record, and a history of how it turned out. These four kinds of business carry that shape most often.
Who it's for ›Mailing depth, reactivation, and who gets the expensive piece. The file already shows who covered their postage last time.
Reorder points set years ago, credit holds released by feel, and order desks keying in email by hand. The ERP holds the history to test all three.
Renewal calls go to the biggest premiums, and everyone else gets the same letter. The agency system already records who left and when.
Win-back offers sent to every lapsed customer, and markdowns set on Monday from a report nobody trusts. The order history knows who comes back and what sells at full price.
Every engagement ends with the same four things in your hands: the result running in your systems, the proof in your terms, the test designed in, and the record. Three samples, each labeled for what it is.
All outcomes ›All of it from the 44,988 names the habit kept mailing at a loss. Mailing everyone netted $10,560. The scored list netted $16,926.
Read the case›Three of six timed workflows get built on a distributor's own accounts. The stopwatch prices the six at $388,600 a year, where the room estimates $229,500.
Read the case›An agency routes the same 480 renewal calls a year by score instead of by premium. The calls then reach 74% more of the commission that leaves, $243,800 against $140,400.
Read the case›Brendan Hoffman spent three years as the sole analyst on a multi-billion dollar DTC e-commerce account. Kinetric is that work, augmented with AI and sized for operations that cannot hire a team. Six areas, three offerings, and one center they all branch from: analysis your own numbers can check.
All expertise ›Response and propensity models that turn one habit into a score, tested against what you do today.
Read more›Customer lifetime value, segmentation, and who to mail, call or win back.
Read more›A weekly readout built around the call someone makes on Monday, with the line that triggers it.
Read more›A/B tests, test cells and control groups, so you learn what a change caused, not only who answered.
Read more›The work your team does by hand, timed first, then running on your accounts and timed again.
Read more›Kinetric directs AI-assisted builds and checks every output before you see it. Each number traces to code that produced it.
Read more›Each sample is published in full, the way the budget owner would receive it. One runs on a public dataset. The others are illustrative engagements with composite figures, and every page says which.
View all results ›A lapsed-donor file of 96,367 names was mailed in full by habit. The score found the half that paid for its postage and beat every simpler cut by 28% to 66%.
Read the case›A composite distributor's order desk, timed with a stopwatch for a week. Three workflows get built, one wired to a score, and a customer portal is declined.
Read the case›A catalog and web retailer swaps a forty-tab sales report for one page built around the markdown decision. The readout shows revenue up 6% while gross margin is down 2.1 points, and $310,000 of markdown going to styles already on pace to sell out.
Read the case›Kinetric is Brendan Hoffman's practice. He spent three years as the sole analyst on a multi-billion dollar DTC e-commerce account: the propensity models, the A/B tests, and the readouts to the people who set the budget.
Kinetric is that work, augmented with AI and sized for operations that cannot hire it. No juniors, no hand-offs, and no new system. If your current rule holds up, you'll hear it from us first, with the comparison that shows it.
About Kinetric ›The Rules Review is a fixed fee for about a week. Everything else is a fixed fee, set once the decision and the data are named, at the scope call. You know the number before any work starts, and it does not move.
A Rules Review or an operations teardown takes about a week. A Scored Decision takes about four weeks, and nothing in your operation changes until the last week. A build sprint runs three to four weeks alongside your desk. A reporting engagement runs four to five weeks across its three depths.
Yes, and it says so on the front page. Kinetric is augmented analysis: the work is done with AI tools and checked before it counts. The models are statistics on your own history, with the expectation written down before the data is opened, and every number traces to the code that produced it. Nothing is predicted that your history cannot check, and the call stays with a person.
Then we'll tell you, with the comparison that shows it. Your current rule is the baseline in every test, and sometimes it wins. Knowing that is cheaper than changing a rule that works.
No. We need one person who owns the decision, one export of past results, and read-only access to the system it runs in. The handover trains that person to keep it going.
Not as Kinetric yet, and we won't pretend otherwise. Brendan did this work for three years on a multi-billion dollar DTC e-commerce account. The samples on this site are a public dataset and illustrative engagements, and each one says which.
With the Rules Review. In about a week it lists the rules that decide something in your operation, puts a money figure on each, and ranks them. You pick from a list, not a hunch.
It stays in your systems. We work from a read-only seat and an export, nothing moves off your system of record, and everything we build is yours to keep. No license.
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