Published June 22, 2026
The recovery business that takes the lost column, holds back a control group, and gets paid only for the customers it truly brings back.
1
The reacquisition tax — and the profit hiding in it
Every brand has a reacquisition budget. It is the money spent each quarter winning back customers who drifted, lapsed, or simply stopped buying. Today most of that money goes to adtech — the brand pays a platform to buy back a customer it already owned, often more than once. That is AdWaste, and at scale it is the largest controllable line in the marketing P&L that nobody has ever named.
Two numbers make it visible. REACQ% is the share of customers being re-bought through paid media — in many catalogues it runs north of thirty per cent. The Adtech:Martech ratio is how much of revenue leaves as a platform tax versus how much is spent owning the relationship. Together they answer one board-ready question: how much of last quarter’s spend was structurally avoidable?
The reason this matters to the CFO and not just the CMO is the margin. Recovered revenue carries almost no marginal cost — the customer is already known, already acquired — so a dollar moved out of AdWaste and into recovery flows almost entirely to contribution margin. This is the difference between buying growth and freeing profit. A brand does not need to sell more to earn more here; it needs to stop paying twice for what it already has. Progency exists to compete for that reacquisition dollar — not as another services vendor, but as a cheaper, measurable alternative to buying the customer back through an auction.
2
What Progency is — and what it deliberately is not
Progency takes one job, and one only: Recover — the lost column, the customers whose attention has gone. It does not run acquisition, it does not run the everyday programme, it does not touch the active base. The CRM team keeps Grow and Protect; the acquisition team keeps Acquire. Progency is the specialist called in for the customers everyone else has given up on.
Its place in the stack is precise: post-CRM, pre-Adtech. It works the gap after the CRM team has done what it can but before the brand falls through to paid retargeting, where the reacquisition waste happens. It is not a replacement for the CRM team and not a competitor to it — it is an aid with skin in the game, a partner paid only if it brings the lost back.
Progency is the missing rung — a measurable recovery route between cheap owned channels and expensive paid media.
And it is bought differently from everything else in martech. A brand does not licence a tool or hire a team and hope for results. It buys recovered customers — an outcome, not a capability. The engines that do the work sit inside Progency and are explained in the companion essay; what matters commercially is that the brand pays for what comes back, not for the machinery that brings it.
| The boundary. Progency takes the lost column, and nothing else. The sentence should sound limiting. It is the limitation that makes the model credible. |
3
Beta, Alpha, Carry — how the pricing works
The pricing rests on three words. Beta is the revenue the brand would have earned anyway — the baseline. Alpha is the incremental revenue created above it. Carry is Progency’s share of the Alpha alone, after the baseline has removed everything that would have happened regardless. There is no retainer and no fixed software fee. If there is no Alpha, there is no Carry.
Carry is a claim on proven lift, never on the baseline and never a fixed fee.
The obvious worry is the baseline: who decides what the brand ‘would have earned anyway’? For recovery, nobody decides it — it is measured live, by holding back a control group the brand can audit for itself. The lapsed customers are split at random into three arms: one Progency works, one left completely untouched, and one given to adtech.
The three-arm holdout: the brand’s own control is the baseline — the vendor cannot move it.
The untouched group is the proof that some of those customers would have returned with no help at all. So Progency is billed not on the customers it recovered but on the lift it added above the control — the lift, never the gross — and at a fraction of what the adtech arm cost to deliver its smaller lift.
| The commercial rule. No lift, no bill. No control, no claim. No fixed fee. |
4
Why you cannot be gamed — and how a deal runs
The sharpest objection to outcome pricing is that the vendor sets the baseline it is paid to beat: lowball the baseline, and the Alpha is invented. The answer is built into the method, not promised in a contract. The brand’s own holdout is the baseline, and the vendor cannot move it. There is no number to negotiate, because the comparison is a live control group drawn from the brand’s own lapsed customers.
Two further worries are settled the same way. ‘Your control still sees my other marketing’ is true, but because the groups are split at random, every other influence lands on both equally — so the gap between them is still Progency’s, reported as a range rather than a falsely exact figure. And ‘your numbers are unproven’ is met by publishing the pilots that underperformed alongside the ones that worked. The test of whether a partner believes its own numbers is whether it will hold back a control group and take its fee only on the lift.
In practice a deal runs in a fixed shape. The lapsed base is defined and split at random into the three arms. A measurement window is agreed up front — long enough for recovery to show, short enough to bill cleanly. Progency works its arm; the control is left strictly alone; the adtech arm runs as the brand would have run it anyway. At the end of the window the arms are compared, the lift is counted, and Carry is billed on the lift at the agreed share. The brand audits the control itself.
The deeper point is who carries the risk. Almost all martech is sold as a fixed fee, which means the brand pays whether or not anything improves. Here the vendor carries it: no lift, no bill. That is what Never Pay Fixed means in practice — the risk sits with the party doing the work.
5
The CFO decision — versus adtech, the sequence, and when not to buy
The number that actually moves a budget is the comparison between a Progency-recovered dollar and an adtech dollar, because the two compete for the very same reacquisition spend. On the same lapsed customers, recovery returns more than adtech does — and that figure is measured by the three-arm holdout, not asserted. Every reacquisition dollar that moves from adtech to recovery buys more customers back, more cheaply, with the proof attached.
But the order is not optional. Exhaust the owned channel first — the everyday Grow and Protect work the CRM team already does. Then recover through Progency. Then, and only then, spend adtech, as the last resort for what neither could reach. The reacquisition dollar should always reach the cheapest proven route before the most expensive one. This is also why Progency competes with adtech, not with the CRM team: it is taking budget from the platform tax, not from anyone’s headcount.
And because the honest case for any doctrine is naming who it is not for, Progency is not for every brand. It needs identifiable customers, repeat or renewal economics, real owned channels, and enough lost-customer volume to run a control group. It does not fit a pure acquisition-stage brand still proving product-market fit, a single-purchase category with no repeat, or a marketplace seller with no path to direct identity.
| The cleanest test. If you cannot hold back a control group, you are not ready for Progency — because without the control there is no honest way to know what it earned. |
* * *
For the far larger set of brands that can, recovery is the missing rung between owned and paid: a way to stop paying twice for customers they already own, priced so the vendor never gets paid for fixed promises. Never Lose Customers. Never Pay Twice. Never Pay Fixed.
