Reward, cashback, affiliate and creator platforms
Can the transaction engine carry the plan you are paying for?
Follow a transaction through the target from the moment it happens to the moment everyone has been paid. Then follow a merchant from signature to the first commission it ever earns.
The join between those 2 is where the plan gets decided, and it is the part nobody has been asked to look at.
I take a small number of engagements at a time and I do the work myself.
A quality of earnings review tells you whether the numbers are true. Legal tells you whether the contracts hold.
This tells you whether the machine that produces the numbers can carry the plan.
What a quality of earnings review does not reach
It is not that anybody missed it. It is that it does not appear as a line.
Commission that was earned and never invoiced sits inside gross commission cost. Goodwill paid out to members because a transaction could not be traced sits in the same place. The people working the exception queue by hand sit in operations headcount, where they look like the cost of running the business rather than the cost of a field nobody captures.
None of it is hidden. All of it is absorbed. And every one of those lines scales with volume, which means the growth case is the thing that makes them bigger.
Publisher payments routinely take longer than 120 days.
I follow the money twice
Once from the sale through to the payment landing.
And once from onboarding a merchant through to the first commission it ever earns.
Most businesses have only ever mapped the first one. Most data rooms only contain the first one.
The second one pays for the first. A member cannot be paid until the merchant has.
The growth plan you have been shown is a journey 1 forecast. Whether it can be delivered is a journey 2 question.
What gets looked at
- Partner onboarding - how a merchant gets from signed to earning.
- Recording what happened - tracking, attribution, matching, the tracking file.
- The checks before money moves - validation, approval, auto-approve windows, reversals.
- Invoicing, payouts and reconciliation - the invoice cycle, payout runs, settlement, what does not reconcile.
- Data into decisions - what the numbers are used for once they exist, and who sees them.
- Team and process resilience - what happens when the person who knows breaks their leg.
Scored against 5 pillars: settlement and cash movement, commercial economics, attribution and allocation, leakage and operational risk, execution and scalability.
What you get
3 pages, inside your timetable.
Page 1. The risk profile across the 5 pillars, with a rating against each, what checks out, and the overall read.
Page 2. Every finding that carries a number, written as what was found, what it costs and what it means for the transaction. Annual margin separated from working capital, with a total.
Page 3. What the first 100 days after completion should do about it, in order of value against effort, with an owner and an effort estimate against each move. Then what the assessment is based on, and what it does not cover.
It is a constructed example on a made-up target, so there is nothing confidential in it and you can pass it round.
Before completion, and after it
Before completion
I work from the data room, the transaction records and 2 conversations with the people who work the queues. You get the 3 pages inside your timetable, and a call to walk through anything that moves the price.
After completion
The first 100 days, with me doing the work rather than writing a plan for somebody else to do it. Stop what is leaking, re-sequence what is funding itself, and put a limit on the terms that carry no limit.
Scope and a fixed fee in writing after a conversation. Nothing on this page is a price list.
Who does the work
I do. There is no team behind this and no associate doing the first pass.
20 years in operations. 9 and a half of them inside a cashback publisher, across 12 affiliate networks and an estate of 4,200 retailers, running the part between a transaction happening and everybody being paid. 4 years since then in payments and embedded finance.
I have been the person whose job it was to explain to a merchant why an invoice was wrong, and the person who had to find out why. That is a different kind of looking from a workstream.
Find where the money stands still.