Deduction recovery is sold three ways, and the pricing model tells you more about a provider than its marketing does. It reveals who is carrying the risk, what the provider is optimising for, and what happens when the engagement goes badly.
This is an even-handed comparison first. We disclose our own model at the end, and we do not claim it is right for everyone.
The three models
Across the CPG deduction market you will encounter flat fee or subscription pricing, contingency pricing as a percentage of recovered dollars, and hybrid arrangements combining a software licence with services. Market surveys of CPG deduction recovery services describe contingency in the range of roughly 15% to 25% of recovered dollars as the most common structure in deduction-focused services, alongside SaaS-plus-services hybrids and per-claim flat fees (Opener).
Flat fee and subscription
You pay a fixed amount — monthly, annually, or per claim — regardless of what is recovered.
Where it works. High, predictable deduction volume; a mature internal process needing capacity rather than expertise; a finance team that values budget certainty above all. Cost per claim falls as volume rises, so at genuine scale a flat fee can be the cheapest option per dollar recovered.
Where it fails. You pay in months when nothing is recovered. The provider’s revenue is unaffected by whether your money comes back, so effort allocation depends entirely on the provider’s professionalism rather than on structure. And for a supplier who does not yet know the size of their deduction problem, a flat fee is a bet placed before the odds are known.
Flat-fee providers frequently pair the model with a guarantee — a commitment to recover some multiple of the fee, with a refund or credit if they do not. That is a real mitigation and worth asking about specifically: what triggers it, who measures it, and is the remedy cash or service credit?
Contingency
You pay a percentage of what is actually recovered. Nothing recovered, nothing owed.
Where it works. Uncertain or unknown deduction exposure; a first engagement where trust has not been established; a supplier who cannot get budget approval for an unproven spend. It converts a fixed cost into a variable one funded out of found money, which is the easiest business case a controller will ever write.
Where it fails. At very high volumes of small, easily-won claims, a percentage can exceed what a flat fee would have cost. And contingency creates a specific incentive distortion: the provider is paid for recovery, not prevention, so a poorly-designed contingency engagement can quietly benefit from your deduction problem persisting.
For context on how contingency behaves in adjacent receivables work, commercial collection agencies typically charge contingency rates in the 25% to 50% range, with flat-fee alternatives around $10 to $25 per account (AgentCollect; TI3). Deduction recovery rates sit well below that, largely because the claims are documented commercial disputes rather than distressed debt.
Hybrid: software plus services
A licence fee for a platform, plus a services layer either bundled or charged separately. This is the dominant enterprise model.
Where it works. Large suppliers with high volume across many retailers who need the reporting infrastructure as much as the recovery, and who have the internal team to operate a platform.
Where it fails. Implementation is a project with a timeline, an integration burden and an internal owner. The software is only as good as the team feeding it. And for a mid-market supplier with a two-person AR function, buying a platform to solve a staffing problem is a category error — you have added a system to operate, not removed work.
Side by side
| Flat fee / subscription | Contingency | Software + services | |
|---|---|---|---|
| Who carries recovery risk | The supplier | The provider | The supplier |
| Cost if nothing is recovered | Full fee | Nothing | Full licence fee |
| Budget predictability | High | Low in dollar terms, perfect in ratio terms | High |
| Cost per dollar at high volume | Falls with volume | Constant | Falls with volume |
| Setup burden | Low to moderate | Low | High — implementation project |
| Provider incentive | Deliver to scope | Maximise recovered dollars | Retain the licence |
| Natural fit | High, steady volume with a mature process | Unknown exposure, first engagement, lean team | Enterprise scale with internal ownership |
The question underneath: who carries the risk
Strip away the structures and one question remains: if the work produces nothing, who is out of pocket?
Under flat fee and software models, the supplier is. Under contingency, the provider is. That is not a moral distinction — it is a risk transfer, and risk transfer has a price. Contingency percentages are higher than the equivalent hourly economics precisely because the provider is absorbing the possibility of working for nothing.
The right question is therefore not “which is cheapest” but “how confident am I in the size of the recovery?” A supplier who knows their exposure precisely and has recovered against it before should probably buy capacity at a fixed price. A supplier who has never quantified the problem should not be writing cheques against a number they cannot yet see.
Incentive alignment and its failure modes
Every model has a characteristic pathology, and knowing them is more useful than picking a favourite:
- Flat fee — effort drifts toward the minimum defensible level, because additional recovery generates no additional revenue. Watch for declining claim throughput after month three.
- Contingency — cherry-picking the large, easy claims and leaving the long tail; and structural disinterest in prevention, because prevention shrinks the fee base.
- Software plus services — the licence renews whether or not the outcome materialised, so the commercial relationship centres on adoption metrics rather than recovered dollars.
The contingency pathology is the one to interrogate hardest, because it is the model we use. The honest answer is that it must be addressed contractually and behaviourally: root-cause reporting as a deliverable rather than a favour, no fee on claims the supplier recovers themselves, transparent reporting on the long tail, and a willingness to tell the client when the right answer is to bring the work in-house.
Questions to ask any provider
- What exactly is the fee calculated on — gross recovered dollars, net of the provider’s costs, or cash actually received?
- When is the fee invoiced: on approval, or after the money lands in our account?
- Are claims we would have recovered ourselves included in the fee base? What about claims already in progress when you started?
- Is there a minimum, a floor, or a monthly commitment underneath the contingency headline?
- What happens to fees on a recovery that is later reversed or re-deducted?
- Is root-cause analysis a contractual deliverable or an informal extra?
- What is the notice period, and who owns the work product and the data on exit?
- Do we have to buy, license or implement any software?
Question 2 is the one that most reliably separates providers. Being invoiced on approval rather than on cash received means you can owe a fee on money that never arrives.
Why Upstream chose contingency
We charge a percentage of recovered dollars, invoiced after the recovery lands in your account. No licence, no retainer, no implementation fee, and nothing owed if we recover nothing.
Three reasons, stated plainly:
- Most suppliers do not know their exposure. Asking a controller to approve a fixed spend against an unquantified problem is asking them to take a risk we are better placed to carry.
- It makes the business case trivial. The engagement is funded entirely out of money that was not coming back otherwise.
- It puts our result and your result in the same place. If we are wrong about what is recoverable, we absorb that, not you.
And the honest caveat: at very high volumes with a mature internal process, a flat fee or a well-run internal team may cost you less per recovered dollar. We would rather say that here than discover it awkwardly in month eight. On the root-cause incentive problem, our position is that remediation reporting is part of the engagement, not an upsell — the firm is named after it, and a client whose deduction volume falls is a client who stays.
See how the engagement runs, or the companion analysis on building the function in-house versus outsourcing it.
Frequently asked questions
What is a typical contingency rate for deduction recovery?
Market surveys of CPG deduction recovery services describe contingency in the range of roughly 15% to 25% of recovered dollars as the most common structure. Rates vary with volume, retailer mix and claim complexity.
Is flat fee or contingency cheaper for deduction recovery?
It depends on volume and certainty. At high, predictable claim volumes a flat fee usually costs less per recovered dollar. Where exposure is unknown or the relationship is new, contingency costs nothing if the work produces nothing, which is why it is the common starting structure.
What should I ask a deduction recovery provider about fees?
Whether the fee is on gross or net recovery, whether it is invoiced on approval or on cash received, whether claims you would have recovered yourself are in the fee base, whether there is a minimum, what happens if a recovery is later reversed, and whether root-cause analysis is contractual.
Does Upstream charge a retainer or require software?
No. Our fee is a percentage of dollars actually recovered, invoiced after the recovery lands. There is no licence, retainer or implementation fee, and nothing is owed if nothing is recovered.
Sources
Every figure in this article is drawn from the publicly available sources below. Retailer programmes, fee schedules and dispute windows change; confirm current terms in the retailer’s own supplier portal before acting.