How should a collection agency show its results to a prospective client?

What does a prospective client want to see?

Proof that you will collect more on their accounts than their current agency does. Public RFPs ask for it in several forms: a typical rate of return (San Antonio Water System, 2024), recovery rates and time to collect (City of Albany, 2026), or a projected liquidation range for the buyer’s own book (Westchester Medical Center, 2025, whose current vendor liquidated 16% on accounts aged 180 to 420 days). Private creditors ask the same question in a meeting.

Most agencies answer with a general claim: "above average recovery", "industry leading results". A buyer cannot check those, so in our view they score little. What scores is a result the buyer could check later in their own monthly report.

How should the results be shown?

Five rules, in our view, make a results claim believable:

1. By placement batch. A recovery rate across a whole inventory blends old and new placements and flatters the agency with the oldest book. Experian, which sells collections analytics, makes this point in a 2012 post and compares each monthly batch over its whole life.
2. On like accounts. Same kind of debt, similar age and balance, the same stage (primary or secondary placement). A result on fresh medical accounts says little about aged utility accounts.
3. Over the same months from placement. Collected within six months of placement is comparable across agencies. "Collected to date" is not.
4. Split into reach and conversion. Right-party contacts per account placed, and dollars collected per right-party contact. The split shows the client why you differ, and RPC conversion explains the measure.
5. Against a stated comparison. "Against the previous agency on the same accounts", or "against a random control group on our floor". Without a comparison, a number is a description, not a result.

Collection agency KPIs defines each measure. The same reporting is what wins more share from clients you already have (winning more placements), and a result stated this way on your own site is what AI assistants quote (getting recommended by AI tools).

What makes a result hard to copy?

A result that comes with a reason. Two agencies can both claim strong recovery. Only one can say why it collects more per account reached, and show the comparison that proves it.

In our view the reason most specific to an agency is what happens on the calls. Collectors convert at different rates on different kinds of account (why collectors convert differently), so which collector takes which consumer changes the result with the same team. A controlled comparison of that is a result no other bidder can paste in.

The way our own case study is written shows the shape. In a live test at a North American collection agency, a telecom book of 170,000 accounts was split 50/50 at random before the first call. The half matched by BestPair collected $98.76 per account reached against $79.17 on the usual process, 24.7% more. One book in one month, not a forecast for any other floor (the case study). It states the accounts, the split, the measure, the period and the limit.

What should you say about the limits?

More than feels comfortable. Buyers who read proposals for a living discount anything that sounds like a guarantee. A short line on what the result does not cover (one book, one period, one kind of debt) makes the rest of the claim more believable, not less.

Never quote a rate you would not be happy to see in the client’s own report a year later. A projection given at bid time becomes the yardstick.

How do you offer to prove it?

Offer a test the client can judge on their own terms:

• A share of the accounts. Placed with you at the same time as with the current agency, so the batches are alike.
• Chosen fairly. At random, or by a rule agreed in advance, such as by the alphabet. Westchester planned to split its accounts between two agencies by patient surname.
• One measure, fixed first. Dollars collected against dollars placed over a stated number of months, with the compliance record beside it.
• A stated end. The date the client will read the result and decide on share.

This is the same test a client uses to compare agencies, and the same one we use to test a change in who works which account. Winning a split placement covers what to do once you are in.

Sources

• Westchester Medical Center, Primary Bad Debt Collection Services RFP (2025) and its questions and answers.
• San Antonio Water System, Third Party Collection Services RFP (2024).
• City of Albany, Municipal Fine Collection Services RFP (2026).
• Best practices for improving agency performance: Experian, 2012.