How is agent matching different from a propensity-to-pay score?
A propensity-to-pay score ranks accounts by how likely each is to pay, so the agency can decide which accounts to work, when and with how much effort. Predictive pairing (agent matching) decides which collector works each account, from the floor’s own call and payment history. They answer different questions, so an agency can keep its score and add pairing, and BestPair does not change which accounts are worked or the order the agency chooses.
New to BestPair? See what it is and how it works
Two different questions
A collection floor makes two decisions about every account. The first is whether to work it, when, and how hard: that is what a propensity-to-pay score is for. The second is which collector works it. The second is often answered by the next available collector, or by whoever ranks highest.
| Propensity-to-pay score | Predictive pairing | |
|---|---|---|
| The question | How likely is this account to pay? | Which collector should work this account? |
| What it produces | A rank or score for each account | An assigned collector for each account |
| What the agency does with it | Decides which accounts to work, when and with how much effort | Loads the assignments the way it loads lists today, or through the dialer’s skills-based routing |
| How the result is judged | How well the score ranks accounts that later pay | Dollars collected per account contacted, against a control group in the same weeks |
Why one score per account leaves a gap
A propensity score gives each account one number, as though the chance of payment were the same whoever makes the call. On a floor where collectors differ, it is not. One collector does better on one kind of account, another on a different kind, and the same account is worth more in the hands of the collector who suits it. Why collectors convert at different rates.
That is why the two answers do not compete. A score can tell you an account is worth working today. It cannot tell you which of your collectors should take it. Predictive pairing answers that second question from your floor’s own call and payment history. Sales floors face the same split between lead scoring and agent matching.
How they run together
Keep the score and the rules you build on it. Your agency still decides which accounts go to the dialer, in what order, and under which client and contact rules. Predictive pairing works inside those choices, the same way it works inside the group your skills routing already picked. How pairing sits with skills routing.
Nothing about your prioritisation has to change for the test to be fair. In live pairing, accounts in scope are split at random: 80% are allocated by BestPair and 20% stay on your current routing as the control group. Because the split is random, both groups carry the same mix of high-scoring and low-scoring accounts, and your prioritisation applies to both. The difference between them comes from who worked the accounts.
What the Floor Assessment shows a floor that already scores accounts
The Floor Assessment (Stage 1) works from up to 24 months of your history and changes nothing on your floor. Its report shows the spread between your collectors on the same kinds of accounts, and an estimated range for how much more might have been collected if each account had gone to its best-fit collector. It is an estimate from past data, not a forecast or a guarantee. If your collectors perform about the same on each kind of account, the report shows that too, and pairing has little to add to your score.
For more ways to raise collections without adding headcount, see raising collections per agent. Agencies with their own data team often ask the next question too: should we build pairing in-house?
Common follow-up questions
We already have propensity models. Does pairing add anything?
It can, if your collectors differ on the same kinds of accounts. Your score already tells you which accounts to work. Pairing works on the next decision, which collector takes each of them. Whether that second decision is worth anything on your floor is what the Floor Assessment measures, from your own history.
Do we have to send BestPair our scores?
The Order Form lists what BestPair asks for: call records, payments, an agent roster and account attributes (balance, days delinquent, placement date, consumer age band, state and ZIP code). A propensity score is not on that list. Your score, and the rules you build on it, stay in your systems.
Will pairing change which accounts our collectors call first?
BestPair changes only which agent works an account. It does not contact consumers, and it does not change whether an account is worked, what a consumer is offered, or any collection terms. Which accounts reach the dialer, and in what order, stays your decision.
What it could mean on your floor
Two minutes
Watch the explainer
What your dialer does with the next account, what pairing changes, and what the Floor Assessment measures. With captions.
Watch the videoYour floor
Try the calculator
Four numbers off your dialer and collections reports give an illustrative figure for your floor. It is at best informed guesswork until your own history is measured.
Open the calculatorA live test
Read the case study
170,000 telecom accounts split at random before the first call, and how the result was counted.
Read it
Would it be a bad idea to spend 45 minutes finding out?
If your shop does not qualify, you will hear it on the call, not after you have paid for a report.
Book a 45-minute fit call