What happens if collectors don’t follow the matched assignments?

Why matched accounts end up with someone else

In our view, skipped assignments come from four places, and only one of them is the collector’s choice:

• Habit. Collectors and team leads keep working the accounts and queues they know, especially in the first weeks.
• Cherry-picking. On lists collectors can work in any order, some pick the accounts that look easy or large and leave the rest. Call Centre Helper’s list of tricks agents use to avoid calls (Megan Jones, 2017) shows how inventive people get with status buttons and queues once a system can be gamed. Its advice is to look for patterns rather than single incidents.
• Dialer settings. On a predictive dialer, a call that connects while the assigned collector is busy can go to whoever is free, as it always has. Lists that load late, unworked accounts that are reassigned, and trading between collectors all move accounts too.
• Schedules. The assigned collector is off shift, on leave or has left.

What adherence is and how to measure it

Routing adherence is the share of contacts in the matched group that reached the collector they were assigned to. In words: the matched-group contacts where the collector on the call was the assigned collector, divided by all matched-group contacts, over the same period. Count it from the dialer’s call records joined to the assignment file by account ID, and break it down by team, shift and campaign as well as for the floor.

The control group has no adherence figure, because it stays on current routing. BestPair’s Order Form sets a floor: in Stage 2, at least 70% of BestPair group contacts must reach the assigned collector, and adherence is reported on every monthly statement.

What low adherence does to a test

Every matched-group account counts in the matched group, whoever worked it. That is what keeps the random split fair. It also means that every skipped assignment is an account worked much as it would have been anyway, so the measured gap between the groups shrinks toward zero. A test with low adherence understates the change it is testing.

The published case study shows this plainly. In a live test at a North American collection agency, on one telecom book in May 2026, adherence, how often the agent BestPair recommended was the agent who actually took the call, was about 50%; when that agent was busy, the call went to whoever was free. Those calls stayed in the BestPair half, and the case study says the test did not measure what happens when the recommended agent takes most of the calls. The case study.

Do not fix low adherence by dropping the skipped accounts from the count. If collectors skipped the hard ones, the accounts left behind are easier than the control group’s, and the comparison is no longer like with like.

How floors raise it

1. Settle pay and fairness first. Collectors on an incentive plan who fear a worse mix will find ways around it. How to get collectors to accept a new way of assigning accounts.
2. Agree the rules before day one: how long an unworked account stays with a collector, when it is reassigned, and whether collectors may trade.
3. Fix the plumbing. Lists that load on time, and dialer settings checked by the dialer admin, remove much of the leakage without anyone changing behaviour.
4. Check schedules against assignments. Leave, shift changes and leavers should be known before lists load.
5. Report adherence weekly to team leads, by team and by collector, so a dip is caught in days rather than at month end.
6. Coach before you police. Call Centre Helper’s guide to understanding call avoidance (Hannah Swankie, 2025) traces avoidance to stress, weak training, poor systems and aggressive targets, and recommends finding the cause and giving feedback before penalties.

In our view adherence is lowest in the first weeks, while habits and settings catch up. Why live pairing can trail at first covers how to read them, and how to prove a routing change worked covers the test itself.

Sources

• Call Centre Helper, the sneaky tricks call centre agents use to avoid calls Megan Jones, 2017, on how agents game status and queues, and how managers spot it.
• Call Centre Helper, Understanding Call Avoidance in Contact Centres Hannah Swankie, 2025, on the causes of call avoidance and how to address them.