A tier 1 Spanish bank measured Murphy against its own call centre
A tier 1 Spanish bank ran Murphy head to head against its own call centre on comparable accounts in early arrears. Over the pilot, the call centre recovered 77% of balance and Murphy recovered 81%, with every call scored against the bank's own rulebook. These are the figures the bank signed off for publication; the rest stays theirs.
Murphy, Editorial team2 October 20264 min read
How was the pilot set up?
The bank split comparable accounts in early arrears between its own call centre and Murphy, same products, same arrears band, same rules. The bank wrote the limits: what could be offered, when calls were allowed, what had to be said at the start of each one. Murphy worked only inside them.
Both sides were measured the same way, on share of balance recovered, so neither could win on a metric the other was not playing.
What did the pilot measure?
Over the pilot the call centre recovered 77% of balance on its accounts and Murphy recovered 81% on its own, a four point difference on the number the bank actually banks. Every Murphy call carried a compliance review against the bank's rulebook rather than a sampled one.
| Measured | Call centre | Murphy |
|---|---|---|
| Share of balance recovered | 77% | 81% |
| Arrears band | 0 to 90 days | 0 to 90 days |
| Calls reviewed for compliance | A sample | All of them |
77%
of balance recovered by the call centre
81%
of balance recovered by Murphy
100%
of Murphy's calls reviewed against the rules
What does the bank get beyond the number?
A recovery figure that survives an audit. Each call is linked to the account it touched, the rules in force at that moment and the person who approved them, so when someone asks why an offer was made, the answer is on file rather than in somebody's memory.
Common questions
Why is the bank not named?
At its request. The figures and the setup are published with the bank's sign off; the name is not. This is common for pilots in regulated institutions.
Were the two groups really comparable?
That was the design: same products, same arrears band, same rules on both sides, measured on the same number. A pilot that compares different books proves nothing, which is also the test to apply to anyone's pilot figures, including these.