What an Operator Actually Sees When They Audit a Claims Operation

A claims operation looks different from the inside than it does from the board pack. The board sees indemnity spend, average settlement times, customer NPS, and a colour-coded recovery rate. What I see, walking onto a claims floor for the first day of a Health Check 360, is the things those numbers do not show.

After nearly three decades running claims operations from the supplier, broker and insurer side, I notice the same handful of patterns on most engagements. Not because the people running these operations are doing a poor job, but because the things that matter most are usually the hardest to see from your own desk. Five of them, in the order I tend to find them.

1. Panel concentration nobody internally measures

Most insurer and broker claims operations have a panel. Few of them can tell you, off the top of their head, the percentage of their motor or property work that goes to the top three suppliers. Fewer still can give you the concentration broken out separately by total cost, by case volume, and by complexity tier.

When concentration sits above forty percent on any of those measures with a single supplier, you have a service continuity risk you may not have priced. When one supplier carries the top quartile of complexity, you have a recovery quality risk hidden inside what looks like a clean operation. Most boards have never seen this view. Most operations directors have never been asked for it.

The first thing I do on day one is build it.

2. KPI sets that reward speed over recovery quality

Claims KPIs almost always include cycle time, first-touch resolution, and customer satisfaction. They rarely include recovery percentage by sub-category of loss, deviation between first offer and final settlement, or rate of post-settlement reopen. The result is a claims operation that hits the targets it is measured on while losing money on the ones it is not.

The fix is not adding more KPIs. The fix is putting the recovery and quality measures where the operational ones currently sit, with the same management focus and the same monthly review. If you measure a thing on the same page as cycle time, it gets attention. If you measure it in a separate quarterly pack, it does not.

3. Broker-supplier loops where loss leaks

This is the one operators on the inside almost never see, because they are inside one part of the loop. A claim goes from broker to supplier, gets handled, settles, and goes back through. Each handover has paperwork, each paperwork step has a small cost, and the cumulative leakage on a year of claims volume is meaningful.

Mapping the loop, end to end, in one diagram with handovers and costs marked, almost always surfaces at least one step that exists for historical reasons rather than current need. Cutting one step a year is usually worth more than the engagement that found it.

4. Single-handler dependencies on the biggest accounts

Your largest broker or insurer relationship is usually managed by one senior handler. That handler has the relationships, knows the personal preferences of the account, and carries the institutional memory of past disputes. When they are on leave, the account either goes to someone less prepared or has to wait.

This is not a hiring problem. It is a documentation and cross-coverage problem. The fix is a one-page account brief per top-ten account, updated quarterly, with a named cover handler. Cheap, fast, and almost universally absent in operations that have not been audited recently.

5. Reserves that haven't moved in months

In every claims book of any size, there are files where the reserve was set at notification and never revisited. Some are appropriate. Many are stale. The pattern shows up as a long tail of mid-value files with reserves that have not been touched for six months or more, sitting alongside settlements that closed well above reserve.

Run the report. Sort by days since last reserve review. Look at the bottom thousand files. There will be money in there, and the conversation about reserve discipline that follows is one of the more useful outputs of any claims audit.

What to do with this

You do not need a consultancy to find any of these patterns. You need someone with permission to look across the operation rather than down through it, and the time to do it properly.

If you have that resource internally, use it. If you do not, or if the answer needs to be defensible to a board that does not want to hear it from inside the operation, that is when an independent senior view earns its keep.

The patterns above are what a Health Check 360 looks for early on. The rest of the work is about quantifying them and writing a defensible fix list. The whole point is to give you a written, costed answer you can act on, not to start an open-ended programme.