BAMML Executive Insights
One Bad Store Is Usually One Bad Set of Controls
Every multi-site group has a bottom site. Arithmetic guarantees it.
What varies is what the group concludes about it — and the default conclusion, almost everywhere, is that it is about the manager.
Sometimes it is. More often, by the time anyone looks properly, two other explanations have never been tested.
What is actually happening
The first untested explanation is comparability.
Sites get measured against a group average or against each other, on the assumption that they are variations of the same thing. Frequently they are not. Catchment differs. Format and floor area differ. Lease cost differs, sometimes dramatically. Trading hours differ. Staffing models differ. Site age differs, and a site that opened last year has a different customer base from one trading for a decade.
A site can therefore be flagged as underperforming against a benchmark that never applied to it. The manager is then held to a target derived from someone else's circumstances, which is both unfair and — more practically — useless as a diagnostic.
The second untested explanation is control.
Groups measure sites on outcomes: sales, gross profit percentage, wage-to-turnover, sometimes footfall. They rarely measure sites on control discipline: stock count accuracy, cash variance, ordering against forecast, waste recording, price file accuracy, compliance with the standard operating routine.
So the group can see which site is failing but cannot see why — because the measures that would explain it are not collected.
Why it matters
Acting on outcome data alone produces two predictable errors.
Replacing a manager who was not the problem. The new manager inherits the same catchment, the same lease, the same control weaknesses and the same support model, and performs similarly. The group concludes that the site is difficult, and often the cycle repeats once more before anyone examines the site rather than the person.
Missing genuine control failure. Weak stock and cash control does not only depress margin. It creates the conditions in which loss — whether error or otherwise — is possible and undetectable. A site with poor control discipline and adequate sales looks acceptable on the outcome report while carrying real exposure.
There is a third, quieter cost. Managers at comparable sites watch how the bottom site is handled. If the group is seen to act on incomplete evidence, the effect on the rest of the network is not neutral.
The Integrated View
| Discipline | What it establishes | Consequence when skipped |
|---|---|---|
| Operational Performance | Whether the standard operating routine is actually being followed | Effort is judged by outcome only |
| Financial Governance | A comparable site P&L, adjusted for lease, format and catchment | Sites compared on figures that were never comparable |
| Workforce Governance | Whether the manager has the capability, staffing and support to succeed | Capability assumed from results |
| Risk & Compliance | Where weak control creates loss exposure | Visible only after a loss |
| Executive Decision Intelligence | Control measured alongside outcome | Which site is failing, but never why |
What good looks like
Run the comparability test before anything else. Normalise for lease cost, format, catchment and trading hours. Some sites move out of the bottom quartile the moment this is done — and that finding changes the entire conversation.
Then measure control, not only outcome. A short scorecard per site is enough: stock count accuracy and frequency, cash variance, ordering variance against forecast, waste recorded versus expected, price file accuracy, routine adherence. None of that requires new systems. It requires deciding to look.
Cross outcome against control and the diagnosis usually resolves itself. Poor outcome with good control points to the site or the market. Poor outcome with poor control points to execution, and tells you exactly which part. Good outcome with poor control is the quietly dangerous quadrant, and it is invisible on any outcome-only report.
Finally, ask whether the manager was supported. Training, staffing levels, equipment, the responsiveness of head office to problems raised. A manager failing without support is a group failure attributed to an individual.
What to do on Monday
- Normalise the site comparison for lease, format, catchment and hours. Does the bottom site stay at the bottom?
- Build a control scorecard — six measures, one page, every site, monthly.
- Plot outcome against control. The four quadrants will tell you more than any individual site review.
- Look at the good-outcome, poor-control sites. Nobody is worrying about them. Somebody should be.
- Before replacing anybody, ask what support that manager received and whether a strong manager would have succeeded in those conditions.
The question is not which site is worst. It is whether you can explain why — and an outcome report alone has never been able to answer that.
