The symptom
Consider an organisation whose core service standard was refined over several years in its original market and performed reliably there. The network expands into a second country. The standard is localised — language, currency, regulatory compliance are all updated, as they always are. Performance in the new market nonetheless lags, on one specific dimension, in a way none of the visible localisation work explains.
The instinctive diagnosis is insufficient localisation: someone did not adapt the standard thoroughly enough for the new market. That diagnosis sends the organisation looking for the parts of the standard that were not translated or adjusted. But suppose, in this case, that search comes up short — the visible layer really was adapted properly. What underperforms turns out to rest on a customer expectation about response timing, one baked so deep into the original design that nobody involved in localising the standard ever identified it as a variable, let alone a cultural one.
Why the invisible parts of a standard don’t get updated
This is a specific and well-studied kind of organisational blind spot. Erik Hollnagel and colleagues, in research spanning the University of Southern Denmark, the University of Florida and Macquarie University, distinguish between work as formally imagined — the documented standard, procedure or script — and work as actually done, the accumulated adjustments that make the documented version function in real conditions. Their central observation is that most standards work not because people follow them exactly, but because people quietly adapt them to fit circumstances the design never anticipated. Franchise standards are a clean example of “work as imagined”: written down, trained on, audited against. What Hollnagel’s research suggests is that some of what makes a standard work in its original market is not in the document at all — it is in the tacit, unrecorded adjustments staff made to fit local expectations, adjustments so embedded that the people making them no longer experience them as adjustments.
This matters for localisation specifically because a localisation process, by design, reviews the document. It asks: what in this standard needs to change for the new market? That question only surfaces assumptions that were written down as assumptions. An assumption so deeply embedded that it reads as simply “how the service works” will not be flagged, because nobody localising the standard experiences it as a variable in the first place. It is not that localisation was done badly. It is that localisation, as a review process, can only find what it knows to look for.
Jeffrey Bradach’s study of large US restaurant chains, published in Administrative Science Quarterly, is a useful complement here. Bradach found that chains combining company-owned and franchised units use structure, control systems, career paths and strategy-making processes together to manage the gap between a standard designed centrally and conditions encountered locally — precisely because no single mechanism, including the written standard itself, reliably captures everything a location needs to perform well. A standard is never the whole story of why a location succeeds; it is one input among several, and the others are harder to see.
What this looks like as it develops
Divergence of this kind tends to follow a recognisable arc inside a growing network. It is close to invisible in the first one or two new markets, because leadership is paying close attention and any gap between the standard and local performance gets caught quickly through direct oversight. It becomes harder to catch as the network adds markets faster than that direct attention can scale, and easiest to miss of all in markets that are performing adequately but not as well as the model predicted — underperformance that is real but not dramatic enough to trigger investigation.
A related pattern shows up within a single market over time, not just across markets. A standard that performed well when the network was small can start to diverge from performance even in its home market, as the customer base itself shifts — new competitors change what customers compare the service against, demographic composition changes, expectations move — while the standard, having proven itself once, is treated as settled and stops being actively questioned. In both cases, what looks like a compliance or execution problem is often a documentation problem: the standard captures less of what actually makes performance work than the organisation believes it does.
The practical difficulty this creates is that the standard cannot simply be asked “what is missing,” because the gap is, by definition, the part nobody currently recognises as part of the standard. What tends to surface it instead is comparison — a location, a market, or a time period where performance genuinely diverges from what the standard predicts, examined closely enough to ask what is different about the conditions rather than what was done wrong.
Maturity of the network changes how much of this is tolerable. A young network, still refining its core standard in its founding market, has an advantage here: the standard has not yet been treated as finished, so questioning it does not feel like undoing settled work. A mature network, with a standard that has been stable and successful for years, faces the opposite dynamic — the standard has accumulated the authority of a track record, which makes it harder, organisationally, to treat any part of it as provisional. Ironically, this means the networks most exposed to this kind of quiet divergence are often the ones with the longest history of getting it right, because success is exactly what makes a standard stop being examined.
Where this shows up
This perspective is written with franchise networks in mind. It sits within the Multi-Site and Franchise grouping, and the questions it raises recur in other sectors too.
Questions worth considering
- Which elements of the current standard were deliberately designed, and which have simply never been questioned because they have always worked?
- When a location or market underperforms despite following the standard, is the response to audit compliance, or to ask what the standard assumed that may not hold there?
- How would the organisation know if a customer expectation embedded in the standard had shifted, in a market where the standard has not changed?
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.