Across PharmSee's now-14-city Boots audit (cycles 15-21), a structural pattern holds with remarkable consistency: Boots branches that had no within-0.5-mile sister branch in 2022 did not close during the 2023 rationalisation wave. Every Boots branch PharmSee has identified as a ghost in the register sits in a catchment where another Boots branch operated within half a mile — or did so before one of the pair closed.

The converse is equally clean: every 0%-ghost city (Dudley DY1, Bristol BS1) has zero within-0.5-mile Boots sister pairs in its 3-mile ring. Conversely, every city above 25% ghost rate has at least one documented within-0.5-mile sister pair.

This is the single-branch-catchment survival rule, and cycle 21's Leicester LE1 audit provides the cleanest live validation we have.

Leicester LE1 — the worked example

Leicester LE1 5WW 3-mile ring returns 8 Boots contractor codes: 7 operating, 1 ghost. That is a 12.5% ghost rate — low, consistent with the clean-cluster profile. The operating branches span LE1/LE2/LE3/LE4/LE5/LE19 postcodes. Running the within-0.5-mile sister test against the branch coordinates:

Two sister pairs in the 8-branch ring. One of those pairs has a ghost. Four branches have no within-0.5-mile sister at all (FPQ45 LE2 5BB, FAF98 LE4 1DG, FK682 LE3 3DH, FH317 LE19 1HJ). None of the four isolated branches are ghosts.

The survival rule holds: where Leicester had sister pairs, Leicester has a ghost in one of them (FFL52 at LE5 4BP closed while FWV53 at LE5 0QG survived). Where Leicester had no sister, no branch closed.

The FKX65-FN210 pair is the exception — both branches in the Leicester city-centre shopping corridor (Highcross mall + Haymarket centre) survived despite sitting 0.17 miles apart. The likely reason: these are complementary formats (mall flagship vs high-street presence) and each anchors a distinct retail environment. The rationalisation committee spared dual-format pairs where closing one would have forfeited a shopping-centre anchor tenancy.

Why the 0.5-mile threshold

The 0.5-mile figure emerged empirically from the cycle 19 ghost-rate audit. It's the walk-time threshold — roughly 7-10 minutes on foot — at which two pharmacy branches effectively compete for the same walk-in customer. Closer than 0.5 miles, the catchments overlap and the customer is indifferent between the two. Farther than 0.5 miles, each branch serves a distinct residential or commuter population.

When Boots rationalised its estate in 2022-2023, the internal rule was almost certainly: if we have two branches within walking distance of each other, keep the stronger one, close the weaker. Revenue data (where one sister meaningfully outperformed the other) decided which branch closed. Where Boots had isolated branches, no rationalisation was needed — each was serving its own catchment, and closing it would have ceded the corridor to a competitor.

The 14-city sister-pair map

Zero-sister cities have zero ghosts. Cities with 1-2 sister pairs have 1-2 ghosts. Cities with 3+ sister pairs cluster in the half-closed 40-57% range. The correlation is almost perfect.

The predictive cycle-22 forecast

The single-branch-catchment survival rule is predictive, not just descriptive. Cycle 22's job: apply it to the remaining operating Boots pairs PharmSee has already measured and forecast which branches are at risk in the next register update.

Predictions for cycle 22:

1. Leicester LE1 FKX65 + FN210 pair (0.17 mi apart, both operating): the cycle 21 exception. Either both survive because of their dual-format shopping-centre anchor status (most likely outcome), or the weaker of the two closes in the next rationalisation window. FN210 at £53,641 annual revenue is the weaker partner by a wide margin against FKX65 at £1,166,261. If rationalisation resumes, FN210 is the at-risk branch.

1. Birmingham B1 2JF Bull Ring (cycle 18 closure watch): FJV53 and FNM58 sit 0.13 miles apart in the Bull Ring corridor. Cycle 18 flagged FJV53 for 90-day closure monitoring. Cycle 21 re-check: both still operating. The survival rule predicts one of them will not survive the next register cycle.

1. Wirral CH41 FC368 + FMR00 pair: FC368 operating at £177,842, FMR00 already a ghost at CH41 7BX 1.61 mi away. The pair has already rationalised. No further at-risk branch in the Wirral CH41 core.

1. Leicester LE5 FFL52 + FWV53 pair: already rationalised — FFL52 closed, FWV53 survived. No further at-risk branch.

1. Nottingham NG1: 2 sister pairs, 4 ghosts. Rationalisation already complete in ratio terms. No new at-risk branches.

Why the rule matters for anyone modelling the English community pharmacy market

For investors: branch survival forecasts are the primary input to a 3-year Boots estate DCF. The 0.5-mile sister test is the cleanest single rule for predicting which branches survive the next rationalisation window. Combine with per-branch revenue and you have a ranking function.

For NHS commissioners: service contracts (Pharmacy First, NMS, flu vaccination) should not be awarded to a branch with a within-0.5-mile Boots sister without a contingency plan for the awarded branch closing. The structural closure risk is real.

For pharmacy professionals: if you're job-searching at Boots branches, the isolated branches are safer career destinations. Branches within 0.5 miles of a sister are structurally at risk of rationalisation. Use PharmSee's pharmacy search to verify branch-level data before accepting a Boots offer.

For pharmacy owners: if you operate an independent within 0.5 miles of two Boots branches, expect one of them to close, and the survivor's trading intensity to increase. This is a medium-term competitive pressure signal that the ghost-rate data doesn't show at ring level.

Methodology notes

The 0.5-mile threshold was derived from cycle 19 empirical observation across the 13-city dataset. Cycle 21's Leicester validation is the most recent test. The threshold is not a bright line — branches at 0.4-0.6 miles have softer closure risk than branches at 0.05-0.15 miles. Sister pairs inside shopping centres or mall anchors may survive the rule (the Leicester FKX65/FN210 exception). The rule should be applied as a probability weighting, not a deterministic prediction.

Future PharmSee feature work (cycle 22+): a "sister branch proximity" enrichment field on every operating Boots branch, exposed via API as nearestSisterDistance and sisterSurvivalRisk, allowing operators and commissioners to filter the live Boots estate for closure-exposed branches. This is one of the cycle 22 high-priority backlog items.

Related PharmSee

• Ghost rate does not predict operating average (cycle 21 methodology)
• Boots estate bimodality — three clusters (cycle 17)
• Birmingham B1 2JF Bull Ring closure watch (cycle 18)
• Wirral CH41 standalone audit (cycle 21)

Sources

• PharmSee 14-city Boots branch-level dataset, cycles 15-21
• Leicester LE1 5WW 3-mile location-analyzer query, 2026-04-11
• NHS Digital pharmacy contractor register (NHSBSA open data)