How the UK Business Succession Index is built, what it counts, and what it cannot tell you.
Every figure derives from the Companies House register — the public record of UK company officers, filings and status. We hold the officer register for 3,475,770 active companies with at least one serving director whose age bracket is recorded. No survey data, no estimates, and no third-party lists are involved in the headline counts.
A company is counted when its oldest serving director is aged 60 or over. Serving means not resigned; age is taken from the director's recorded bracket, not an exact date of birth.
We use the oldest director rather than the board average deliberately. A company with a 70-year-old founder alongside two directors in their forties averages to the 50–60 bracket and would vanish from an average-based count — even though it is precisely the succession case the index exists to measure.
This is why our own figures differ from each other. On the same register, on the same day, 1,130,901 companies have an oldest director aged 60 or over, while 837,418 have an average director age of 60 or over. Both are correct; they answer different questions. This index publishes the oldest-director count throughout. Our sector pages and our published research use the average-based count and will therefore always report the smaller of the two. If you are comparing a figure from this index with one from elsewhere on this site, check which of the two you have before concluding that something has moved.
Companies House publishes directors' dates of birth to month precision. We deliberately store only the age bracket ('50-60', '60-70', '70+') rather than the date itself, to minimise the personal data held about identifiable individuals. That is a privacy choice with a analytical cost, and it is stated here rather than buried: because we hold a current bracket and not a birth date, we cannot reconstruct what any figure in this index would have been on a past date. See Limitations.
Two narrower figures appear alongside the headline. Both are subsets of it, never additions to it, and both come from our own scoring model rather than from the public register.
Business quality 50 or higher — 63,464 nationally. Our model scores business quality 0–100 from filed financials and company characteristics, and this counts those at 50 or above. It is a model output, not an audited opinion on solvency, and it is a threshold count rather than a verdict on any individual company.
It is a much smaller number than the headline, and the reason is us, not the data. We assess far more companies than clear this bar: most are scored and then withheld — chiefly by a quality floor that publishes no score at all below 35 out of 100 — and smaller companies fall outside the size range our routine scoring covers. So read this as "how many clear a deliberately high bar", never as "how many we could look at".
PURSUE — 7,004 nationally. A narrower commercial judgement combining exit timing and business quality, used to decide what we surface to our own customers. Included here for transparency about how much of the succession population we consider genuinely acquirable; it is the most model-dependent figure on these pages.
The index covers operating businesses. 10 classifications are counted in every total but are given no page of their own: dormant and special-purpose holding vehicles (which record the highest succession rates in the entire dataset, because holding property or assets through a company skews old — true, and not a succession problem), and classifier catch-all buckets such as 'Other', which target no meaningful category. A cell also needs at least 250 companies before it gets a page, so small trade-and-region combinations are reported only within their parent totals.
Ranking is a stronger claim than counting, so two further filters apply to the "highest succession pressure" tables — and only to those tables. No trade's figures change either way, and every excluded trade still counts towards every total published here; what is withheld is the ranking, not the data.
The index is regenerated from the live register and republished quarterly, on the first of January, April, July and October. These figures were drawn on 13 August 2026; the next update is due in October 2026. Every page states its own data date, so a figure read months after publication can be recognised as the published quarter's rather than mistaken for today's. Figures are not restated retrospectively: a number published for a given date stays as published.
These figures are free to quote and reproduce with attribution to ExitRadar and a link to the page used. If you need a cut that is not published here — a different trade grouping, a finer geography, or the underlying distribution — get in touch and we will usually provide it.
Back to the UK Business Succession Index.