UK Tech Startup Survival Curves
Only 37% of UK tech companies born in 2019 were still active five years later, tracked from ONS Business Demography official cohort data. Data pulled 2026-07-23.
Source: Business Demography, UK: 2024 (Office for National Statistics). Licence: Open Government Licence v3.0.
Key findings
- Of 30,080 tech companies (Information and Communication) born in 2019, 37.0% were still active five years later, against 38.4% across all industries in the same cohort.
- Tech survival tracks close to the all-industry average at every horizon from 1 to 5 years, running against the popular narrative that tech startups fail at dramatically higher rates than other small businesses.
- The steepest drop for tech companies happens between year 2 and year 3: roughly a fifth of the survivors at 2 years are gone by year 3, before the rate of further loss slows in years 4 and 5.
- One-year survival for the most recent 2023 birth cohort is 94.9%, in line with prior cohorts.
Source: Business Demography, UK: 2024, under the Open Government Licence v3.0. Figures may be cited with attribution to Founder Tax Partners.
Survival curve: 2019 birth cohort
Percentage of the 2019 birth cohort still active at each anniversary, tech (Information and Communication) versus all industries.
One-year survival across birth cohorts, 2019 to 2023
One-year survival is stable across cohorts, including through the pandemic-era 2020 birth cohort, which some might expect to show a Covid-era dip.
Two-year survival by industry, 2022 birth cohort
How tech compares to other broad industry groups at the two-year survival mark, highest first.
Bars show two-year survival percentage. Source: ONS Business Demography (Table 4.2), OGL v3.0.
Methodology and honest limitations
Source
All figures are read directly from ONS Business Demography Table 4.2 (Survival of newly born enterprises by broad industry group). An enterprise 'births' in a given year if it is a new legal unit with no predecessor. Survival at N years means the enterprise was still active N years after birth, regardless of employment or turnover change. 'Information and communication' is ONS's broad industry group covering publishing, software, telecoms, broadcasting and IT services (SIC sections 58 to 63), a broader grouping than software-only. Counts are control-rounded to the base 5 by ONS. Only the 2019 birth cohort has a complete 5-year survival window in this edition; more recent cohorts (2020 to 2023) have progressively shorter windows because fewer years have elapsed since birth.
Caveats
- 'Information and communication' is a broad ONS industry group (SIC sections 58-63: publishing, software, telecoms, broadcasting, IT services), not a software/SaaS-only cut. It is the closest available official grouping to 'tech startups'.
- Survival counts are control-rounded by ONS to the nearest 5, so percentages calculated from the published counts may differ very slightly from ONS's own published percentages (both are shown here; ONS's own percentage is used).
- Only the 2019 birth cohort has a full 5-year survival reading in this edition. Later cohorts (2020-2023) are shown with only the survival horizons that have actually elapsed; cells marked as unavailable in the source ('suppressed pending data') are omitted, not treated as zero.
- 'Survival' means the enterprise remained on the statistical business register (broadly, still trading or dormant with an active VAT/PAYE registration); it does not measure profitability, growth, or founder outcome. A company can survive while shrinking, and a company can be a successful trade sale or wind-down (not a 'failure') while counting as dissolved.
- This is an enterprise-level series (ONS Business Demography), separate from and not directly comparable to the Companies House incorporation/dissolution snapshot used in this site's other formation-tracking research pages.
Re-running the pull
The pull script is at startups-tech/pipeline/pull_tech_survival_index.py. It downloads ONS's current Business Demography reference tables and regenerates the JSON file that powers this page.
Sources and how to cite
Primary source
Business Demography, UK: 2024. Publisher: Office for National Statistics. Licence: Open Government Licence v3.0. Data pulled 2026-07-23.
Cite this index as
UK Tech Startup Survival Curves, derived from ONS Business Demography, UK: 2024 (Table 4.2, reference tables). Published under OGL3. Data pulled 2026-07-23.
Download the full dataset (CSV)
Last updated: 2026-07-23.
Building a company that lasts?
Runway management, board-ready reporting and getting your structure right from day one all improve your odds of being in the surviving majority. Our fractional CFO and core compliance services support founders through the years where most attrition happens.
Frequently asked questions
What percentage of UK tech startups survive five years?
Of UK companies in the Information and Communication industry group born in 2019, 37% were still active five years later, according to ONS Business Demography data. This compares to 38% across all industries for the same birth cohort. 'Survival' means the enterprise remained on the statistical business register, whether trading, dormant, or scaled down; it is not a measure of profitability or founder success.
Why does ONS use 'Information and communication' rather than 'tech'?
ONS's Business Demography statistics classify enterprises by broad industry group derived from Standard Industrial Classification (SIC) codes, not by a bespoke 'tech' category. Information and Communication (SIC sections 58 to 63) is the closest official grouping: it covers software publishing and development, IT consultancy, data processing and hosting, telecoms, and broadcasting. It is broader than software/SaaS alone, so the true survival rate for a narrower software-only cohort may differ from the figures shown here.
Is tech riskier than other industries for a new business?
The data shows tech survival tracking close to, and in this edition slightly below, the all-industry average at the five-year mark (37% vs 38%). This runs counter to the popular narrative that tech startups fail at dramatically higher rates than other sectors: at the enterprise-survival level (not funding or profitability), most UK tech companies are not meaningfully riskier than the average small business.
What does 'survival' actually measure here?
An enterprise is counted as having survived at year N if it remains active on the UK statistical business register N years after birth, regardless of whether it is trading profitably, dormant, or has shrunk substantially. It does not distinguish a genuine failure from a successful trade sale that removed the original legal entity, or from a founder who deliberately wound up a profitable company. It is a register-continuity measure, not a success or failure judgement.
Where does this data come from?
All figures come from ONS Business Demography, the UK's official annual statistics on enterprise births, deaths and survival, published each November on gov.uk under the Open Government Licence v3.0. This index uses Table 4.2 from the reference tables, which tracks each annual birth cohort's survival at one to five years after birth. Only the oldest cohort in each edition has a complete five-year reading; more recent cohorts have progressively shorter observed windows.
How does this differ from the UK Startup Formation and Survival Index?
This page uses ONS cohort-based Business Demography survival curves (what share of a birth cohort is still trading after 1 to 5 years). The site's UK Startup Formation and Survival Index instead reports a live Companies House active-versus-dissolved snapshot across a broader 8-SIC cluster.