How to identify fast-growing companies in one country or region

To identify fast-growing companies in a specific country or region, two decisions come before any data: what counts as fast, and what counts as being in the country. This guide takes the first from the statistics offices, which define a high-growth enterprise by a rate, a period and a minimum size. It takes the second from the records a company leaves in each country: its listing, its job postings and its website. You get queries with the country as a parameter, a table of measures by what each needs, and a worked example, written for the analyst who has to say why each name is on the list.

How do you identify fast-growing companies in a specific country or region?

To identify fast-growing companies in a specific country or region, settle three things. Fix one growth test: a rate, a period and a minimum size, as official statistics do. Choose what places a company in the country: its domicile, its hiring or its selling. Compute the same measure for every company, with the country as a parameter.

1. Fix the growth test. Take the threshold, the period and the minimum size from an official definition and write them as parameters.
2. Choose the test of place. Domicile, hiring and selling give three different lists for the same country.
3. Compute one measure per company. Stated headcount where a company has two counts, open postings by country for every company that publishes roles.
4. Open the evidence. A posting records an intention to hire, and a stated count speaks for one date, so read the postings and statements behind each name.

Sourcing private companies for venture capital scores companies on several signals at once, and catching companies as they enter a new market builds the timeline of one entry. This guide holds one growth test still and changes the place.

How official statistics define a high-growth enterprise

Two thresholds are in use, each with a rate, a period and a minimum size.

Eurostat's glossary defines a high-growth enterprise as one whose number of employees grows by more than 10% a year on average over a three-year period, with at least 10 employees at the start. An enterprise born in the starting year is left out, and the entry adds that growth can also be measured by turnover. Eurostat's information on business demography data repeats the test, calls high-growth enterprises that are four and five years old gazelles, and bases the data on European business statistics legislation from reference year 2021.

Eurostat's metadata for its business demography tables of 2004 to 2020, last updated on 10 February 2021, records the definitions of the Eurostat-OECD Manual on Business Demography Statistics: growth above 20% a year over three years is high growth, and growth between 10 and 20% is medium growth. The same page gives the reason for a size test, since an enterprise going from one employee to two in three years would otherwise count.

The UK's Office for National Statistics applies the 20% threshold. Its release Business demography, UK: 2024, published on 20 November 2025, covers businesses registered for VAT or PAYE, and at Companies House. It measures those with average growth in employment above 20% a year between 2021 and 2024 and a size of 10 or more employees. It counts 14,330 high-growth businesses in the UK in 2024, a rate of 4.9% of active businesses with 10 or more employees, and breaks the count down by region.

Read as a compound rate, 10% a year for three years means an end count above 1.331 times the start (1.1 × 1.1 × 1.1), and 20% a year means above 1.728 times (1.2 × 1.2 × 1.2).

Eurostat and the ONS count these enterprises and name none of them. Eurostat's regional figures follow the NUTS classification, down to NUTS 3 regions, and in the 2004 to 2020 tables the unit counted is the enterprise, which many countries reported as the legal unit. The ONS states in its quality and methodology information that statistical disclosure control is applied so that information attributable to an individual organisation is not disclosed. Eurostat's metadata cites Regulation (EC) No 223/2009 on European statistics and states that data flagged confidential under it are not published. A list of names therefore comes from company-level data, measured on the same test.

What "in a country" means for a company that hires and sells in several

The ONS gives each business one place. It states that the location of a business is generally its main operating site or its head office. That is a test of domicile, and it is one of three a list can use. A company with a head office in one country, engineers in a second and customers in a third passes a different test in each.

The country on a posting is the country of its first location. The country on a listing is the issuer's, and the venue its shares trade on is a separate field, the market identifier code. A listed company's rows carry its listing identifiers whichever country its postings name, so filter on one country and report the other beside it.

A region is a list. For a region made of countries, run the queries below once per ISO 3166-1 alpha-2 code and add the counts. Inside a country, each posting location carries a city and a region beside its country, which you map to the regions you report on.

Which metrics screen high-growth companies by country?

Four metrics screen high-growth companies by country. Annualised growth in stated headcount over three years applies the official test to one company. The change in open postings located in the country measures hiring pace there. A first posting in the country marks an entry, and the weekly country series gives the baseline to judge each company against.

The queries are PostgreSQL over datasets loaded from bulk files, with JSON columns as jsonb. A name that starts with a colon is a parameter: the country, every date and every threshold. The hiring and announcements and scale pages describe the datasets, and the data dictionary defines each field.

The official test on stated headcount

with rated as (
  select e.company_id,
         s.employees as employees_start, s.as_of as start_date,
         e.employees as employees_end,   e.as_of as end_date,
         power(e.employees::numeric / s.employees,
               365.25 / (e.as_of - s.as_of)) - 1 as growth_a_year
  from company_headcounts e
  join company_headcounts s
    on s.company_id = e.company_id
   and s.source = e.source
   and e.as_of - s.as_of between :min_days and :max_days
  where e.as_of between :end_from and :end_to
    and s.employees >= :min_employees
)
select *
from rated
where growth_a_year > :threshold
  and company_id in (
    select company_id from listed_securities where country = :country
  )
order by growth_a_year desc;

In Employee Headcount, employees is the count as stated and as_of the date it refers to. The join pairs two counts of the same source, with the later one dated between :end_from and :end_to. :min_days and :max_days set how close to three years the pair must be: 1,065 and 1,125 allow thirty days either side. :min_employees is the size test and :threshold the rate, 0.10 or 0.20. In the listed securities table country is the issuer's country, so the last filter is the domicile test.

Hiring pace in the country

with ends as (
  select company_id,
         max(case when day = :start_day
                  then coalesce((by_country ->> :country)::int, 0) end) as open_start,
         max(case when day = :end_day
                  then coalesce((by_country ->> :country)::int, 0) end) as open_end,
         max(case when day = :end_day then open_postings end)          as open_all
  from company_hiring_daily
  where day in (:start_day, :end_day)
  group by company_id
)
select company_id, open_start, open_end,
       round(100.0 * (open_end - open_start) / open_start, 1) as change_pct,
       round(100.0 * open_end / nullif(open_all, 0), 1)       as country_share_pct
from ends
where open_start >= :min_open
order by change_pct desc;

In Hiring Activity, by_country holds the postings open at the end of each day by the posting's country, which is the country of its first location. :min_open is the minimum base, the counterpart of the 10-employee test. This query needs no listing: the place comes from the postings alone, so it covers private companies on the same terms.

A first posting in the country

select company_id, min(first_seen_at) as first_opening
from job_postings
where country = :country
group by company_id
having not bool_or(found_on_first_read)
   and min(first_seen_at) >= :from_date
   and min(first_seen_at) <  :to_date;

In Job Postings, found_on_first_read marks a posting already on the board when it was first read. The having clause keeps a company only when every posting it has had in the country was seen opening, the earliest of them inside the window, so load postings from well before :from_date. On the website, Technology Changes writes a market event with the country code as its key and added as its change when a site begins to declare a version for a country.

Keeping one definition of growth across countries

• One set of parameters. Thresholds, window, minimum base and dates are written once. Only the country changes between runs.
• A baseline per country. Market Series gives weekly figures by headquarters country on same-store cohorts, among them new postings and the median year-over-year change in stated headcount, each with its growth against the previous window. Set a company's rate beside its country's.
• Rates across countries. Compare countries on rates over the companies covered.

A worked example: one country, four invented companies

Acme Robotics is an invented company, as are Company B, Company C and Company D, and this example is illustrative.

Assumptions:

• The country parameter is DE, an arbitrary choice.
• The growth test is more than 10% a year over three years from at least 10 employees, with the 20% threshold read beside it.
• Stated headcounts are as of 31 December 2023 and 31 December 2026.
• Open postings are read on 31 March and 30 June 2027, with a minimum base of 10 in the country on the first day.
• Acme Robotics: issuer's country DE, 400 employees then 620, 30 open postings in DE then 45, of 80 in all countries on the second day.
• Company B: issuer's country US, 2,000 employees then 2,300, no posting in DE on the first day and 24 on the second, the first seen opening on 14 April 2027.
• Company C: issuer's country DE, 8 employees then 20, 2 open postings in DE then 5.
• Company D: issuer's country DE, 1,000 employees then 1,400, 12 open postings in DE then 10, and 5 in PL then 60.

The working:

1. Headcount ratios. Acme Robotics 620 ÷ 400 = 1.55. Company B 2,300 ÷ 2,000 = 1.15. Company C 20 ÷ 8 = 2.50. Company D 1,400 ÷ 1,000 = 1.40. The 10% test asks for a ratio above 1.331 and the 20% test for one above 1.728.
2. Rates a year. The cube root of each ratio, less one: 15.7% for Acme Robotics (1.157 × 1.157 × 1.157 = 1.549), 4.8% for Company B, 35.7% for Company C and 11.9% for Company D.
3. The size test. Company C starts with 8 employees, under the minimum of 10, so its rate is left unread.
4. Hiring in the country. Acme Robotics: 45 less 30 is 15, and 15 ÷ 30 = 50.0% up, with 45 ÷ 80 = 56.3% of its open postings in DE. Company D: 12 less 10 is 2, and 2 ÷ 12 = 16.7% down. Company B has no base, and Company C's base of 2 is under the minimum.

One country gives three lists. Domiciled there and above the 10% test: Acme Robotics and Company D. Growing there by hiring: Acme Robotics alone. Entering: Company B. No company passes the 20% test, since Company C fails on size. Company D shows why the tests stay apart: by domicile it is a high-growth company of the country, while its open postings there fell and those in PL went from 5 to 60. A list of companies based in the country keeps it, and a list of employers growing there leaves it off.

How to identify high-growth companies by country: revenue growth, hiring, funding, startup databases such as Crunchbase and the official methodology

To identify high-growth companies in a country, start from the official methodology: a growth rate over three years, in employees or turnover, from a minimum size. Revenue growth needs published turnover. Where a company publishes none, stated headcount and hiring carry the test. Funding and startup databases such as Crunchbase add capital events and a headquarters.

• Revenue growth. Turnover is the second measure the official definition allows, and Eurostat lists high-growth enterprises by turnover among the data collected on a voluntary basis. It comes from the accounts a company files or publishes. Apply the same threshold and period, and keep the turnover list beside the employee list.
• Hiring. The two hiring queries above place growth in a country for any company that publishes roles.
• Funding. Company Funding holds private capital raises reported in regulatory filings, each with its filing date, the amounts offered and sold and the issuer's principal place of business, and Company News dates the funding rounds companies announce. Capital dates new means, so keep it as a column beside the growth measure.
• Startup databases. Crunchbase is one. Its data page lists funding rounds and revenue ranges under financials, industry, legal status and employee count under firmographics, and heat and growth scores based on signals such as news activity, funding trends and user engagement. Its guide to advanced search describes location as the company headquarters and size as the number of employees, which answers the domicile test. Our guide to Crunchbase alternatives describes it at more length.

Fokals delivers the records that place a company: every posting with its locations, daily open postings by country and the markets each website declares, with stated headcount over time, joined on one company ID and delivered by REST API or as bulk files. The listed-company index covers equity listings in 79 countries, and the coverage page shows this week's figures.

Frequently asked questions

The queries and code on this page are examples to adapt. Test them in your own environment before you rely on them.

What this page says about the products it names was checked against their public documentation on 7 October 2026. Product and company names are trademarks of their owners. Fokals is not affiliated with them or endorsed by them.