How we did the study
2,357 companies 30 countries Website evidence captured 21–22 June 2026
Where the companies came from
We began with 11,119 listings in the EU-Startups directory. We matched each listing to its website, consolidated duplicate domains and used public website evidence to identify companies selling software to businesses.
The final cohort was the 2,357 companies with estimated organic traffic above zero. We joined other sources through each company's main website domain because names and URLs varied.
This directory-based cohort is not a census of European software companies.
European countries represented in the study
View the data behind this chart
| Country | Companies |
|---|---|
| United Kingdom | 479 |
| Germany | 317 |
| France | 224 |
| Spain | 174 |
| Netherlands | 136 |
| Sweden | 86 |
| Italy | 85 |
| Austria | 71 |
| Switzerland | 63 |
| Belgium | 63 |
| Denmark | 61 |
| Poland | 55 |
| Finland | 54 |
| Ireland | 50 |
| Portugal | 50 |
| Norway | 44 |
| Estonia | 43 |
| Romania | 36 |
| Lithuania | 33 |
| Czechia | 29 |
| Bulgaria | 29 |
| Cyprus | 27 |
| Greece | 25 |
| Latvia | 25 |
| Hungary | 22 |
| Slovakia | 20 |
| Slovenia | 18 |
| Luxembourg | 18 |
| Croatia | 11 |
| Malta | 9 |
What data we collected
Found and not found are both observed results. Unknown means the available evidence could not establish either.
Company classification. OpenAI We used public website text to establish what each company sold, who paid and its offering type. All 2,357 selected companies sold software to businesses.
Offering types recorded more than once 2,356 companies
- Software product
- 2,187 of 2,356
- Infrastructure or connectivity
- 111 of 2,356
- Marketplace or network
- 24 of 2,356
- Content, media or directory
- 21 of 2,356
- Professional service
- 13 of 2,356
Search evidence. We used Ahrefs for the study companies and DataForSEO for competitor domains.
- Ahrefs
- 2,357 study companies Cohort selection and organic-search estimates.
- DataForSEO
- 10,481 researched competitor domains Search visibility and supporting company-location evidence.
Website technology evidence. Wappalyzer OpenAI We checked Wappalyzer signals against saved website evidence and grouped the 783 technologies by what they do.
- Technology found: 2,345 (99.5%)
- Not found: 12 (0.5%)
Technology purpose 783 tools
- Website content and experience
- 225 of 783
- Website infrastructure
- 224 of 783
- Marketing and customer acquisition
- 93 of 783
- Measurement and insights
- 71 of 783
- Customer interaction and support
- 53 of 783
- Commerce and payments
- 32 of 783
Employee count for the study cohort. Apollo.io Apollo supplied estimated employee counts. Missing values stayed Unknown.
- Available: 2,121 (90.0%)
- Unknown: 236 (10.0%)
Products and buyers. OpenAI We identified 3,088 products across 2,277 companies and recorded their buyers and main jobs. Missing or unverifiable public evidence stayed Unknown.
- Complete: 2,277 (96.6%)
- Unknown: 80 (3.4%)
Products per company 2,277 companies
- One
- 1,869 of 2,277
- Two
- 222 of 2,277
- Three or more
- 186 of 2,277
Mistral AI We checked saved homepage text and controls for cookie banners. Completed checks without a banner were Not detected; incomplete evidence stayed Unknown.
- Detected: 1,179 (50.0%)
- Not detected: 1,090 (46.2%)
- Unknown: 88 (3.7%)
Google paid search. Ahrefs Ahrefs supplied paid-search estimates on 7 July 2026.
- Found: 502 (21.3%)
- Not found: 1,855 (78.7%)
Meta Ad Library. Meta We searched Meta Ad Library for globally active ads on 20–22 July 2026. Advertisers we could not match stayed Unknown.
- Found: 193 (8.2%)
- Not found: 1,378 (58.5%)
- Unknown: 786 (33.3%)
- Found: 283 (12.0%)
- Not found: 1,871 (79.4%)
- Unknown: 203 (8.6%)
We checked banner presence across all 2,357 saved homepages. We could read the controls on 878 banners.
Across 825 automated visits, we checked whether known tracking services were contacted before a visitor answered the banner and whether any new ones appeared after acceptance.
We compared each detected tool with the privacy policy from the same website.
The supplier review covered 361 of the 783 technologies: the tools that could receive customer or visitor data.
We checked official documentation for who controlled each tool, where data could be processed and how the supplier said it could use that data. If the documents did not say, we recorded Not documented.
Competitor research. OpenAI We researched direct competitors for each study company. Incomplete research stayed Unknown.
- Competitors found: 2,166 (91.9%)
- None met the criteria: 123 (5.2%)
- Unknown: 68 (2.9%)
How we compared competitors
Search overlap alone was not enough. We reviewed candidate products and did not count the study company as its own competitor.
Amie shows how the review worked for one company.
Europe
- Granola
- Jamie
- tl;dv
- MeetGeek
United States
- Fireflies.ai
- Otter.ai
- Fathom
The same process produced the European and US competitor groups across the study.
| Source | Output | Distinct competitors |
|---|---|---|
| 2,357 study companies | European competitors | 3,525 |
| 2,357 study companies | US competitors | 3,572 |
Each study company and its accepted direct competitors formed one market. We compared 2,035 markets with at least one European or US competitor across 20 product categories.
We calculated the European and US split within each market and gave every market equal weight. Companies elsewhere or with unknown locations stayed in the data but outside that percentage.
On average, competitors were almost evenly split between US and European companies, with large differences by product category.
US companies averaged 71.4% of competitors in AI platforms and agent building; European companies averaged 78.9% in real estate software.
European and United States competitor shares by product vertical
View the data behind this chart
| Vertical | Markets | European share among Europe and United States competitors | Unique European companies | European examples | United States share among Europe and United States competitors | Unique United States companies | United States examples | Other or unknown share among all competitors |
|---|---|---|---|---|---|---|---|---|
| AI platforms & agent building | 106 | 28.63% | 92 | Make, Lovable | 71.37% | 211 | OpenAI, Perplexity | 16.39% |
| Customer service AI & chatbots | 89 | 30.88% | 81 | YouCanBookMe, Doodle | 69.12% | 184 | Zendesk, Intercom | 19.07% |
| Learning & education | 77 | 38.45% | 103 | systeme.io, Seneca Learning | 61.55% | 172 | Udemy, Instructure | 26.16% |
| Advertising technology | 73 | 38.93% | 130 | Adjust, Criteo | 61.07% | 163 | Google, PubMatic | 20.98% |
| Sales outreach & marketing automation | 119 | 39.11% | 233 | Typeform, Brevo | 60.89% | 313 | Mailchimp, HubSpot | 25.91% |
| Hosting & cloud infrastructure | 73 | 39.6% | 137 | TeamViewer, UptimeRobot | 60.4% | 225 | Cloudflare, AWS | 16.4% |
| Identity, fraud & financial crime | 131 | 39.89% | 198 | Whistlelink, Yoti | 60.11% | 324 | OneTrust, Tenable | 13.98% |
| Data analytics & business intelligence | 148 | 42.43% | 245 | TradingView, LSEG | 57.57% | 341 | Microsoft, Qualtrics | 18.95% |
| AI content & media generation | 117 | 42.95% | 191 | Prismic, Framer | 57.05% | 277 | WordPress.com, Adobe | 30.31% |
| Healthcare & clinical software | 73 | 45.11% | 153 | Elsevier, Clarivate | 54.89% | 137 | Veeva, athenahealth | 19.79% |
| Recruitment & HR | 105 | 45.82% | 205 | Teamtailor, Personio | 54.18% | 192 | LinkedIn, Indeed | 13.81% |
| E-commerce & product data | 90 | 45.94% | 198 | Shopware, Trusted Shops | 54.06% | 188 | BigCommerce, WooCommerce | 17.48% |
| Contracts & document workflows | 147 | 49.77% | 319 | SAP, Assently | 50.23% | 339 | Jotform, Dropbox Sign | 15.34% |
| Industrial operations & asset monitoring | 148 | 52.51% | 322 | ABB, Schneider Electric | 47.49% | 327 | Datadog, Honeywell | 20.43% |
| Work & business management | 205 | 52.7% | 450 | Odoo, Cargoclix | 47.3% | 447 | GitHub, Oracle | 20.78% |
| Hospitality & booking systems | 86 | 54.76% | 241 | GetYourGuide, Eventim | 45.24% | 158 | Yelp, Toast | 24.39% |
| Crypto & digital assets | 76 | 58.17% | 177 | Revolut, Coinranking | 41.83% | 131 | Chainalysis, Coinbase | 19.76% |
| Payment processing & orchestration | 75 | 64.47% | 188 | Klarna, NOWPayments | 35.53% | 78 | PayPal, Stripe | 14.86% |
| Sustainability & carbon accounting | 59 | 69.29% | 114 | Siemens, EcoVadis | 30.71% | 46 | S&P Global, IBM | 15.01% |
| Real estate software | 38 | 78.85% | 113 | Zoopla, Rightmove | 21.15% | 30 | Esri, Yardi | 16.06% |
The employee comparison used Apollo estimates for 5,238 European companies and 3,511 US companies. The European group combined the study cohort with accepted competitors and counted overlaps once. Missing values were not filled in.
The US median was 79 employees, compared with 21 in Europe.
Employee-count distributions for European and American companies
View the data behind this chart
| Region | Company | Employee band | Reported employees |
|---|---|---|---|
| Europe | Hyperping | 0-1 | 1 |
| Europe | Numista | 2-5 | 5 |
| Europe | Brandfetch | 6-14 | 7 |
| Europe | Publer | 15-38 | 23 |
| Europe | Flightradar24 | 39-99 | 91 |
| Europe | HappyScribe | 100-250 | 210 |
| Europe | Kahoot! | 251-630 | 450 |
| Europe | SoundCloud | 631-1584 | 1100 |
| Europe | Klarna | 1585-3980 | 3800 |
| Europe | Wise | 3981-plus | 6000 |
| United States | RSS.com | 0-1 | 1 |
| United States | SavvyCal | 2-5 | 2 |
| United States | Gimkit | 6-14 | 8 |
| United States | Kapwing | 15-38 | 34 |
| United States | Blooket | 39-99 | 50 |
| United States | FlightAware | 100-250 | 130 |
| United States | Quizlet | 251-630 | 500 |
| United States | ZipRecruiter | 631-1584 | 1000 |
| United States | Khan Academy | 1585-3980 | 2300 |
| United States | 3981-plus | 188000 |
Public pages and ad libraries can change after observation. A visible tag does not establish whether or when it ran. The browser checks covered 825 automated visits.
Ahrefs and Apollo provide estimates. Advertising sources cover different observation windows. Supplier silence is not proof of a prohibition.
The study reports what we observed, not what caused it. It does not make legal-compliance judgments.