TL;DR: This piece started as a happy accident. I set out to check whether a small chemical company called “Yang Li” on Qian Nong 1st Road in Xiaoshan was undercutting business, and discovered that the moat in chemical manufacturing — hazardous-material licenses, process know-how, and upstream supply lock-ups — is completely impenetrable to outsiders. But the idea that emerged from that research, a “supply-chain data product,” translates perfectly to football scouting data — a niche small enough for a one-person company. Below is a full breakdown of the five major players in this space, and the blind spot they all share: the amateur and lower-league market, ignored globally.

Origin: From Chemical Hidden Champions to Scouting Data

No. 10 Qian Nong 1st Road, Xiaoshan District, Hangzhou — Hangzhou Yangli Petrochemical Co., Ltd. (also known externally as Hangzhou Yangli Industrial Co., Ltd.) — founded in 1996, registered capital of RMB 27 million. The three high-rise buildings are factory structures (distillation columns / tank farms / workshops), not office towers.

Its core product is a fine chemical raw material — dicyclopentadiene (DCPD) — one of China’s larger refined DCPD producers, with an annual capacity of roughly 5,000 tonnes at 99%+ purity. Downstream it extends into tetrahydrodicyclopentadiene and adamantane (DCPD deep-processed derivatives).

The profit model is straightforward: upstream, DCPD feedstock is separated from the C5 fraction produced as a by-product of refinery cracking; midstream, refined purification (the core barrier is the distillation process that achieves 99%+ purity) — a know-how forged over years of tuning, not written down anywhere; downstream, it’s sold to pharmaceutical intermediate factories (adamantane is a precursor for adamantane amine / rimantadine and other Alzheimer’s drugs), photoresist manufacturers (adamantane derivatives used as scaffolds in 193 nm ArF photoresists), and speciality materials producers.

But this path is a dead end for ordinary people:

• Hazardous materials production license — environmental impact assessment + safety assessment + production safety permit; just the approval process takes 2–3 years
• Process know-how — 99%+ purity is the result of years of distillation parameter tuning; it’s not on paper
• Upstream supply lock-in — feedstock is refinery by-product; you need a long-term supply agreement with a refinery, and newcomers can’t secure allocation

The money locked behind these three barriers is observable but unreplicable. Physical “buy low, sell high” of chemical products also requires hazardous materials warehousing licenses and significant capital — cash-strapped, unlicensed individuals can’t touch it.

What’s worth carrying away is the idea triggered during the research: use AI to aggregate fragmented public information into supply-chain signals and sell the output as a data product to industry players. That capability only makes sense when transplanted onto a track that fits your own strengths — rather than being locked to a niche chemical product with only 20 players in the world.

And football scouting data is exactly that kind of track.

Landscape of the Five Main Competitors

① Wyscout 🇮🇹 Italy — the de facto standard for football scouting video

• Founded: 2004, Genoa (later moved to Chiavari)
• Founders: Matteo Campodonico, Simone Falzetti, Pier Maria Saltamacchia
• Revenue: €13 million (2019, per Wikipedia)
• Headcount: ~80 (2019)
• Status: Acquired by US-based Hudl

The business is a football video analysis platform plus a player database, supporting scouting, match analysis, and transfer workflows. It is the de facto standard for football scouting video — virtually every professional club uses its video library for player screening.

The profit model is B2B SaaS subscription: clubs, football associations, and agents pay to access the video library and data tools. There is also the Wyscout Forum — an offline transfer-matching trade show that extends the video platform into a transactional marketplace.

What this means for a one-person company: a ceiling to reference. A company with 80 people and €13M in annual revenue is not something a single person can replicate. But its “video → scouting” path is the best template for entering the scouting market through video analysis.

② SciSports 🇳🇱 Netherlands — the prototype closest to the vision

• Founded: Circa 2015–2016 (industry common knowledge; needs further verification)
• HQ: Zeist, Netherlands
• Revenue: Not publicly disclosed
• Headcount: ~30–50 (estimated)

The business is AI-driven player data analysis — scouting reports, player profiles. Data sources include its proprietary BallJames optical tracking system plus aggregated public data.

The profit model is B2B SaaS, offered in two product lines: Professional and Youth (verified via their website). Clients span clubs, football associations, player agents, and academy programs.

This is the single most important finding in this research: SciSports’ product architecture — AI analysis + scouting reports + an academy-focused line — is essentially the standard template for “AI-powered scouting data.” But it focuses exclusively on the professional and elite youth levels — it does not cover amateur or lower-league football. What a one-person company should build is, in essence, SciSports’ “Chinese amateur/lower-league” version.

③ InStat 🇧🇾 Belarus — pioneer of event-level data reports

• Founded: Circa 2007 (industry common knowledge)
• HQ: Brest (Belarus), later Ireland
• Revenue: Not disclosed
• Headcount: ~200–500 (estimated)

The business is match tactical analysis, generating event-level data reports for every match — player ratings, tactical heatmaps, action sequences. It covers global leagues and pioneered the post-match analysis model.

The profit model is B2B service fees: clubs pay for post-match analysis reports and database subscriptions.

Technical lineage: The video2script tool (match video → structured event JSON) in my arsenal follows the same technical route as InStat’s event-level analysis. The difference is that InStat covers professional leagues using a manual annotation team, whereas a one-person company can use AI to compress that cost to a single operator.

④ StatsBomb 🇬🇧 UK — premium event data licensor

• Founded: 2018, London
• Revenue: Not disclosed
• Headcount: ~30–50 (estimated)
• Status: Acquired by Hudl, now called Hudl StatsBomb

The business is advanced football event data — models such as xT (Expected Threat), VAEP (Valuable Actions by Effective Probability), and others — licensed to clubs, media, and betting companies. Data quality is top-tier and the professional barrier is high.

The profit model is B2B data licensing + SaaS subscription (Pro Suite).

What this means: a reference for the data licensing model. But it serves professional clubs that need top-tier event data. What a one-person company should build is “amateur-grade, low-cost data” — not a head-on competition on data precision.

⑤ Stats Perform 🇬🇧 UK — the ceiling / leviathan of the track

• Founded: 1981 (as STATS, US) → merged with Perform in 2019
• HQ: London
• Owner: Vista Equity Partners (private equity)
• Revenue: Hundreds of millions of USD (estimated)
• Headcount: Several thousand, global

The business is a sports AI + data oligopoly covering multiple sports, spanning data collection, predictive analytics, media, and betting. It began investing in AI in 2015.

The profit model is B2B enterprise: subscriptions from clubs, leagues, media, and betting companies.

What this means: proof of the industry ceiling. It has no direct competitive relationship with a one-person company, but its scale alongside StatsBomb and Wyscout demonstrates that this is a real track — not a pseudo-demand.

Reading the Board: Where Is the Vacuum?

Placing these five side by side, a clear fault line emerges:

Every player focuses on the “professional league” market. Wyscout’s video library is professional matches. SciSports’ Professional + Youth lines target professional clubs and elite academies. InStat covers global professional leagues. StatsBomb sells top-tier event data to professional teams. From the Premier League down to China League Two, the professional scouting data space is carved up completely among these five (and the Hudl ecosystem behind them).

But amateur football, school football, and China League 3 and below — this is a collective blind spot for the global giants. Not because they can’t see it, but because they look down on it: the market is too fragmented, the per-match value is too low, and covering it with manual annotation teams is uneconomical. This is precisely the gap that AI-assisted solo production can fill: using AI to convert video into structured events, a single person can produce the post-match analysis reports that previously required an annotation team.

The second takeaway is the consolidation trend among giants: both Wyscout and StatsBomb have been acquired by Hudl. This shows that pure-data small companies are being absorbed into video platforms — platforms with distribution channels combined with data companies’ analysis capabilities are more competitive than standalone data firms. For a one-person company, this means don’t build a “pure data company.” Instead, build a composite of “data + methodology + local relationships” that makes you impossible to simply acquire.

The Real Moat Isn’t “Nobody’s Doing It”

The biggest lesson from the chemical research is: “few players” does not equal “can make money.” Fewer than 20 adamantane players in China — but the customer base is too small, information asymmetry is too low, and I’m not an insider. Narrow and poor — a dead end. Three conditions must be met simultaneously:

1. Niche but with sufficient per-customer willingness to pay (narrow and fat, not narrow and poor)
2. Aligns with your unique capabilities (otherwise others can enter too, and “few players” is only temporary)
3. Coverable by one person’s output capacity

The football scouting data track satisfies all three for a one-person company:

• Capability fit — AI data analysis skills, Position Play methodology (referencing the La Masia model), and the event-structuring technology already accumulated in video2script; the combination of the three is a composite advantage no one else has
• Solo production viable — AI-assisted video analysis → structured reports; a single person can produce enough
• Vacuum zone — amateur and lower-league football is not covered by global giants

And the real moat is not the temporary state of “nobody in China is doing this,” but rather the knowledge barrier that Spanish Position Play methodology is unfamiliar to Chinese practitioners. Giants understand AI but don’t understand football tactics philosophy; Chinese football people understand the game but not data — a one-person company positioned at that intersection is fundamentally uncopyable.

Go-to-Market: Don’t Start by Selling Reports

To be pragmatic, the poverty of lower-league clubs is a real problem. So don’t start by selling reports directly. Instead, use video2script to provide free opponent analysis for 1–2 China League Two or China League 3 clubs, build case studies and relationships, and then move to paid engagements.

Position Play knowledge is scarce to those who understand it — but to cash-strapped clubs, you need to prove value before asking for payment. Start with free cases to earn trust, use trust to acquire data assets, and use data assets to convert paying clients. This is the reverse application of the lesson from the chemical research: first make yourself an irreplaceable observer, then talk about business.

Data Sources and Honest Disclosure

Among the competitive intelligence in this piece, the following are verifiable hard data from public sources:

• Wyscout: Founded 2004, revenue €13M (2019), headcount ~80, acquired by Hudl — all cited from Wikipedia
• Stats Perform: Founded 1981 (as STATS), merged 2019, owner Vista Equity Partners, CEO Carl Mergele — from Wikipedia
• SciSports: Product structure (Professional/Youth dual lines), client segmentation (clubs/associations/agents/academies) — from the “About” page on their website
• StatsBomb: Now Hudl StatsBomb — confirmed via their website

The following are industry-estimate figures that could not be verified against public financials — use caution when citing:

• SciSports founding year (~2015–2016), headcount (~30–50)
• InStat founding year (~2007), headcount (~200–500)
• StatsBomb headcount (~30–50)
• Stats Perform revenue (hundreds of millions of USD tier)

The recollection of “a French scouting company” could not be confirmed after verification — the closest matches are the Netherlands-based SciSports or UK-based Stats Perform (which has French operations). France’s sports data space is dominated by Sportradar (Swiss-rooted, bias toward betting data), not the scouting vertical.

The next research step is to dig deeper into SciSports’ fundraising history and actual revenue to validate the track’s profitability.