Case study

A mid-size EU sportsbook's move to in-play micro-markets

Short-duration markets promised engagement and brought latency, integrity and player-protection problems with them. How one operator handled each.

A mid-size EU sportsbookSports betting5 min read
61% (from 48%)
In-play share of stakes
2.9s (from 6.5s)
Average bet acceptance delay
-63%
Bets voided for latency abuse
+22%
In-play users with an active limit

The problem

The operator in this study, anonymised at its request, holds licences in several EU member states and takes most of its revenue from football and tennis. By 2025 its pre-match margins were under sustained pressure from price comparison and from larger competitors, and in-play growth had stalled.

Its in-play product was built on conventional markets: match result, next goal, set winner. Settlement was slow, bet acceptance delays averaged more than six seconds, and markets were suspended for long stretches during attacks and break points, which is when customers most wanted to bet. The trading team also faced a persistent problem with courtsiding and fast-feed abuse, where a small number of accounts with quicker information than the book placed bets in the gap.

The operator decided to introduce micro-markets: short-duration markets such as the winner of the next game in tennis or the outcome of the next five minutes in football. It was aware that these products raise two concerns beyond the commercial ones. Small, discrete events are more exposed to manipulation, and rapid bet cycles are associated in the research literature with higher risk of harm.

What they did

  • Official low-latency data. The operator moved its in-play coverage to official data feeds with sub-second delivery and licensed a supplier's micro-market pricing models, applying its own margin and limits on top. Faster, more reliable data allowed acceptance delays to be shortened safely.
  • Automated trading controls. Suspension and reopening were automated from feed signals, with dynamic stake limits by market type and customer profile. Human traders moved to supervising exceptions.
  • A restricted market list. After consultation with its regulators and its integrity monitoring partner, the operator excluded markets on events that a single participant can easily determine with little sporting consequence, such as an individual point or the next throw-in. It offered micro-markets only on competitions above a defined tier, and reported alerts through the integrity body's standard channel.
  • Protection measures designed in. Micro-markets launched with lower default stake limits, a prompt to set an in-play deposit limit on first use, session reminders on by default, and an addition to the operator's risk model that scores rapid repeat betting. Accounts flagged by that model receive an interaction and, where indicated, a cooling-off period.

Results

Twelve months after launch, in-play betting accounted for 61 percent of stakes, up from 48 percent. Average acceptance delay fell from 6.5 to 2.9 seconds. Bets voided for latency abuse fell by 63 percent, which the trading team attributes mainly to the official feed.

The share of in-play users with an active deposit or loss limit rose by 22 percent, driven by the first-use prompt. The operator reported a higher rate of safer gambling interactions among micro-market users than among other in-play users, which it regards as the monitoring working as intended and not as a success metric. Three integrity alerts were raised over the period; none was confirmed.

What others can take from it

The commercial gain came chiefly from the data and trading infrastructure, which improved every in-play market, not only the new ones. Operators considering micro-markets should cost that investment first.

The restricted market list cost some turnover and avoided the categories that draw most regulatory concern. Several European regulators have limited or questioned markets on minor in-game events, and an operator that excludes them voluntarily is less exposed to a rule change.

Building limits and monitoring into the product before launch was cheaper than retrofitting them, and gave the operator evidence to show its regulators when they asked, as two of them did.

Latest stories

All stories

Worth a look

All guides