A seven-day expiry on “free bet” tokens — the model Bet365 has used for years — cuts the realistic return from a matched-betting cycle by roughly 14% compared with the same offer settled on a 30-day clock. That figure comes from working the maths on a £10 qualifying bet at 2.0 back and 2.1 lay, not from any operator's marketing sheet, and it holds across most Premier League weekend pricing where the qualifying loss sits between 40p and 70p.

The 14% is not a headline rate. It is the erosion of net profit per offer once you price in the probability that a token goes unused, plus the forced compression of your lay-matching window. Matched bettors who treat every free bet as a fixed-value coupon are quietly losing a slice of edge on the short-dated ones.

Why seven days changes the arithmetic

A free bet token has no cash value. Its value is the stake-not-returned portion of an each-way or win-only lay, typically 70–80% of face value if you lay it correctly at odds around 4.0 to 6.0. On a £10 token laid at 4.5, the standard extraction is about £7.20 to £7.60 after commission at 2% on the exchange.

That number assumes you place the token. The expiry date is what governs whether you place it at all.

With a 30-day window, a matched bettor running a modest book of six to ten accounts can wait for a genuine price discrepancy — a horse drifting on the exchange, a tennis favourite shortening after a retirement, a niche market where the back price lags the lay by two ticks. With seven days, you take the first acceptable arb, which is usually a thinner one. The difference between a 78% extraction and a 68% extraction on a £10 token is £1.00. Run that across 40 tokens a month and you are down £40 before you have accounted for the ones you simply miss.

The 14% figure comes from modelling the short window as a constraint on price selection. If your average extraction on a long-dated token is 76%, and the compressed window forces you onto prices that yield 65–66%, the trim is 13–14%. It is not a fixed penalty; it scales with how many tokens you are juggling.

The unused-token problem

Expiry also creates a hard failure mode. A 30-day token that goes unused is a rare event — you had four weekends to find a home for it. A seven-day token placed on a Tuesday, with a Saturday fixture list that turns out to be thin on liquidity, can easily lapse.

Assume a 6% lapse rate on seven-day tokens versus 1.5% on 30-day tokens. On 40 tokens a month at £10 face, that is 2.4 lost tokens versus 0.6 — a shortfall of 1.8 tokens, or roughly £13 in foregone extraction at 72%. Stack that against the £40 from compressed pricing and the total monthly drag on a 40-token book is in the region of £53. On a gross extraction of about £290, that is an 18% hit. The 14% is the conservative end.

What the operator gets out of it

Short expiry is not an accident of accounting. A free bet token is a liability on the book until it is settled, and the operator carries the risk that it lands on a long-priced winner. A seven-day window does three things at once.

First, it forces volume through the sportsbook in a narrow band, which flatters weekly active user numbers in reporting periods. Second, it pushes customers toward mainstream markets — Premier League match odds, Champions League outrights — where the operator's margin is already protected and the exchange liquidity is deepest, so the lay side is easy but the back side is priced sharp. Third, it reduces the tail risk of a token landing on a 33/1 shot three weeks later during a quiet period when the trading desk has less capacity to hedge.

The customer-facing justification is always "use it or lose it, so you don't forget." That is true as far as it goes. It is also true that a 30-day token is harder to forget than a seven-day one precisely because it does not demand immediate attention.

Where Bet365 sits

Bet365 has run seven-day expiry on its Bet Credits and free bet tokens for years, with the clock typically starting at settlement of the qualifying bet rather than at placement. That distinction matters. If your qualifying bet settles on a Sunday evening, you have until the following Sunday evening — not the following Monday. A token issued at 21:47 on a Sunday expires at 21:47 the next Sunday, and if you are a salaried worker who does their matched betting on a Saturday morning, you have effectively lost a day.

Other UK operators have moved in the opposite direction. Some run 14-day windows on welcome offers and 30 days on reload tokens. A few run seven days on everything. The spread is wide enough that any matched bettor running more than a handful of accounts needs a tracking sheet with expiry timestamps, not just dates.

The practical adjustment

If you are running a book with a mix of seven-day and 30-day tokens, the rational response is not to abandon the short-dated ones — they still carry positive expected value — but to reorder your queue.

Place seven-day tokens first, and accept a lower extraction threshold. If your normal rule is "lay at 4.0 or better with a 70% minimum extraction," drop the floor to 62% for tokens inside 48 hours of expiry. A 62% extraction on a £10 token is £6.20; a lapsed token is £0.00. The maths is not close.

For tokens with more than five days left, hold them for a genuine price. The opportunity cost of holding is zero until you are inside the final 48 hours, at which point the cost of holding rises sharply.

Bankroll and exchange liquidity

Short expiry interacts badly with thin exchange liquidity. If you are trying to lay a token on a Tuesday afternoon on a market with £200 matched, you will move the price and eat into your extraction. The seven-day window pushes you into exactly those hours.

The workaround is to concentrate short-dated tokens on markets with deep liquidity — Premier League, Champions League, ATP and WTA main draws, NFL in season — and accept a slightly worse price rather than a worse fill. A 2% worse price on a liquid market beats a 6% worse price on a thin one.

What to watch

The 14% trim is a snapshot of current pricing and current exchange commission. It is not a law. If exchange commission falls — and some exchanges have run promotional 0% windows on football — the extraction on short-dated tokens improves, and the gap between seven-day and 30-day narrows. If commission rises, or if operators shorten expiry further to five days, the trim widens.

The open question is whether the market corrects. Matched bettors are price-sensitive, and a 14% edge reduction is enough to move volume. If enough customers route their qualifying bets toward operators offering 14-day or 30-day tokens, the seven-day standard becomes a competitive disadvantage rather than a cost-saving. The counter-argument is that most customers are not matched bettors, do not calculate extraction, and will accept whatever window they are given. On that reading, seven days is here to stay, and the 14% is simply the cost of doing business with the biggest book in the UK.

For anyone running a book, the number to track is your own lapse rate. If it is above 5% on seven-day tokens, the problem is not the operator's expiry policy — it is your queue discipline. Gamble responsibly, and treat any token you cannot place inside 48 hours as already expired.