A UK-licensed operator running a £20-first-deposit promotion cut its live-chat first-response target from 15 seconds to 40 and saw deposit completion rise 23% among new registrants who opened chat before funding. The change ran for six weeks from 3 February 2025 across roughly 18,400 eligible sessions, with the uplift measured against a matched control group held on the old 15-second SLA. The headline number is not a claim that slow service sells. It is evidence that a chat agent appearing inside four seconds on a deposit page reads as a sales tactic, and that the resulting drop-off costs more than the queue time ever did.

What the test actually changed

The intervention was narrower than "make support worse". Three things moved together:

• First-response target went from 15 seconds to 40 for pre-deposit chat only. Post-deposit, withdrawal and complaint queues stayed on the fast SLA throughout.
• Agent greeting changed from a scripted upsell opener to a plain availability line — no bonus mentions in the first message.
• Queue position messaging was added: "You're 2nd in the queue" replaced a spinner with no estimate.

The 23% figure covers deposit completion within 30 minutes of opening chat, not conversion to a funded account overall. Registrants who never opened chat were excluded from both arms, which matters: the effect is concentrated in a self-selected group who were already hesitating.

Why the control group was necessary

February is a poor month to read raw conversion from. Cheltenham build-up spend lifts sportsbook registrations from mid-February, and the operator's own CRM calendar had a reload prompt going out on 17 February to a slice of the base. The matched control — same acquisition source, same device mix, same registration week — held those effects roughly constant. Without it, the 23% would have been indistinguishable from seasonal noise, and the sample of 18,400 split unevenly would have produced a confidence interval wide enough to swallow the result.

The mechanism: speed as a sales signal

The plausible explanation is not that customers enjoy waiting. It is that sub-20-second responses on a deposit screen pattern-match to the behaviour of a sales floor. Anyone who has used a UK bookmaker's chat widget has seen the same sequence: instant greeting, immediate bonus offer, a nudge toward the £20 qualifying deposit. Users have learned what that looks like.

A 40-second wait, by contrast, reads as a queue — a real one, with other customers in it. The queue-position message reinforces that. And the removal of the bonus opener from the first message means the agent's first contact is not a pitch. Deposit completion rose because the interaction stopped feeling like a funnel.

There is a second, duller mechanism worth taking seriously: agents on a 15-second SLA were closing chats early to protect their average. A longer target lets them actually answer the question — usually about payment method, pending status, or whether a Skrill deposit qualifies for the promo (it usually doesn't). Faster first response had been buying worse conversations.

Where the effect probably doesn't hold

Two caveats. First, this is pre-deposit chat only. A 40-second first response on a withdrawal query or a complaint is a different proposition entirely, and UK Gambling Commission expectations around complaints handling — eight weeks to a final response before ADR referral — leave little room for queue experimentation there. Second, the uplift is likely to decay. Novelty effects in chat UX are real, and a 40-second queue that becomes the norm may stop reading as "real" and start reading as "understaffed". The operator would need to re-run the test at six months to know whether the 23% is a level shift or a spike.

The £20 deposit problem specifically

The test design is inseparable from the amount. £20 sits at the awkward end of UK first-deposit behaviour: high enough that a hesitant registrant wants reassurance, low enough that the reassurance has to be cheap to deliver. A £10 threshold would have produced less pre-deposit chat volume; a £50 one would have pulled in a different, more decisive cohort and probably a smaller relative effect.

It is also the figure most UK welcome offers are built around, which is precisely why the chat interaction had become so contaminated. When every £20 first deposit is the trigger for a matched bonus, the chat agent's presence on that screen carries an obvious commercial meaning. Slowing the response and stripping the opener decoupled the two.

Operators reading this should be careful about the inference. The headline invites a lazy conclusion — "slow down support, make more money" — that would be wrong applied generally. What the test isolates is the value of not looking like you are selling at the exact moment a customer is deciding whether to trust you with card details.

What to measure if you run this yourself

The 23% is a completion-rate delta, and completion rate is a soft metric. Anyone replicating the test should instrument three things the original operator reportedly did not publish:

1. Deposit value distribution, not just completion. If the uplift came entirely from £20-and-nothing-more deposits, the revenue case is thinner than the conversion case.
2. Seven-day and 30-day retention of the uplift cohort. A customer reassured by a calm chat may still churn at the same rate once the bonus is gone.
3. Agent handle time and chat length. A 40-second first response that produces longer, better-resolved chats is a different operational change from one that produces the same chats, later.

There is also a compliance angle worth flagging. Chat transcripts are part of the customer interaction record, and any change to greeting scripts should go through the same approval as marketing copy. Removing a bonus mention from an opener is easy to get wrong in the other direction — an agent improvising a verbal offer is a worse outcome than a scripted one. Responsible gambling signposting in the chat widget should not be affected by response-time targets at all, and if it is, the test has overreached.

The open question is whether the effect is about time or about tone. The test bundled a slower response with a plainer greeting and a queue estimate, and the write-up attributes the gain to the SLA change because that is the number that moved. But a 15-second response with the plain greeting might have performed just as well — and if so, the lesson for UK operators is not about queue management at all. It is that the first line an agent types on a deposit screen should never be an offer.