Responsible Play

Monitor at-risk buying behaviour, review safer-gambling alerts, manage self-exclusions ("take a break") on behalf of your customers, and keep an eye on customer-set spending limits (which only customers themselves can change). This page covers running scans, AI analysis, resolving alerts, setting or removing an exclusion for a customer who phones in, exporting the exclusion register for compliance, and what customers can do for themselves in their own account.

Where to find it: Admin → Responsible Play (/admin/responsible-play)

The page has two tabs:

• Alerts — behavioural alerts raised by the platform's monitoring scans. The tab label shows a badge with the number of pending alerts.
• Users — every customer with a responsible-play setting: self-excluded customers and customers with a spending limit.

Understand the Alerts tab

Four stat cards summarise the current position:

• Pending Alerts — alerts awaiting review. Shows in red when there are any, green when the queue is clear. When several pending alerts belong to the same customers, a sublabel such as "3 unique users" clarifies how many actual people are involved.
• High Risk and Medium Risk — pending alerts by risk level.
• Alerts This Week — how many alerts were raised in the last seven days, resolved or not.

Alert types you will see

The platform raises these alert types automatically:

• Spending Limit 80% / 90% — a customer with a self-set monthly limit has reached that share of it. The row subtitle shows the actual figures, e.g. "£45.00 of £50.00".

• Spending Limit Reached — the customer has reached or passed 100% of their limit. This is the name shown in the Alert Details panel and the Review Alert window. In the alerts table, the row title instead shows the severity as a multiplier where the figures are available, e.g. "Spending Limit Exceeded — 2.50× limit", so a customer at 250% stands out from one at 101%.

  Because checkout blocks any card payment that would take a customer past their limit, most of these alerts sit at exactly 100%: the customer has used their budget up, not overshot it. A row well above 100% means one of two specific things, and both are worth a look:

  • The customer set or lowered their limit to below what they had already spent this month. Open the alert and look for the yellow "Limit set after spending had already happened" banner, which names the date and the amount already spent. Someone cutting their own budget after a heavy run is telling you something.
  • Several orders completed at the same moment and slipped past the check together. This shows as a small overshoot with no such banner.

• Rapid Purchases — an unusually fast run of orders.

• Failed Payments — repeated failed payment attempts.

• Late Night Activity — purchasing at unsociable hours.

• Spending Spike — this month's spend is a multiple of last month's; the title shows the ratio, e.g. "Spending Spike — 3.1× last month".

• Limit Removed Then Spent — the customer removed their spending limit and then spent.

• Possible Exclusion Evasion — a self-excluded customer appears to match one or more active accounts on shared details. See Investigate a possible exclusion evasion below.

Each alert carries a Risk level (High, Medium, Low) and a Status: Pending, Reviewed, Actioned, or Dismissed.

Grouped ("re-raised") alerts

When repeated scans flag the same customer for the same reason, the pending view groups them into a single row rather than filling your queue with near-identical entries. The Activity column shows "re-raised 3× since 12 Jun" on grouped rows, and the alert detail panel repeats this with a note that resolving the row clears every underlying occurrence at once — you do not need to review each re-fire individually. Grouping only applies while you're viewing pending alerts; historical views (Reviewed, Actioned, Dismissed) list each alert individually.

Run a scan

Alerts are generated by the platform's behaviour scans. To check all customers right now rather than waiting:

1. On the Alerts tab, click Run Scan.
2. The scan runs immediately — there are no parameters to choose.
3. When it finishes, the alerts list and stat cards refresh with anything new.

Run AI analysis

AI analysis writes a plain-English behavioural summary onto pending alerts, giving you context (patterns, history, severity) without piecing it together yourself.

1. On the Alerts tab, click Run AI Analysis.
2. When it completes, analysed alerts show a blue AI badge in the AI column.
3. Open an analysed alert's details to read the AI Behavioral Analysis section — a written assessment of the customer's behaviour.

To see only analysed (or only unanalysed) alerts, use the AI Analysis filter — "AI Analyzed Only" or "Not Analyzed".

Filter the alerts list

Above the table, three dropdowns control what you see:

• Status — defaults to Pending (your work queue). Switch to All Statuses, Reviewed, Actioned, or Dismissed to see history.
• Risk Level — All Levels, High, Medium, or Low.
• AI Analysis — All Alerts, AI Analyzed Only, or Not Analyzed.

Review an alert

For every pending alert you should decide an outcome and record it — this builds your audit trail.

1. On the Alerts tab, click the eye icon on the row to open the Alert Details panel. It shows the customer, alert type, risk badge, status, when it was raised, the raw details behind the alert, and the AI Behavioral Analysis if one has been run.
2. Click Review Alert (or use the green tick icon directly on the row to skip the detail panel).
3. In the Review Alert window, choose a Resolution:
   • Reviewed (No action needed) — you looked into it and the behaviour is acceptable.
   • Actioned (Took action) — you did something: contacted the customer, set an exclusion, restricted the account, etc.
   • Dismissed (False positive) — the alert was wrong or irrelevant.
4. Optionally describe what you did in Action Taken — strongly recommended when choosing Actioned.
5. Click Submit Review.

The alert leaves the pending queue and the stat cards update. If the row was a grouped alert, all of its underlying occurrences are resolved together. Once resolved, the detail panel permanently shows Reviewed By (who resolved it and when) and the Action note — your audit trail if a regulator or auditor asks.

Reviewed vs Actioned vs Dismissed — which to pick

• Pick Reviewed when you checked and no intervention was needed (e.g. a customer nearing their limit, but well within their means and their own cap is doing its job).
• Pick Actioned when you intervened in any way, and say what you did in Action Taken.
• Pick Dismissed only for genuine false positives, so your reporting reflects real risk accurately.

Review or dismiss many alerts at once

After a scan you may have a batch of low-value alerts to clear:

1. Tick the checkbox on each row you want to act on (or the header checkbox to select the page).
2. The page header switches to a bulk bar showing "N selected" with Mark Reviewed, Dismiss, and Clear Selection buttons.
3. Click Mark Reviewed or Dismiss to resolve all selected pending alerts in one go.

Bulk actions only apply to pending alerts; selected rows that are already resolved are ignored. Grouped rows resolve all of their underlying occurrences, just like a single review.

Read the warning banners in alert details

Two banners can appear in the Alert Details panel and change how you should read the alert:

• "Limit set after spending had already happened" (yellow) — on spending-limit alerts, this tells you the customer set their limit after already spending that month, e.g. "User set this limit on 12 Jun after already spending £120.00 this month." The "over limit" figure includes spend they committed to before capping themselves — so this is often a customer taking control, not blowing through a cap. Factor that into your resolution.
• "This customer appears to have other accounts" (red) — appears on Possible Exclusion Evasion alerts. See the next section.

Investigate a possible exclusion evasion

A Possible Exclusion Evasion alert means a self-excluded customer matches one or more active accounts on shared details — potentially someone circumventing their own break.

1. Open the alert's details (eye icon).
2. Read the red banner: "This customer appears to have other accounts."
3. Click View linked accounts. This takes you to Users → Account Flags with the matching group of accounts pre-selected, so you can compare them side by side before deciding.
4. Decide your action — for example, extending the exclusion to the linked account or locking it — then come back and resolve the alert as Actioned with a note, or Dismissed if the match was coincidental (e.g. family members in the same household).

See Account Flags for how linked-account matching works and what you can do from that screen, and User Management for account-level actions.

Set a self-exclusion for a customer

When a customer phones or emails asking to be excluded (or you decide to exclude them following an alert), you can apply the exclusion on their behalf:

1. Go to the Users tab and click Set User Exclusion.
2. In the Select User window, search by email or name (at least two characters). Customers who are already excluded show an "Excluded" badge in the results.
3. Click the customer. If they are already excluded, you'll be asked whether you want to Modify the existing exclusion or Remove it.
4. In the Set Self-Exclusion window, choose a Duration: 24 Hours, 7 Days, 30 Days, or Permanent. These are fixed durations — you cannot type a custom end date.
5. Optionally record a Reason (e.g. "User requested via support ticket") — recommended, as it documents why the exclusion exists.
6. Click Set Exclusion.

The exclusion takes effect immediately: the customer can no longer add items to their basket or place orders, and their own account area shows the exclusion type and end date. If a customer already excluded asks for a longer break, use the same flow — picking them from the search and choosing Modify lets you set a new duration.

You can also start from the table: the pencil icon on any row opens the same Set Self-Exclusion window for that customer.

Remove a self-exclusion early

Customers cannot lift their own exclusion — the removal control only exists on your side, deliberately. If you decide to remove one:

1. On the Users tab, find the excluded customer (filter Status to Self-Excluded if needed).
2. Click the red bin icon (Remove self-exclusion) on their row. The icon only appears for customers who are currently excluded.
3. Read the warning: "This will immediately allow the user to place orders again." There is no cooling-off delay — removal is instant.
4. Enter a Reason — this is required, and forms your record of why the exclusion was lifted.
5. Click Remove Exclusion.

Treat early removal seriously, especially for permanent exclusions: the customer chose (or you imposed) the break for a reason. Note that removing a spending limit and then spending triggers its own alert type (Limit Removed Then Spent), and exclusion evasion monitoring continues regardless.

Find customers with exclusions or spending limits

The Users tab lists every customer with a responsible-play setting. Stat cards at the top show Currently Excluded Users, Permanent Exclusions, Users With Spending Limits, and High Spenders (>75%) — customers who have used 75% or more of their monthly limit (a customer at exactly 75% is counted).

Filter and search with:

• Status — All Users with Settings, Self-Excluded, or Has Spending Limit.
• Exclusion Type — 24 Hours, 7 Days, 30 Days, or Permanent (unavailable while filtering to spending limits).
• Search — free-text search by name or email.

Each row shows the customer, their current status (Excluded or Active), the setting type, Exclusion Until (the end date, where applicable), their Spending Limit, and Monthly Spend.

Spot customers approaching their limit

The Monthly Spend column shows the customer's spend this calendar month with a percentage badge against their limit, colour-coded: green below 75%, amber from 75%, red from 90%. Combine the High Spenders (>75%) stat card with the Has Spending Limit filter to check on customers nearing their cap — these are the same customers who will shortly generate Spending Limit 80/90% alerts.

Note that spending limits are set by customers themselves, are always monthly, and reset at the start of each calendar month. You cannot set, change, or remove a customer's spending limit from the admin panel — this screen shows you limits and usage so you can monitor them.

Export the exclusion register (CSV)

For compliance reporting, audits, or record keeping:

1. On the Users tab, set the filters to the population you want (e.g. Status: Self-Excluded).
2. Click Export CSV.

The file downloads immediately, named with today's date, and contains the rows currently shown in the table with these columns: Email, First Name, Last Name, Status, Exclusion Type, Exclusion Until, Spending Limit, Monthly Spend, and Spend %. Each export contains only the page currently on screen (up to 20 rows) — moving between pages does not build up a bigger file. If the list spans several pages, either narrow your filters until everything fits on one page, or export each page in turn and combine the files.

What customers can do themselves

Your customers manage their own responsible-play tools from Account → Responsible Play on your site. Knowing what they see helps you handle support queries:

• Status banner — shows "Active Protections" (with their limit and/or exclusion end date) or "No Active Restrictions".
• Monthly spending limit — customers can set a limit using preset amounts (£25, £50, £100, £200) or any custom amount from £1. They see their card spend this month, remaining budget, and a colour-coded progress bar. Limits are monthly only, reset at the start of each calendar month, and take effect immediately. Card payments that would exceed the limit are blocked — at checkout the customer sees a "Spending Limit Reached" notice with a link to manage their limit.
• What the limit counts — the limit is a budget on money the customer pays you by card, so it counts the amount actually charged to their card. Anything paid for with site credit (prize winnings, refunds, referral rewards) does not count towards it, and a customer with enough credit can keep entering after reaching their limit. Expect the occasional support query about this; the customer's own screen now says so, but it is worth knowing. If you need a customer to stop entering altogether, a Take a Break exclusion is the control that does that, and it applies no matter how they pay.
• Limit increase/removal cooldown — lowering a limit applies immediately, but increasing or removing it is held behind a cooling-off delay; the customer sees exactly when they'll be able to make the change. This is deliberate friction, and you cannot override it.
• Reality check reminders — customers can switch on periodic pop-up reminders (every 15 minutes, 30 minutes, 1 hour, or 2 hours) showing their session time and spending summary, and can preview or disable them at any time.
• Take a Break (self-exclusion) — customers can self-exclude for 24 hours, 7 days, 30 days, or permanently. They are warned before confirming that they won't be able to add to basket or place orders and that the exclusion cannot be cancelled early. Their basket is cleared when the exclusion starts. Timed exclusions end automatically; permanent ones do not.
• Your site also has a public responsible-play guide page with support resources, linked from the account area.

Customer-set and admin-set exclusions are the same thing under the hood: both appear on your Users tab, both block purchasing identically, and both can only be removed early by you.

Troubleshooting

A customer says they can't add anything to their basket or check out

Check the Users tab for their account. If their status is Excluded, a self-exclusion is active — either they set it themselves or an operator did. The admin panel does not show which, or the reason recorded when it was set, so check your own support records if that matters. They cannot lift it themselves; if you judge removal appropriate, follow Remove a self-exclusion early. If they are not excluded but see "Spending Limit Reached" at checkout, they have hit their own monthly limit — it resets at the start of the next calendar month, and they can only raise or remove it after their cooling-off delay.

A customer wants their exclusion lifted early

Only you can do this — the customer-facing page states the exclusion can't be cancelled early. Use the bin icon flow on the Users tab, record a clear reason, and remember the effect is immediate. Consider the original reason for the exclusion before agreeing.

A customer asks you to raise their spending limit

You can't — spending limits are entirely customer-controlled. Point them to Account → Responsible Play on your site. If they recently set or lowered the limit, they'll see a cooldown notice telling them when they can increase or remove it.

The Pending Alerts number doesn't match the rows in the table

Pending alerts for the same customer and reason are grouped into one row. The stat card counts individual alerts, and its sublabel (e.g. "2 unique users") tells you how many people are involved. Resolving a grouped row clears all of its occurrences, so the pending count can drop by more than one per review.

An alert says the limit was massively exceeded but the customer seems fine

Open the alert details and look for the yellow "Limit set after spending had already happened" banner. If present, the customer set their cap after spending that month, so the excess reflects pre-limit spend — usually a sign of someone taking control rather than losing it. Resolve as Reviewed with a note if no action is needed.

The same customer keeps getting re-flagged

That's expected while the underlying situation persists — each scan re-raises it, and the pending view rolls the repeats into one row ("re-raised N× since …"). Resolve the row once; if it reappears after future scans, the behaviour is continuing and may warrant escalating from Reviewed to Actioned.

Related pages

• Account Flags — linked-account detection used by exclusion-evasion alerts.
• User Management — locking accounts and other account-level actions you might take from an alert.
• Orders — reviewing the orders behind a spending alert.