Brand Compliance Monitoring

So what goes wrong on a page nobody has touched

Nobody touched it is not quite true, it is just true of every individual person you could go and ask. Pages stopped being documents a long time ago. What you publish now is assembled out of parts. A promo module pulling from a feed, a banner slot on a rotation somebody set up two campaigns back, a legal footer component shared across every brand in the group, a terms block that renders from whatever the CMS holds today rather than whatever it held the day the page was approved. Change one of those parts centrally, for a perfectly sensible reason, and you have gone and edited four hundred pages without opening a single one of them. And the approval, whenever it happened, was given on a page, whereas the change went into a component, and there is nothing sitting between those two that would ever put the one in front of the other.

Then there is the plainer sort, where a person did edit it and simply did not think of it as an edit. A promotion gets extended, so the end date moves on the main page and nowhere else at all, on account of the main page being the one open in front of whoever did it. A game gets pulled by a provider and the offer that named that game keeps naming it. A banner gets swapped for a seasonal one and the new artwork carries the offer headline but not the conditions that used to sit under the old one. Somebody localises a page properly and carefully, and the localisation carries a figure the English version has since moved off. That last one compounds so quietly that we wrote a whole separate piece on the same offer said four different ways, and none of the people in that story did a single thing you could fairly call careless either.

And then the third kind, which catches the tidiest operators of the lot, where the page never moved at all and the rule underneath it did. Guidance gets updated, a regulator publishes a view on a phrasing it used to leave alone, a market you have held a licence in for years tightens what has to sit beside an offer. Your page is exactly as it was of the day it passed. It just is not passing any more, truth be told, and there is nothing in your own systems that would ever go and tell you so, on account of nothing in your own systems having changed.

The bits of a page that carry the licence

The part of a page that gets checked hardest is nearly always the offer headline, on account of it being the loudest thing on the screen and the thing the campaign meeting argued about. It is rarely where the problem is, mind you. The problems collect in the furniture around it, the pieces nobody feels much ownership of, the pieces that are somebody else's component and therefore nobody's.

Significant conditions are the first of them. Under the CAP Code the conditions that would change a person's understanding of an offer have to be communicated clearly and prominently with the offer its self. The rest of the terms have to be no more than one click away. That is rule eight point seventeen, and it is about as plainly written as these things get. So when a banner swap leaves the headline sitting on its own with the wagering requirement now living two pages back, that is not a design decision that went slightly wrong, that is the rule. The same goes for the word free, which CAP is clear should not be used where a customer has to risk their own money to qualify, and for risk free, which should be left alone entirely unless the player genuinely cannot lose funds of their own. Those two turn up in old creative constantly, usually pasted in from a market where they were grand at the time.

Then the display obligations, which sound like the easy ones and are not. A British remote licence carries a condition on displaying your licensed status with a link through to the Commission's register. That is licence condition eight point one one. It is the sort of thing that stays correct for two years and then goes wrong the week a developer rebuilds the footer. Age marks go the same way. So does responsible gambling messaging, which carries the added trouble of being market- specific in both its wording and the helpline it points at, so a footer shared across the whole of your estate is a footer that is wrong somewhere by definition.

And then the judgement-shaped ones, the rules about what an advert may suggest. Marketing has to be socially responsible, it must not portray or encourage gambling behaviour that could do somebody harm, it must not present a bet as a way out of personal or financial or professional trouble, and it must not be likely to be of strong appeal to under-eighteens. Those are rules sixteen point one and sixteen point three of the CAP Code and they are the ones that catch creative rather than copy. A cartoon treatment that felt fun in the brief, a sports personality who turns out to have a large teenage following, a line about a bad week and a fresh start that read as warm to the person who wrote it. Nobody goes and publishes those on purpose. They arrive through a template, from a partner, in a seasonal refresh, and then they sit there quietly being the version of you that people see.

One brand, two licences, and two different right answers

Here is the bit that makes this a harder job in our industry than it is in most others. A page is not compliant or otherwise, it is compliant under a licence, and the very same page, word for word, can be sound under one authority and a problem under the next one over. So the question a check has to answer is never just whether the page follows the rules, it is which rules, in the market that page is actually being served into.

Take the plainest example of it. A page carrying a British licence shows eighteen plus and points at British safer gambling support. The same brand operating in New Jersey is looking at twenty one plus and a different helpline entirely, and a different set of expectations about what has to be disclosed and, in several US states, about who has to see the creative before it goes anywhere at all. Neither version is the wrong version. They are both right, each in its own market. The only failure going here is serving one into the other, which is precisely what happens when a group builds a shared component and rolls it out across the estate because it was correct in the market somebody wrote it for.

So when you connect a brand to MARC you tell us which licence covers which market, and every page gets checked against the authority that governs it rather than against one flattened checklist that tries to keep everybody happy and ends up strict in the wrong places and slack in the others. We hold the rulebooks for a hundred and fifty five of them, which is a number that only matters on account of this, it was never put there to be impressive on a slide. It is there so that a group running four brands across nine markets gets nine answers and not one average.

Why the estate you own is the bit nobody checks

Most operators check, and they check properly, and that is what makes this a frustrating thing to sit and explain to somebody who is doing the work. The checking gets pointed at whatever is important. The new market, the flagship brand, the campaign the commercial team is nervous about, the page that carries the quarter. Choosing by importance feels like the only sensible way to choose. For this particular problem it is precisely backwards, since whichever pages hold the attention are the very ones getting updated, so the review lands over and again on the pages that had the least chance of going wrong in the first place. The market page nobody has opened since a launch two years back is the safest looking page you own and it is the one quietly serving an expired offer to whoever lands on it.

Volume finishes the argument off. A group with a handful of brands across a handful of markets is looking at several hundred live surfaces where an offer or a licence detail or a piece of safer gambling messaging is stated, and that is before anybody counts the terms pages behind them. A person can hold maybe a couple of dozen pages in a session before their attention gives out on them. After that the eye starts reading whatever it expects to be sitting there rather than what actually is. That is not carelessness at all. It is only what reading is. And a mismatch between a page and a rulebook is not even the hard case. The hard case is a mismatch between two of your own pages, which no amount of reading one page at a time will ever surface. You would have to hold the other one in your head the whole while you read.

What we line up when we check a page

Say the offer is fifty spins on a named slot, ten pence a spin, winnings capped at a hundred, forty times wagering on those winnings, seven days to use the lot of it. The figures are made up for the example, mind you, though the shape of it is one you will know well enough.

The check does not ask whether the page looks about right. It goes field by field. The headline claim against the conditions sitting beside it. Those against the full terms the page links to. Those against whatever the market variant of the page says. Then against however the promotion is really set up to pay out inside the platform its self, which teams skip more than any other comparison on the list, and which settles whether somebody claiming the offer receives the thing the page showed them. Then the furniture, separately, and against the market rulebook rather than against your other pages. Is the age mark the right age mark for this licence. Whether the licensed status is displayed at all, and whether the register link still resolves rather than going off somewhere that moved in a rebuild two years ago. The safer gambling wording gets read against what this market expects, helpline and all, since a helpline number is a country-level thing and travels badly. Then the small ones. Is the game the offer names even live still. And is the page being served somewhere the offer was never available in the first place.

Then it does the same again the next time the page changes, which is the part people underestimate. We take a clean state for every page on the first pass and diff against it after that, so what arrives is not a page score that got worse, it is the exact element that moved, with a screenshot of it, timestamped. The screenshot does more work than you would think of it doing. A good deal of the value in any finding is only being able to put the thing in front of somebody, and there is more again in being able to show a regulator, eventually, what you knew and the day you went and knew it. If you would rather catch the problem before the page is live at all, that is the Landing Page Checker doing the same work at the cheaper moment.

The pages you did not publish are stating the same offer

Your own estate is only ever half the surface an offer lives on. Affiliates copied that offer into their review pages and comparison tables at some point, accurately at the moment they copied it, and they have not been back since on account of the page being finished as far as they are concerned. The offer shifted under them and nobody went and told them so.

That is the Affiliate Audit rather than this module, and the two are separate on purpose, but they answer to the same problem and most operators run both. What matters here is not to mistake one for the other. Watching your own pages tells you whether your estate is sound. It tells you nothing at all about the pages you did not publish, and those are frequently where the older wording survives longest, on account of nobody there having any reason to go looking.

What a score out of a hundred is good for

We give you a single number, zero to one hundred, across every page and brand and market you have connected. The trend sits under it. It is worth being honest about what that is for. A score is a management object. It is the thing that goes in front of a board. It is how a head of compliance shows that last quarter's work moved something at all, and it is how you come to notice that a brand nobody has mentioned in months has been sliding quietly for three weeks. What it is not is the thing you act on. Nobody ever went and fixed a page by looking at a seventy four.

The trend is more useful than the level, in truth. A score of ninety one that has dropped four points since a deploy is a far more interesting object than a score of eighty two that has sat still all year, and the second one tells you mostly about how old your estate is. The findings underneath are the work of it, and the number is only there to tell you whether that work is getting done.

The bits of it we do not solve for you

Fair is fair, mind you, so here is the honest shape of what we leave on your side of the desk.

What we hand you is a disagreement between a page and a rulebook. Whether that disagreement matters is a separate question and it stays with a person who knows the market. Some disagreements are a genuine problem. Others are a page correctly doing a different thing because its market requires it to. Our checks are generated by a model, they can be incomplete and they can be wrong, and they are not legal advice and were never meant to read as it. What we are really selling, if we are straight about it, is that the call turns up on a short list of a Monday morning rather than sitting unread in several hundred pages nobody ever had the hours for.

We read published pages, which is a real boundary and worth saying plainly. We do not sit in an ad auction, we do not watch what was bid on your brand terms overnight, and we do not report on paid search. Where an ad points at a landing page we can tell you about the page at the end of the click, and that is the honest edge of the thing. Same goes for anything that never gets a URL. Email copy, push messages, in-app promos and the creative sitting in a partner's pack are all advertising and they all sit under the licence, and none of them are pages, so they belong to the Content Checker rather than to this.

Finding a thing is quick. Fixing it is not, and that gap belongs to you rather than to us. A drifted component takes us minutes to spot and it takes your development queue however long your development queue takes, and a shared footer that is wrong in three markets is the one ticket four separate teams will have opinions about. We can shorten the finding end down to nearly nothing. The far end sits inside your own organisation, and no software anywhere has ever fixed that part for anybody.

And the plainest limit of the lot, which is that monitoring is not a compliance function. A live score and a queue of findings is a good thing to have on a Monday morning and it does not decide anything. Somebody still has to own the problem their own self, read the findings, know the market and make the call. All we change is that they make it on purpose, off a list, rather than of a Wednesday afternoon on account of a stranger having sent an email about a page from March.

Where to start, if any of this sounds familiar

Start with a market rather than a brand. Pick the licence you would least like to receive a letter about and write down every page you publish into that market, homepage, promotions, each offer's terms, the deposit flow, whatever landing pages are still live from campaigns that ended. Most people find that list is longer than they expected before they have finished writing it. That is itself a finding. Then take the oldest three pages on it, not the newest, and go through the furniture on each one your own self. Age mark, licensed status, register link, safer gambling wording and helpline, and whether the conditions on any offer are sitting with the offer or a click and a scroll away from it.

An hour of that turns something up more often than not, and it is rarely the thing anybody went in expecting. If any of that bothers you, and it usually does, that is about where the hand version of this quietly runs out of road. We can then run the same check continuously across the whole estate, against the rulebooks of a hundred and fifty five licensing authorities rather than a checklist somebody wrote for one market and carried into the others after. Best way to judge whether any of that is real is a demo run over your own live pages, on account of twenty minutes of that telling you more than another page of ours going on about compliance ever will.