Three doors opened in four months. Alcohol walked through all of them

Three changes, not one, and the order matters#

Most of the trade coverage this week treated Google's announcement as a standalone gift to drinks brands. Read on its own it is exactly that. Read against the two changes that preceded it, it is the last of three deliberate steps, and the sequence tells you far more about where alcohol advertising is heading than the headline did.

The scent was there for anyone following the trail since midsummer. Search Engine Land reported the October change as a discrete announcement, which it is, but Google had already moved the boundary stones twice this year, and a fox reads the whole hedgerow rather than the last gap in it.

Start at the end. Google's personalised advertising policy update, published on 3 September, sets the date plainly: Starting on October 30, 2026, targeted advertising becomes available for alcohol on YouTube inventory where local law permits it. The scope is broad, covering alcohol, alcohol-related products, and alcohol alternative beverages.

Four markets are held back. Advertisers in Egypt, India, Indonesia and Poland do not get this yet, and Google offers no reasoning beyond saying it is not available there with this update. Google's stated rationale for the change itself is growing customer interest, emerging categories, and improved controls and data protections.

The July change was about expanding the approved geographic locations for the sale and informational promotion of alcoholic beverages. The September change is a rebuild: Google says it will introduce a more unified global advertising framework, and warns that it revises country-specific guidelines, expands allowed advertising locations, and provides clearer details on specific advertiser requirements. The October change hands back the targeting. Territory, then rulebook, then reach.

What the alcohol advertising rules actually released#

To understand the size of this, you need the cage it opens. Alcohol sits on Google's list of sensitive interest categories, alongside health, gambling, religious belief and sexual orientation. The restriction that list carries is specific and severe: Advertisers promoting products and services that fall within sensitive interest categories are unable to use advertiser-curated audiences.

In practice that has meant no customer match, no first-party data segments, no lookalikes and no audience expansion. A drinks brand could buy enormous reach and almost no precision. Every campaign was a broadcast, which is why alcohol marketing on video has looked like television for a decade: it had no other option available to it.

The content restrictions in the underlying alcohol policy survive intact. Ads still may not target people below the legal drinking age in any location being advertised to, may not imply that drinking delivers health or therapeutic benefits, and may not present heavy drinking approvingly or feature competitive or binge drinking.

There is a neat clue in that same policy about one of the four excluded markets. Google notes that In Poland, brand or informational advertising for alcoholic beverages is limited exclusively to beer. A market whose national law already narrows the category to a single product type is an awkward place to launch broader targeting, which makes Poland's exclusion look less arbitrary than it first appears. That reading is ours, not Google's, and worth holding lightly.

For anyone running paid social advertising in a restricted category, the practical shift is that audience strategy becomes a live question again after years of being a settled one. That is a bigger change to a media plan than it sounds.

Worth saying plainly for anyone new to this thicket: none of it removes alcohol from Google's advertising policies. The category stays restricted, which means permission is conditional, revocable and market by market. What has changed is the size of the clearing inside that restriction, not the fence around it.

The evidence this policy is walking past#

Alcohol is a sensitive interest category for a reason, and it would be dishonest to write this piece as though the loosening happened in a vacuum. The most recent systematic evidence is not ambiguous, and any brand walking through the new door should know what is written above it.

A meta-analysis published in The Lancet Public Health in November 2025 pooled 31 studies covering 62,703 participants, drawn from six databases and covering research from January 2004 to February 2025. It found consistent associations between exposure to digital alcohol marketing and drinking behaviour across every outcome it measured.

The confidence intervals are worth stating rather than hiding: 1.22 to 2.67 for binge drinking, 1.39 to 2.20 for past 30-day use. Heterogeneity is high, above 88 per cent on every outcome, which the authors report openly and which means the studies vary a great deal in design. The direction, though, is consistent, and the effects were larger among adolescents than adults.

None of that makes targeted alcohol advertising unlawful, and none of it is a reason for a drinks brand to sit out a channel its competitors will use. It is a reason to expect scrutiny, and to be able to answer it. The full paper is open access, so the people who will quote it at you have read it.

Where the rules still bite hardest#

Google's permission is not the only permission. In the United Kingdom the binding constraint is the CAP Code, and its alcohol section contains a rule that sits awkwardly beside individual-level targeting.

Rule 18.15 states that No medium should be used to advertise alcoholic drinks if more than 25% of its audience is under 18 years of age. That is a media-composition test. It asks what a channel's audience looks like in aggregate, and it was written for a world of media buys rather than audience buys.

An aggregate rule meeting an individual tool#

Targeted advertising does not work in aggregates. It works one impression at a time, which means a compliant channel-level composition and a badly built audience can coexist comfortably. The rule is not broken by the new targeting, but it stops being the safety net it used to be, and the burden moves onto the advertiser's own audience construction.

The ASA's own guidance on alcohol advertising makes the practical point that alcohol marketers must not imply that drink can change mood or behaviour, nor link it to social or sexual success. Those are the traps that catch good creative teams working at speed, and they do not care how precisely the audience was built. A brand can outfox the targeting rules and still walk straight into a content ruling.

The wider CAP requirements still apply too: alcohol advertising must not be directed at under-18s through media selection or context, must not appeal particularly to under-18s, and must be socially responsible. In the European Union the Audiovisual Media Services Directive bars audiovisual commercial communications for alcoholic beverages from being aimed specifically at minors or encouraging immoderate consumption, and it reaches on-demand and video-sharing platforms rather than only broadcast.

What to do before the bolt is drawn#

There is a trap in this timetable that will catch slow teams. The September rebuild revokes existing permissions and requires a fresh application. An advertiser who waits until late October to think about this may find they have lost the permission they already had and gained nothing, because the thing that unlocks on 30 October is useless to an account that is no longer approved.

So the order of work is the reverse of the order of the announcements. Handle September first. Treat October as the reward for having done so.

After that, the real work is audience construction, and it is unfamiliar work for this category. A decade of broadcast-only buying means most drinks teams have no tested first-party audience, no suppression discipline and no age-assurance logic beyond what the platform enforced for them. Building that carefully is the difference between a channel that opens and a channel that opens and then closes again after a complaint.

There is a quieter opportunity in here for smaller drinks brands, and it is the one worth denning down with. For ten years the category's media has favoured whoever could buy the most reach, because reach was the only lever anyone had. Precision changes that arithmetic. A regional distiller with two thousand known adult customers has, for the first time, a way to spend against them rather than against a demographic cloud.

Good alcohol brand marketing in this new shape will look conservative for a while, and it should. Narrow audiences, heavy suppression, creative that would survive a regulator reading it aloud. The brands that move first and carelessly will write the case law everyone else has to live under.

Small drinks marketing teams should not read that as an invitation to rush. It is an invitation to prepare, because the brands that qualify early and build their audiences carefully will be prowling a channel their larger competitors are still filling in forms for. Being three moves ahead in a restricted category is worth more than being loud in an open one.

If you sell in a category the platforms keep at arm's length, this is familiar ground for us. It is the same patient prowl we run for iGaming and healthcare clients, where the rulebook moves faster than the media plan and the quiet quarry is a channel nobody else has bothered to qualify for.

The wider lesson holds beyond drink. Every restricted category moves this way eventually: a long freeze, then a thaw arriving in instalments that nobody announces together. Reading the instalments as a pattern rather than as news is most of the advantage, and it is available to anyone willing to keep a moonlit eye on the policy pages.

Frequently asked questions#