Running Account-Based Advertising Globally: The Regional Fault Lines Nobody's Mapped Yet

A reader asked recently whether we could offer a point of view on trends, audience behaviour, messaging and regional nuance across markets. This intentionally isn’t a data-backed victory lap of what I have seen in each market, because my weight of evidence isn’t equal enough across all territories. Instead, it’s fifteen-odd years of watching this industry lurch from market to market, held up against the best third-party evidence I could find, with the gaps left as gaps rather than papered over. If you’re running, or thinking about running, an account-based programme across more than one region, I think it’s worth your ten minutes.

---

What you’ll learn in this article:

• Why UK and EU privacy law are quietly pulling apart for the first time since GDPR, and what that means for addressability on each side of the Channel

• Why the US isn’t “looser” than Europe, it runs on an entirely different consent architecture, and why the difference is more significant than most vendors let on

• Whether the old floor trader’s instinct, hashed email in the US, clusters and personas in Europe, actually holds up against the evidence

• Why Singapore tops the global AI adoption tables while running a stricter consent regime than most of the region, and why those two facts sit together perfectly well

• What high-context and low-context culture actually means for B2B messaging and campaign sequencing, and why the “one deck, three markets” approach still quietly underperforms in 2026

---

The UK and the rest of Europe used to be one privacy story. They aren’t any more.

For most of my career, “European privacy law” has meant one thing. GDPR plus the ePrivacy Directive, enforced with local flavour but pointed in the same direction everywhere. I watched that direction get set in 2018, when GDPR went live and half the industry spent that May frantically re-papering consent flows they’d been ignoring for a decade. It caused carnage in agency land and a wave of premature in-housing happened to manage legal workflows at the cost of client performance. I watched it harden further when the Belgian regulator went after the IAB’s own Transparency and Consent Framework and the CJEU backed them up in 2023, which told every publisher and every DSP that the industry’s own self-built consent plumbing wasn’t going to save anyone. Reject the banner, and that was that, wherever you were standing.

That assumption is now out of date, and it’s the UK that’s broken the pattern.

Since February 2026, UK cookie law has been changing shape under the Data (Use and Access) Act. New categories of storage and access technology no longer require consent at all, and the ICO has gone further than that. It’s signalled openly that it intends to submit evidence to government on which advertising activities could earn a statutory exemption from consent, provided the privacy risk is genuinely low. This is something I have been privately advocating for yonks, and especially in B2B where the targeting currency becomes ‘the account’, with measurement mapped to that aggregation of people in an office; I genuinely think B2B sits squarely in that very low risk category.

The regulator’s own finalised guidance makes an argument I’ve been making in agency pitch rooms for a decade without much luck: treating a cross-site behavioural profile and a single contextual ad as the same category of risk was always a very blunt instrument. That’s the ground I covered in the piece on PECR reform a couple of months back, and nothing’s actually switched on yet, advertising still needs consent in the UK today. But the regulator is the one holding the door open, which is a genuinely new posture for a UK data authority to take. For the first time in a generation the privacy direction of travel has changed, and I really do think that is positive. The wider ecosystem of an ad-funded internet enabled by addressable advertising is what has kept the ‘web free since ninety-three’ as OG adtechers like to say, and that privacy cost has been blown out of all proportion. The ad ecosystem isn’t looking at bank account numbers and personal phone numbers, its pseudo anonymous identifiers that map back to signals like which business you may work for or what news story you were reading in which region of which country.

Before we get excited and think the whole continent of Europe, of which the UK has always had one foot in and probably a bit more than that out since Brexit in 2016, they are in-fact on another track entirely. The EU has gone the other way.

The long-stalled ePrivacy Regulation, which had been in negotiation since I was still cutting my teeth on RTB, was formally withdrawn in February 2025. Rather than loosening anything, the Commission’s Digital Omnibus package folds cookie consent straight into GDPR through two new articles. Article 88a restricts how often a site can re-ask for consent after someone’s said no. Article 88b will eventually make browser-level consent signals legally binding on every controller, which hands more power to the rejection, not less. And enforcement hasn’t slowed down to match the political mood. The French regulator fined Google and Shein a combined €475 million in a single day in September 2025 for cookie violations, and Brussels separately fined Meta €200 million over its consent-or-pay model.

So for the first time since GDPR landed, “Europe” isn’t one privacy regime moving in one direction. The UK’s exploring a risk-based exemption for low-risk advertising. The EU’s consolidating and tightening. Anyone running the same programme across the UK and the rest of the continent on the assumption that the compliance posture is identical is working off an assumption with an expiry date, and that date is this year, not some hypothetical future one.

All of this is explored in my ultimate guide to B2B programmatic advertising in 2026

The US was never “looser.” It’s built on a different foundation entirely.

Every European media buyer I’ve ever trained has walked into their first US campaign with the same instinct: this market’s simpler, less regulated, more room to move. I had exactly that instinct myself the first time I ran budget into the US, many years ago, and it’s not quite right, and getting it wrong costs you.

The UK and EU run on opt-in. Nothing gets stored on a device for advertising until the user actively says yes, and the default state is off. That’s true whether you’re dealing with the original 2011 UK implementation of the EU cookie directive, which is the piece of regulation that actually started the banner era before GDPR existed, or the version we’re living under now. The US runs on opt-out, and it still does even in its most regulated states. As of March 2026, no US state requires affirmative consent before a cookie is placed. The twenty states with comprehensive privacy laws, California’s CCPA/CPRA leading the pack, give consumers a right to opt out of targeted advertising and data sales after the fact. The default is collection. The user has to go and switch it off, and almost nobody does. That’s not a looser version of the European rule. It’s a different rule built on a different premise about who has to do the work.

That single structural difference is why individual-level, device-level signal has persisted at far greater scale in the US by default, and it’s exactly why deterministic identity, hashed email matched to mobile advertising ID, grew up there as a discipline in the first place. This manifests into being the go-to identity tactic for North American adtech businesses and about the last tactic to try for a European business. These legal norms determine the entire approach to identity that European versus North American adtech businesses take, which is an interesting nuance which probably often gets forgotten.

I strongly believe that any B2B vendor with global advertising ambitions are better working with a European adtech business, precisely because they have had to adopt the widest range of tactics to build deep addressability. The combined scale of the US market combined with the default to ‘on’ for data collection makes everything easier - and a rude awakening when US businesses come across the pond and collide with Europe’s default to ‘off'‘. As a UK based founder, I am bias, but I do think there’s validity in this argument.

I remember the identity graph wars properly kicking off around 2016 to 2018, Drawbridge and Tapad slugging it out over cross-device matching, LiveRamp building what eventually became RampID out of the old Acxiom onboarding business, and all of it assuming a US-style ocean of persistent, matchable signal to work with. Then GDPR hit in 2018 and a huge amount of that infrastructure simply didn’t translate to European traffic, not because the vendors got worse at their jobs, but because the underlying signal they needed to match against had been legislated into scarcity on this side of the Atlantic. Many packed their bags and left the UK/EU market entirely.

One identity vendor is blunt about what happens when a US-built graph gets pointed at non-US traffic: match rates fall off a cliff, because the hashed-email-to-device associations that make the whole thing work were never built at density outside their home market. It isn’t purely a regulatory story. It’s a data-density story too, and the two feed each other. Roqad, one of the European identity players, is upfront that it exists specifically because the US-style deterministic model doesn’t fit the region it operates in, which tells you the industry itself has already made this call, it’s just rarely said out loud to the client.

The old trading-floor instinct, hashed email in the US, clusters in Europe, mostly holds up

I’ve heard some version of this from nearly every experienced trader I’ve worked with. Individual-level resolution in the US, aggregate towards persona and account clusters in Europe. I went in wanting to test it rather than just repeat it, and it holds up better than I expected, with one real gap.

The mechanism runs in both directions at once, which is what makes it durable rather than coincidental. Regulation constrains what can be collected and how long it can be kept, and pushes European and UK activation towards contextual and account-level signal rather than individual identifiers, and that’s been true since 2018 and is about to get more true again as the EU tightens further through the Digital Omnibus. But underneath the regulation, the raw identity plumbing is thinner in Europe regardless of what the law technically permits, because the panel data and the deterministic HEM-to-MAID (matching hashed email signal to targetable device level signals) pairings that make US-style resolution work were built in the US first and were never rebuilt at the same density elsewhere.

Bombora’s own account of how B2B identity resolution actually works is a useful tell here. It leans on a blend of deterministic hashed business emails, IP-to-domain matching, SSO data and a proprietary work-from-home probabilistic model, precisely because no single deterministic method reaches workable B2B coverage on its own, and I’d bet money that blend leans harder on the probabilistic side the further you get from North America. This approach has led to Bombora re-gaining ground in the EU where their US cousins have lost it.

The gap, and I want to be straight about this rather than pretend I found something I didn’t, is Singapore and the wider APAC region. There’s a reasonable amount written comparing US and European identity coverage. There’s almost nothing written with any rigour about Singapore-specific match rates or graph density, and that silence is itself informative. It tells me the industry hasn’t scrutinised APAC identity infrastructure with anything like the attention it’s given the transatlantic comparison, and anyone who tells you with confidence what B2B match rates look like in Singapore today is very likely extrapolating from the US or European picture rather than working from actual regional data. Given we’ve got a genuine Singapore desk, that’s arguably a more useful place to point our own delivery data than re-proving the US-Europe split, which the wider industry has already covered reasonably well. All of this is exasperated by the markets focus on demand and direct response tactics like content syndication.

Singapore is the AI adoption story I didn’t see coming - and this directly impacts INTENT data capture

If you’d asked me a year ago to guess which market leads the world on AI usage intensity, I would not have said Singapore, and I’d have been wrong more than once, because it isn’t a fluke ranking. Anthropic’s own Economic Index put Singapore second globally behind Israel on usage relative to working-age population back in September 2025, and by the March 2026 update, Singapore had moved into first place outright. This is important because of the dark funnel and the fact AI is starting 51% of all B2B buying journeys now and is a part of upwards of 92-94% of them - which is explored in detail in the below piece around steering the dark funnel

The Singapore ranking is interesting. The reason behind it is more useful, and it maps directly onto how B2B buying decisions actually get made in that market. Forrester’s regional AI research found that in APAC, 33% of enterprises put the CEO in direct ownership of AI strategy, against 18% in North America and just 8% in Europe. APAC firms also back that ownership with real budget: 26% invest between $400,000 and $500,000 in generative AI, against 19% in North America and 17% in Europe.

Forrester’s own read is that European firms operate under tighter regulation and stronger labour protections, which pushes their AI posture towards governance rather than speed, while APAC firms are competing in faster-growing markets where speed itself is the edge, and they’re prepared to fund that with capital and put it directly under the CEO rather than diffusing it across a committee. This is a fasinating global split in attitude, pace of market and general approach

If you’re advertising to, or selling into, B2B audiences in Singapore, that’s not a trivia point. It means you’re more often talking to a buying committee where AI-adjacent purchasing decisions sit close to the top, and where the appetite to move on a new platform or data partnership is structurally higher than it is in the UK or the rest of Europe. That’s a different sales motion, not just a bigger TAM.

Its also worth considering the impact on intent data - which is often largely formed out of observable behaviour in publishers and vendor websites. If their market is driven more by AI, which gives off no exhaust fumes and they mainly focus on content syndication, which captures data off-domain (not on the vendors core-site) then this market is more limited in its ability to capture intent, period.

It’s worth setting that against the UK’s own mood, which sits at the opposite end of the same spectrum. A UK and Ireland survey of 277 B2B marketing leaders, published in January 2026, described the finding in its own title as a sense of hesitancy: strong budget growth and near-universal tool adoption sitting right alongside real reluctance to hand AI anything genuinely strategic. That isn’t a contradiction of the Singapore story. It’s the other end of the exact spectrum Forrester’s ownership data describes, and if you’re running one global content calendar across both audiences, you’re pitching confidence to one room and caution to the other, and most decks I’ve reviewed don’t seem to know that.

Messaging doesn’t travel as cleanly as most global campaigns assume, and I learned this the expensive way

This last one is less about infrastructure and more about the actual words on the ad, which ties to general advertising tactics, and it’s the mistake I’ve made personally, more than once, before I knew better.

Years ago I ran a campaign built for a UK and US audience out into an APAC launch with nothing more than a language swap. Same structure, same direct ask up front, same “here’s the number, here’s the CTA” cadence that had worked perfectly well at home. It underperformed badly, and for a while I assumed it was a targeting problem, because that’s always the first thing you check. It wasn’t a targeting problem. It was a message architecture problem, and the anthropologist Edward Hall had already explained why decades before I ran into it, if I’d bothered to read him sooner.

Hall’s distinction between high-context and low-context cultures still holds up as a genuinely useful practical filter for B2B messaging. The UK, the US, Germany and the Netherlands are consistently cited as low-context markets that respond to direct messaging and explicit value propositions: state the claim, back it with a number, close with an ask.

Much of APAC runs on the opposite instinct, where relationship and shared understanding carry more of the message than the literal copy does, which is exactly why B2B campaign sequencing across the region so often opens with thought leadership or a webinar to build trust before anything resembling a direct lead-gen ask ever appears. One localisation specialist puts it about as plainly as I’ve seen it put: North American programmes can lead with a direct lead-gen ad from day one, while APAC programmes typically need the trust-building content to run first, or the direct ask simply doesn’t land, which is precisely what happened to me.

Singapore sits in an odd middle position here, English-speaking and commercially direct by regional standards, but still embedded in a higher-context region, which is probably why so much generic “APAC” creative guidance doesn’t fit it especially well either. Honestly, this is another spot where our own delivery data could add something the wider commentary hasn’t nailed down.

What this actually means if you’re running an account-based programme across two or more of these markets

Pull the threads together and a handful of practical calls fall out fairly cleanly.

Treating “Europe” as a single addressability zone is no longer safe. The UK and the EU are diverging, and a programme built on the assumption of identical consent posture on both sides of the Channel is going to misjudge one of them within the year. Treating the US as a bigger, looser UK misreads the actual mechanism, it’s an opt-out architecture sitting on top of denser deterministic identity infrastructure, and the two together are why individual-level resolution travels so well there and so poorly everywhere else. The instinct that Europe leans on clusters and personas rather than individuals is basically right, but it’s driven as much by thin identity plumbing as by regulation, and Singapore in particular remains under-documented enough that anyone with a confident opinion on APAC match rates is very likely borrowing it from the US or European picture rather than the region itself.

The takeaway is that a vendor with global advertising aspirations would be better using an EU-native business to stand it up, and translating that architecture into the US works better because it IS easier. Taking the US approach to the EU makes the market feel smaller then it is.

And on messaging, a single global deck with a language swap is a weaker starting point than most teams assume, especially running UK or US-style direct-ask creative into APAC without the trust-building sequencing that market usually needs first, a lesson I paid for once so hopefully you don’t have to.

None of this is gospel and is my own interpretation on data. It’s a first, properly sourced pass at a question worth answering with our own numbers eventually. The honest next step, and the one a reader effectively asked for at the start of all this, is pulling our actual delivery data across London, New York and Singapore and seeing where it agrees with the wider industry picture and where our own patch of the market turns out to be the exception. If it holds up, that’s a piece worth writing on its own, and a better one than this.

If you’re the person in the room who has to explain why the UK deck doesn’t work in Singapore, this is what The B2B Stack is for.

Fifteen years in, most of what I read on global B2B advertising is either a generic localisation checklist or a regulatory update with no practitioner behind it. This is written from the trading desk out, one genuinely useful read a week on B2B programmatic, intent data, and the regional detail that never makes the vendor deck. No recycled listicles, no press releases dressed up as insight.

---

Sources and further reading

The case for UK privacy law diverging from the EU rests on the ICO’s finalised guidance on storage and access technologies and Covington’s analysis of the ICO’s more permissive signalling, read alongside our own earlier piece on the PECR climbdown. The EU’s opposite trajectory is covered in Secure Privacy’s breakdown of the Digital Omnibus and Consenteo’s 2026 practitioner guide, which carries the CNIL and Digital Markets Act enforcement figures. The US opt-out architecture is set out in Recording Law’s state-by-state guide, and Singapore’s consent framework in Chambers and Partners’ 2026 Singapore data protection guide.

On identity resolution, Echo Analytics’ piece on coverage asymmetry, Roqad’s identity buying guide and Bombora’s explanation of B2B identity resolution between them cover the deterministic-versus-probabilistic split referenced above. The AI adoption data comes from Forrester’s regional AI adoption research and Anthropic’s own Economic Index, with the March 2026 Singapore ranking update via the Singapore AI Observatory. The high-context/low-context framing draws on Oban International’s cultural differences primer and Elevation Marketing’s B2B localisation guidance.