The Digital Markets Act: What It Gives Advertisers, and What It Withholds
If your ads reach people in Europe, EU law grants you a free daily record of what you actually paid, per ad. Almost no one claims it.
Acquisition
Channels, budgets, creative, measurement and choosing a provider: what decides whether a dollar of advertising produces qualified demand rather than a statistic.
Paid acquisition is the area where the most confident numbers circulate and the fewest are sourced. A benchmark gets quoted, the original study turns out to have measured something else, and a budget is set on it anyway.
Most of these articles start by going back to the source. Several of them end by showing that the source does not say what everyone repeats.
One rule holds across all of it: a channel is not good or bad, it is appropriate or not to a buying cycle, an average deal size and a volume. Everything here is written to make that judgement yours rather than a vendor's. Our full approach is on the B2B paid acquisition page.
68 articles in this category. See all resources
If your ads reach people in Europe, EU law grants you a free daily record of what you actually paid, per ad. Almost no one claims it.
No US institution measures channel performance. And a landmark study says the number you want cannot be cheaply produced, by them or by you.
At one audited show, 39 per cent of the headline attendance was exhibitor staff. The space is about half the cost of being there.
An eighth of an inch too thick doubles the postage. And the only response benchmark in the channel is published by the channel's own trade association.
The Do Not Call Registry is written for residential subscribers. The rule that does reach business calls keys on the number type, not the person.
The bulk sender rules do not formally apply to B2B. No public registry holds email addresses. And open rates broke in 2013, not 2021.
A view runs from zero to thirty seconds depending on who sells it. On two platforms a views objective bills by impression. Completion is tied to no sale.
Under ten domains, 0.02% junk inventory. Over five hundred, 19.8%, open-auction quality at double the cost. The deal ID verifies none of it.
Of 67 advertisers who wanted access to their own delivery data, 46 never got it: their contracts forbade it. Here is what to fix, and in what order.
Of every $1,000 entering a demand-side platform, $150 buys impressions whose visibility was never determined. It appears on no report, and it beats the fees.
CPM is not a price, it is a quotient. The denominator is not standardized, most of the numerator never reaches the publisher, and no average is published.
A journey map has one lane. A buying group runs from five to 16 people, and 74 percent of buyer teams show unhealthy conflict. Map the group instead.
Every channel costs money or hours. What each one demands, which figures used to compare them have no source, and the one effect that was actually measured.
A business with no storefront has its own published rules, and they decide what you can own. What a lead platform sells you, and what accumulates instead.
The thresholds have not moved since 2024. The enforcement has. And three of the rules people quote have no primary source at all.
No platform has ever published an average organic reach figure, and the metrics differ so much that a cross-platform benchmark is arithmetically impossible.
Google defines the third local ranking factor as how well known a business is. That is not a search problem, and it changes what advertising can buy you.
The peer-reviewed evidence is weaker than the vendor claims, the index everyone optimises has no published formula, and the famous statistic has vanished.
Google's eligibility test is one sentence, and many B2B companies fail it. What the rules say, the three ranking factors, and the statistic with no source.
The company credited with coining it swore a first-use date of 2008 to the trademark office, and the phrase appears in the book record from the 1970s.
A retailer switched off its brand-keyword ads and organic search caught 99.5 percent of the clicks. Its measured return on the rest was negative 63 percent.
Not the design. One visitor typed a question, the other was interrupted. Both platforms publish what they judge your page on, and the lists differ.
Your customer filmed it, you paid for the campaign, and you own none of it. Three separate bodies of law govern a customer video, each needs its own document.
The most-quoted number in marketing automation has no source. The rules your sequences must satisfy do, and so do the thresholds Gmail and Yahoo enforce.
Third-party cookies are not going away in Chrome, and Google said so three times. The reason to send conversions from your server is a rule few have read.
Same panel, same year: LinkedIn and Google Ads clicks cost almost the same. The leads do not. Why the comparison you have been shown is the wrong one.
A regulator measured it: publishers receive about 65p of every advertiser pound. A US study found 36 cents reaching the consumer. The whole chain, sourced.
The funnel descends from AIDA, whose authorship historians dispute. Buying does not descend. What the three stages still explain, and where they mislead.
On a handful of deals a quarter, a customer acquisition cost swings by chance more than by decision. What is actually readable, and what the 3:1 rule never was.
A countdown timer is a factual claim. So is 'only 3 left'. US law requires you to have a reasonable basis for both before you publish them.
Your max CPC, your actual CPC and your average CPC are three different objects. The third is an amount nobody was ever charged, and your benchmark is a median.
There is no opt-in requirement and no B2B exemption. Both facts surprise people, and the second one is the expensive surprise.
A Meta ad account created in an agency's portfolio can never be moved out of it. Google says account-level history does not exist. Plan the handover on that.
Media buying is now an auction, not a negotiation. Where the money actually goes, the eleven words you need, and what US survey data says about agency fees.
CPM is no longer a rate you negotiate, it is an auction output. The published US benchmarks disagree by a factor of three, and a low CPM often costs more.
The CPL formula is trivial. What is not is deciding what counts as a lead, and the one standards body that defines it means something else entirely.
CTR goes up when you narrow targeting, prune a list or overpromise in a headline. None of that improves anything. And when a gap is actually real.
Search volume is the wrong selection criterion in B2B. How to structure by intent, and three places Google's documentation contradicts common practice.
Six or fewer per ad set was official guidance in a February 2025 capture. It is gone from the live page. What remains is a ceiling, a duration, and no budget.
Meta publishes exact thresholds everywhere else: 50 results in seven days, 17 purchases plus 5. On budget changes it publishes no percentage at all.
Reach, frequency and GRP have exact published formulas. The three-exposure rule inverts what its author argued, and no platform publishes a frequency.
The figure is real and published. The scale was never validated on B2B advertising, and everyone who produced it sells to advertisers who believe it.
The platform says the words do not have to match. It says the promise does. And the score people optimise for is not an input to the auction at all.
No budgets, no impressions, no end dates, and outside the EU no commercial ads at all. The gap between what people expect and what the law actually requires.
The ads guide documents four formats. The specs change with every placement, one objective hides your call to action, and thirty enhancements alter it.
One recommends under 15 seconds with no data. One recommends 15 to 30 for placement eligibility, not attention. The two with the most data say nothing.
It appears in no industry glossary, no measurement guideline and no platform documentation. The two definitions in circulation divide incompatible metrics.
Every platform names a different number for the opening seconds. Only one of them cites research, and its number turns out to be the skip button.
The 4x figure has no published study behind it. What does exist: a federal rule since October 2024 that puts the advertiser, not the creator, on the hook.
On the same B2B panel, static images beat video on cost per lead across three platforms. And the cheapest format of all is neither.
Two different jobs share one label. One is measured in a quarter, the other is not, and that asymmetry is why the second one keeps getting cut.
Google answers a question somebody asked. Meta interrupts a scroll. On one B2B panel of 153 advertisers, a Meta lead cost $145 and a Google Ads lead $524.
Google's own documentation works the example: 12 channels and two years of weekly data give four data points per parameter, which it calls too low.
eBay switched off paid search and kept 99.5% of the clicks. Then a second experiment found the opposite on contested brand terms. Both results are right.
The comparison is usually run on hourly cost, the one dimension where the answer does not matter. What the published data says to compare instead.
Who owns the ad account, what the agency earns on your spend, and what happens the day you leave. Three answers you can verify before you sign anything.
Three platforms, three answers to who owns the ad account. One documents that it can never be transferred. Run this audit before you say you are leaving.
Modern Meta delivery rewards fewer ad sets, each with more signal. When to consolidate, the ~50-conversion learning rule, and CBO vs ABO.
Server-side tracking recovers conversions lost to Safari ITP and iOS, but only past a certain volume. When it pays off, what it costs, and what it will not fix.
Forget the magic percentage of revenue. Your advertising floor comes from your cost per result and the volume a platform needs in order to learn.
Meta Advantage+ can complement your manual campaigns or cannibalise them. The real conversion thresholds, the independent data, and what B2B should expect.
Frequency is a weak signal for creative fatigue. The metrics that actually flag a dying Meta ad: cost per result, the CTR and CPM crossing, and cadence.
The customer acquisition cost formula is one line. What changes everything is scope: fully-loaded costs, blended CAC vs paid CAC, and CAC by channel.
ROAS drops when you scale for four separate reasons: audience saturation, creative fatigue, learning resets and measurement noise. How to tell them apart.
Facebook ads have no single price but two costs: auction media spend and management. Real 2025 US CPM, CPC and CPA benchmarks, sourced and explained.
How to split budget between Google and Meta Ads in B2B: the three variables that decide it, an ideal split, the 70/30 rule, and when LinkedIn is worth it.
No single acquisition metric tells the truth alone. How ROAS, MER, CAC and LTV differ, why a 3:1 LTV:CAC is the floor, and which one to run on by stage.
GA4 and Meta rarely show the same conversions, and neither is wrong. Why the numbers diverge, how big a gap is normal, and which one to trust.
Would you rather talk it through? A conversation often beats a long article. Let's discuss your situation.