ATYP

ATYP At ATYP, we grow businesses by putting the customer at the centre – and aligning everything else around them.

Share your perspective in five minutes.We’re asking leaders at established businesses how they’re thinking about growth,...
20/07/2026

Share your perspective in five minutes.

We’re asking leaders at established businesses how they’re thinking about growth, AI, capability and external partners.

ATYP will review every submission and invite selected respondents to a complimentary 30-minute Executive Growth Review with the member of our Executive Team best suited to the challenge or opportunity they’ve shared.

Every member of ATYP’s Executive Team has at least 18 years’ experience across growth, customer, product and technology.

We’ll also share a summary of the key industry findings once the survey closes.

Take the survey: https://lnkd.in/dHaRPNEp

Thanks in advance for helping us build a clearer, more evidence-led picture of where established businesses are heading.

A simpler Google PMAX structure delivered 54% stronger efficiency.That was not the result across every business we teste...
20/07/2026

A simpler Google PMAX structure delivered 54% stronger efficiency.

That was not the result across every business we tested. But it was a useful reminder: the more layered option is not always the more effective one.

We’ve been testing this across four businesses of different sizes. The question was simple: can a keyword-led search themes structure outperform a more layered signal setup, even when that setup is built from good data and sensible audience inputs? In several cases, the answer has been yes.

In Google’s highly automated campaign type, we tested search themes in isolation. In plain English: instead of giving Google several inputs at once, we stripped things back and let one signal guide the campaign more directly.
For these tests, we kept things simple. We did not give Google extra audience signals to work from. We only used search themes based on keywords that had already delivered value.

The reason was simple: we wanted to understand whether search themes could perform comparably to, or better than, a more layered signal structure, rather than assuming “more data” automatically improves the system’s chance of success.

The results have not been uniform, which is exactly why the test has been useful. In two of the four businesses, the search themes-only structures outperformed campaigns using combined audience signals. In one business, we saw roughly 54% stronger efficiency in aggregate, with some specific themes delivering 80% stronger efficiency. In another, efficiency improved by 20% - 30%.

In the third business, the same approach underperformed, with efficiency approx. 20% weaker than the more layered signal structure. In the fourth business, performance is currently more or less on par. But underneath the average, there’s a much wider spread. Some search theme subsets are significantly overperforming. Others are significantly underperforming.

Here the test has still given us something valuable: clarity. We can now see pockets of opportunity, and where the structure may be holding performance back. That gives us a better basis for optimisation than if everything had stayed bundled under a more general structure.

What’s been most interesting is that a relatively uncomplicated approach can sometimes outperform a more layered signal structure, even when that structure is built from good data and sensible audience inputs. In theory, the broader mix should give the system more context. In practice, that is not always what produces the clearest or most efficient outcome.

That does not make search themes-only the “better” approach by default. But it does make the case for testing the uncomplicated option properly, rather than assuming the more layered setup is automatically more sophisticated or more effective.

One reason data problems persist is that they rarely sit neatly in one team’s remit.Marketing feels the pain when perfor...
20/07/2026

One reason data problems persist is that they rarely sit neatly in one team’s remit.

Marketing feels the pain when performance cannot be read clearly. Tech may own parts of the implementation. Sales or CRM may own parts of the source data. Finance may have the numbers everyone trusts. Leadership needs the answers.

But no single team fully owns the system.

That is where things tend to get stuck.

Marketing adds caveats to the report. Tech logs the issue somewhere in the backlog. Sales cleans fields when the problem becomes visible. Finance reconciles what it can. Leadership asks why the numbers still do not line up.

And somehow, the underlying issue remains.

Not because no one cares. Usually, plenty of people care.

The problem is that everyone owns a piece, but no one owns the whole.

That is why data problems are rarely just technical problems. They are ownership problems, definition problems, prioritisation problems and decision-making problems.

Better tools can help. Better dashboards can help. Better integrations can help.
But at some point, someone has to be responsible for making the system coherent.

Not doing everything.

Not owning every platform.

But understanding the definitions, dependencies, gaps, trade-offs and decisions well enough to keep the whole thing moving in the right direction.

That means agreeing which numbers matter most. Defining the terms people keep using differently. Knowing where data is created, changed, enriched and reported. Deciding which gaps are merely annoying, and which gaps are actually limiting better decisions.

It also means turning recurring snags into funded fixes, rather than letting them become permanent footnotes.

Because if no one owns the system, every team keeps working around the same problem from a different angle.

And eventually, the workaround becomes the operating model.

AI hasn’t reduced the cost of underinvestment. It’s made that cost easier to miss.For years, businesses have underinvest...
15/07/2026

AI hasn’t reduced the cost of underinvestment. It’s made that cost easier to miss.

For years, businesses have underinvested in areas that create long-term competitive advantage: SEO, data infrastructure, CRM, lifecycle marketing, CX, measurement, content and technical capability.

That was always limiting, but there was usually some recognition of the trade-off. If you didn’t invest in SEO, you didn’t expect organic search to become a major growth engine. If your data infrastructure was neglected, you accepted that reporting, attribution and optimisation would have limits. If CRM was underfunded, expectations around personalisation and lifecycle marketing were adjusted accordingly.

That gap feels much wider today. The underinvestment remains, but the expectations have changed.

Businesses want to compete in AI search without properly investing in SEO. They want sophisticated CRM without the data, strategy or specialist capability behind it. They want automated advertising platforms to perform better while the signals feeding those platforms remain weak.

In other words, they want the outcome of investment without making the investment. AI makes that easier to rationalise, not because AI can’t help, but because many businesses are beginning to treat AI as a substitute for the foundations themselves.

That’s where the problem sits.

AI can amplify capability. So can automation. So can better tools. But amplification isn’t substitution.

It still depends on expertise, judgement, infrastructure, standards and people who know what good looks like. Without those things, AI doesn’t close the gap. In many cases, it simply makes the gap easier to ignore.

The longer that continues, the more expensive the problem becomes.

If competitors have spent years building SEO authority while you deferred the basics, you’re not starting from zero when you decide to invest. You’re starting from a deficit.

If your data infrastructure has been patched together over years, the fix doesn’t become easier because you waited. If your CRM strategy was never properly developed, AI personalisation won’t suddenly create the relevance or maturity the business hasn’t built.

Workarounds aren’t infrastructure. Prompts aren’t strategy. Automation isn’t expertise.

Businesses either invest in the foundations, or they adjust their expectations. They can’t underinvest indefinitely and still expect the outcomes of maturity.

That gap doesn’t close by itself. It compounds.

Address

Cape Town

Website

Alerts

Be the first to know and let us send you an email when ATYP posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to ATYP:

Shortcuts

Featured

Share

Category