AI Search, MarTech & Data Intelligence

Search stopped being a list of links

For twenty years the objective was a position. Rank first, collect the click, convert the visit. In 2026 that model describes a shrinking share of how people find things. Published research puts roughly 60% of Google searches as ending without a click to any external site, and dedicated AI search products — ChatGPT, Perplexity, Google AI Mode — run zero-click rates measured between 60% and 93%. When an AI Overview appears, the top organic result has been measured losing around 58% of its clicks.

The strategic consequence is uncomfortable but clear: visibility and traffic are now different metrics. A brand can be prominently described, compared and recommended inside an AI answer and receive no session at all. Conversely, absence from that answer is invisible in analytics — you cannot see the enquiries you never received. If your reporting only counts sessions, the most consequential change in search is happening where you are not looking.

What we actually do about it

The work divides into three parts. First, entity clarity: making it unambiguous to a machine what your organisation is, where it operates, what it sells, who works there and what evidence supports your claims. That means structured data, consistent naming, and factual statements written to survive being lifted out of context. Much of the E-E-A-T discipline that search engines rewarded now does double duty as AI-citation groundwork.

Second, answer-shaped content. Content engineered to be quoted has different properties from content engineered to rank: a direct answer near the top, specific figures with attribution, definitions stated plainly, and claims that are checkable. Our own service pages are built this way deliberately.

Third — and most neglected — earned sources. AI answers lean heavily on third-party material rather than the brand's own pages, which means citation share depends partly on what exists about you elsewhere. Directories, industry publications, client-side references and accurate mentions all contribute. This is why GEO overlaps with PR and cannot be solved purely on your own website.

Measurement that reflects the new reality

We report brand mention rate across the major answer engines, citation share of voice against named competitors, AI referral volume and conversion quality, and the correlation between citation frequency and branded search demand. Alongside those sit conventional GA4 and Search Console metrics, because organic search remains the larger channel — AI referrals still account for around 1% of website traffic on average, and treating them as a replacement rather than a leading indicator would be a serious misread.

There is a genuine upside in the data. Brands cited inside AI Overviews have been measured earning roughly 35% more organic clicks and 91% more paid clicks than uncited brands on the same query, and AI-referred visitors have been reported converting at several times the rate of traditional organic traffic — because the model has already pre-qualified them. Lower volume, materially higher intent.

The MarTech and reporting layer underneath

None of the above is measurable without a sound measurement foundation, and most reporting disputes we inherit are event-architecture problems rather than data problems. We define a conversion once, implement it properly in GA4 and Tag Manager, and report against that single definition consistently. Paid media then optimises on accurate signals instead of inflated ones, and executives get Power BI dashboards they can read in two minutes without an analyst present.

Media buying and bidding are increasingly agentic — Performance Max and equivalent systems make allocation decisions automatically. That makes the quality of the signals you feed them the primary lever a human still controls. Feed a bidding system a mismeasured conversion and it will optimise diligently toward the wrong outcome.