AcademicarXiv, 11 Sep 2026

The source-state term made up 66.34% of the total GEO effect in the controlled design

Performance measurement

66.34%

of the simulated total GEO effect came from the source-state term in the controlled design

Reader Note:: This share comes from the paper's simulated decomposition of GEO effect components, not from a live campaign.
Evidence:: Modelled / projected
Vendor interest:: Low
Sample:: Main evidence is simulation-based. The paper also collected 2,240 complete AI answers from two GPT models answering 56 product-recommendation questions in English and Japanese, with 20 repeated answers per question-model pair.

Good for

Planning with a rough expectation, and sanity-checking your own assumptions.

Not for

Reporting as what happened. This is what a model expects, not what anyone counted.
Nobody counted this directly — it's a calculation. The publisher worked it out itself, or took the ad platform's word for what its ads achieved.

Platform engagement and real effectiveness only line up in about one out of three creator ads, so most engaging content isn’t what will grow your brand

Most marketers are making or planning changes to performance measurement for the AI web, mainly tracking brand visibility and citations in AI tools

AI's impact on planning and measurement approaches: current/planned changes to performance measurement for the AI web
Change	Share of respondents (%)
Measuring brand visibility/citations in AI tools	48%
Using branded search/direct traffic as proxies	44%
Using 3P AI-driven discovery analysis tools	40%
Monitoring AI platform interactions/referrals	31%
Increasing use of incrementality tests	30%
Using modeled measurement (probabilistic)	30%
Refining fraud detection for legitimate agents	28%
Placing less weight on website traffic	26%
Establishing direct AI-platform partnerships	23%
Using consumer surveys or self-reported data	21%

Open full graph

Buyers often use Google Search or visit a cited website directly after seeing a brand mentioned in an AI answer

Go deeper on this topic

Our Substack turns the data behind these cards into full breakdowns: what changed, why, and what to do about it.

Fixing your influencer marketing campaign: How to get into the top 10% of advertisers

151

is the long-term ROI index for influencer campaigns in the UK, contributing 6.2% of total sales (excluding campaigns that produced no result)

Marketers are much more likely to track engagement rate than efficiency metrics, with engagement monitored by 91% but EMV per working spend tracked by only 9%

Share of marketers tracking each influencer campaign KPI (%)
KPI	Tracked by (%)
Engagement rate (likes, comments, shares) — How many people interacted	91%
Impressions / reach (How many people saw it)	84%
Click-through rate (How many people clicked)	78%
Sales / revenue attribution (Whether people bought)	72%
Customer acquisition cost (CAC) — What each customer cost	72%

Open full graph

71% of marketers say proving ROI is their biggest influencer challenge

Top challenges marketers face with influencer marketing (% citing each)
Challenge	Share of marketers citing this challenge (%)
Determining the ROI	71%
The time it takes to run programs	50%
Content quality	48%
Selecting the right creators	46%
Working with less budget	46%
Consistency in influencer rates	35%
Scaling my influencer program	35%
Internal hurdles (leadership support, legal and regulatory)	25%
Brand safety	10%
Choosing the right agencies	8%

Open full graph

The source-state term made up 66.34% of the total GEO effect in the controlled design

Good for

Not for

Related cards

Go deeper on this topic