Imagine someone asks ChatGPT to recommend three companies in your sector. They compare options, ask follow-up questions and choose the one they trust most. Days later, they find that company via Google and make an enquiry.
Your analytics credits organic search for the visit, or perhaps direct traffic. The AI conversation that helped win the customer? Nowhere to be seen. So how do you measure marketing’s impact when you can’t see the full customer journey?
AI search makes this problem harder to ignore. People can discover businesses, compare products and make decisions through AI answers without visiting a website.
Why AI search challenges marketing attribution
Marketing has always struggled to connect activity with sales. TV research could estimate audience exposure but not reliably isolate its contribution to sales. Econometric modelling and marketing effectiveness research developed to address this problem.
Digital analytics made customer journeys easier to track, but uncertainty remained. AI search brings marketers back to the challenge of combining different signals to understand what contributes to sales, rather than assigning every interaction a precise share of the credit.
Charlotte Deacon, Client Services Director at Oban, says: “We’re talking to clients more and more about how to understand what their marketing is really delivering, especially as customer journeys become harder to track. Just because we can’t measure something doesn’t mean it hasn’t influenced a decision. With AI changing how people discover and compare brands, it’s even more important to look across channels and join the dots to understand what’s contributing to business results.”
Three AI search measurement challenges
AI search creates three measurement challenges:
- Traffic is an incomplete picture. People can discover and assess a business through AI answers without visiting its website, so fewer organic sessions don’t necessarily mean less visibility.
- Conversions can be credited to the wrong channel. Someone influenced by AI might later arrive through branded search, paid advertising or a direct visit. Analytics records the route, not necessarily what created the demand.
- Visibility isn’t the same as value. Appearing in AI answers doesn’t establish whether the right audiences see the brand or whether that exposure generates business.
The answer is to combine different signals, understand their limitations and look for patterns that help explain business performance.
How to measure AI search performance
A practical measurement framework starts with three questions:
1. Can AI systems find and understand your content?
Before tracking AI visibility, check that AI systems can access and understand your content. Audit four areas:
- Technical access: Can AI systems crawl relevant pages, or are they blocked by robots.txt rules or server issues?
- Content quality: Does your site answer potential customers’ questions clearly and specifically?
- Evidence of expertise: Does your content explain what you do, what you offer and why customers should trust you?
- External credibility: Is your business represented accurately by credible sources beyond your website?
AI systems draw on information from across the web, so your website is only part of the picture. What credible external sources say about your business can also influence how it is represented.
What to do: Establish a baseline, fix the most important technical and content issues, and review the results after significant changes. Oban’s SEO services can help strengthen discoverability.
2. Does your business appear in the answers that matter?
Understand how your business appears in AI answers to questions potential customers might ask, from initial research to product comparisons and recommendations.
Analyse brand mentions, citations, recommendations and the competitors that appear. Look at differences across AI platforms, countries and languages to understand where your business is visible and where competitors feature more prominently.
A brand visibility report can provide a useful starting point, showing how often your business appears in relevant AI answers compared with competitors. It gives you a clearer picture of your current visibility and helps identify areas for further investigation.
What to do: Establish a baseline using commercially important topics, identify gaps and track changes alongside other visibility measures and business results. Oban’s AI visibility reporting can help you do this.
3. Is that visibility contributing to business performance?
No single measure can reliably isolate AI search’s contribution. Treating any signal as definitive risks false precision, particularly when AI may influence a decision long before a website visit. Combine several measures, considering what each reveals and what it can’t tell you.
| Measure | What it can tell you | The limitation |
| Branded search | Changes in interest in your business | Other marketing and PR can drive demand |
| Direct traffic | Changes in visits where the source isn’t identified | Doesn’t reliably identify AI-influenced visits |
| AI referral traffic | Visits identifiable as coming from AI platforms | Misses visits without a detectable referral |
| Leads, sales and revenue | Changes in commercial outcomes | Many factors influence performance |
| AI visibility | Changes in presence in relevant AI answers | Doesn’t prove commercial impact |
Interpret these measures together rather than treating any one as proof. If AI visibility, branded searches, identifiable AI referrals and qualified enquiries all increase, that is a useful pattern to investigate, but it doesn’t prove AI visibility caused the increase.
If visibility improves but enquiries remain flat, investigate whether you’re appearing for the right topics and reaching the right audience, whether you’ve allowed enough time for results to emerge, and whether other factors are affecting conversion.
What to do:
- Define commercial outcomes. Agree whether success means qualified leads, sales, revenue or pipeline value.
- Establish a baseline. Record visibility and business performance before significant changes.
- Monitor measures together. Track AI visibility, branded search, identifiable referrals and commercial outcomes.
- Account for other influences. Record significant changes in paid media, PR, pricing, promotions and seasonality.
- Review the evidence. Distinguish between what you observed, what the data suggests and what you have demonstrated.
Where the data allows, marketing mix modelling, econometric analysis and controlled experiments can help estimate marketing’s incremental contribution. These methods can support better decisions than last-touch attribution alone.
AI search measurement across channels and markets
AI search measurement shouldn’t sit in an isolated report. SEO teams need to understand how traditional rankings, website traffic and AI visibility relate to one another. Paid media teams need to consider how demand created earlier in the customer journey may eventually be captured through advertising or other channels.
Oban’s integrated search approach brings these perspectives together, linking SEO and paid media performance to wider marketing objectives.
AI responses vary across markets and languages, so visibility in one country doesn’t guarantee visibility in another. Oban’s Local In-Market Experts help assess how visibility, competitors and customer needs differ across markets and languages.
Measure what you can, and be honest about what you can’t
AI search makes it harder to trace the full customer journey. The aim isn’t to achieve perfect attribution, but to use the available evidence to make better marketing decisions. If you want a clearer picture of how AI search affects your brand, contact Oban to discuss how our SEO, Paid Media and AI visibility expertise can help.
Let’s accelerate action together
At Oban, we believe change happens when we act, support each other, and keep moving forward. These stories show how small steps can make a big difference. If you want to improve your digital marketing, get in touch. Let’s get started.
