Organic Traffic AI Exposure Audit
Classify search traffic to reveal which content AI will cannibalize first.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 96%
The audit converts an undifferentiated content portfolio into an intent-based risk map. Begin by assigning each page and its organic traffic to informational, navigational, branded, local, commercial, or transactional categories. Because informational content varies in defensibility, divide it further into generic answers, education with perspective, unique data, factual references, and opinion. Calculate the share of traffic in each group, treating generic informational queries as the most exposed to AI summaries. Establish a baseline, then monitor category-level clicks and impressions over time as AI search features expand. The output is not merely a traffic-loss forecast: it is a prioritized portfolio showing which pages should be enriched, consolidated, repositioned, or monitored and which commercial or differentiated assets are comparatively resilient.
Origin
Extracted from Marketing Against The Grain. Kieran Flanagan credited SEO consultant Aleyda Solis with a template for classifying content and described HubSpot's plan to add more nuanced informational subcategories.
Core principles
- 01Search intent determines exposure to AI-generated answers.
- 02Generic informational content carries the greatest near-term risk.
- 03Unique data, perspective, and opinion can distinguish educational content.
- 04Traffic must be segmented before its disruption can be measured.
- 05Repeated measurement is more useful than a one-time estimate.
How to run it
- 1
Build the content inventory
Export every indexable content URL with organic clicks, impressions, conversions, and other available performance measures. Use a consistent reporting period.
Pro tip Include page templates and primary query themes to make classification faster.
Watch out Excluding low-traffic pages can hide a large aggregate exposure across the long tail.
- 2
Classify primary intent
Assign each page to informational, navigational, branded, local, commercial, or transactional intent according to the searcher's dominant goal.
Pro tip Use the queries generating actual impressions rather than relying on titles alone.
Watch out Do not classify pages solely by funnel stage; intent and funnel position are related but different.
- 3
Subdivide informational content
Separate generic answers from education containing unique data, perspective, opinion, or other distinctive value. These subcategories reveal different levels of AI exposure.
Pro tip Create mutually exclusive labels and document examples for reviewers.
Watch out A page is not differentiated merely because it is long.
- 4
Quantify category exposure
Aggregate traffic and conversions by category and calculate how much depends on vulnerable informational queries. Rank categories and pages by potential business impact.
Pro tip Weight conversion contribution alongside raw sessions.
Watch out Traffic loss is not equally important when pages have different commercial value.
- 5
Establish a performance baseline
Record category-level clicks, impressions, click-through rates, rankings, and conversions before broader AI-search changes occur.
Pro tip Annotate known search feature launches in the reporting timeline.
Watch out Without a baseline, seasonality or ranking changes may be mistaken for AI cannibalization.
- 6
Monitor and respond
Repeat the analysis periodically to identify categories declining faster than the portfolio average. Enrich, consolidate, or reposition the highest-value exposed content.
Pro tip Test whether unique evidence and perspective improve resilience before rewriting the entire library.
Watch out Do not delete informational content based only on a forecast; validate the trend with observed data.
In the wild
A SaaS company classifies its organic library by intent, then splits informational pages into generic definitions, tactical education, proprietary research, and opinion. It discovers that generic definitions supply 35% of visits but only 8% of assisted conversions, while proprietary research drives fewer visits but much higher conversion value.
→ The team monitors generic pages closely and directs investment toward differentiated education and research.
A multi-location service company separates local landing pages, branded searches, buying guides, and generic how-to articles. The audit shows that most leads come from local and commercial pages, while AI exposure is concentrated in informational articles with little conversion value.
→ The company protects high-value local assets while selectively upgrading vulnerable educational content.
Common mistakes
Treating all education alike
Generic answers and education supported by unique evidence have different exposure profiles and should not occupy one undifferentiated bucket.
Measuring traffic without value
Raw visits can exaggerate risk when vulnerable pages contribute few leads, sales, or assisted conversions.
Running the audit only once
AI search coverage and user behavior evolve, so a static classification cannot reveal the timing or scale of actual disruption.
Is it for you?
Best for
It is best for publishers and businesses with substantial organic traffic distributed across informational, commercial, transactional, branded, navigational, and local intent.
Not ideal for
It is not ideal for businesses with negligible search traffic or no page-level analytics history.
From the transcript
“you would want to look at how much of your traffic is in that informational bucket because that is the category at risk”
“you can start to break apart your content and your traffic in this way”
“that educational content bucket regardless of how you break that out is going to be up for cannibalization”
From the episode
Look Dumb or Get Disrupted: Google's Risky AI Overviews Launch