Search visibility
Google Still Matters in the AI Search Era
Use your buyer’s journey to set priorities
This article began in 2012 with a question about Google’s staying power. The current review addresses AI-assisted search. It is not a prediction that Google’s position, traffic or interface will remain unchanged.
For a business, the decision is where discovery produces useful attention and what a prospective customer needs after finding you. Measure that journey before moving effort from one channel to another.
- Discovery: where do suitable prospects first encounter the problem, product or business?
- Verification: where do they check specifications, examples, reviews or company details?
- Action: where do they request a quote, book or buy?
Ask customers as well as reading reports. One recorded referral may describe only the last part of their journey.
AI answers can be part of Google Search
Google’s AI Overviews and AI Mode can show links to supporting websites. Google says the normal SEO foundations still apply: supporting pages must be indexed and eligible for a search snippet. Inclusion is not guaranteed, and no special AI file or additional schema is required. See Google’s guidance for website owners.
That is a reason to keep useful pages accessible and accurate, rather than treating AI visibility as a separate collection of tricks. It does not establish how much traffic your site will receive.
Google’s guidance covers its own Search features. Do not assume that another assistant uses the same sources, access rules or reporting. Check the specific product before making changes for it.
A summary does not answer every buying question
Consider a hypothetical buyer looking for an industrial computer for an inspection station. A general answer might explain processor, enclosure and connection choices. The buyer still needs to check the particular supplier’s configuration and support.
- Fit: which camera interfaces and operating systems does the offered model support?
- Evidence: which test results apply to this exact configuration, under which conditions?
- Availability: which parts, replacement options and supply terms can the supplier confirm?
- Next step: what drawings, workload details or environmental requirements are needed for an assessment?
A product page that answers these questions gives the buyer a reason to inspect the source. A generic paragraph saying “reliable industrial solutions” leaves the evaluation unfinished.
This is a planning example, not evidence that an AI system will cite the page or send a customer. Design the page so it remains useful whichever route brings the reader.
Five checks for pages buyers rely on
A clear question
Write the decision the page helps someone make. Keep the main answer near the start and use headings that explain the remaining questions. Separate distinct buyer needs instead of publishing near-identical keyword pages.
Evidence that can be checked
Show real examples, methods, specifications or comparisons. Name the source and relevant author or reviewer. Distinguish observed findings from assumptions, and keep limitations beside the claim.
A working route to the answer
Open the page from its internal links and on a phone. Check that important text and images load and the intended action works. If a page intended for public search is missing, investigate its indexing and access before rewriting it.
Consistent business facts
Compare your service, product and contact information with the relevant profiles and listings you maintain. For an eligible local business, check Business Profile details such as hours, category and address. See Google’s local visibility guidance.
A reason to continue
Offer the detail needed for the next decision: a worked example, usable instructions, a relevant demonstration video or a clear assessment route. Keep product and comparison pages specific enough to support a real choice.
Read visibility and outcomes separately
Record whether people discover the business, visit the relevant page and complete a useful action. A ranking screenshot, citation or impression does not establish that a buyer understood the offer.
Google reports traffic from its AI Search features within Search Console’s Web performance data. That reporting context limits what you can infer about a particular feature from overall changes.
- Search visibility: inspect relevant queries, landing pages and changes in impressions and clicks.
- Recorded visits: compare available referral information, remembering that it may not show every prior interaction.
- Qualified demand: review whether enquiries fit the service and progress through the sales cycle.
- Customer accounts: ask how buyers found and checked the business. Keep their answer distinct from analytics attribution.
Investigate a decline before naming its cause
If visits fall, compare the affected pages, queries, devices and periods. Check whether demand is seasonal, pages moved, access changed, tracking broke or the result presentation changed. A decline during an AI rollout does not by itself prove that AI caused it.
Likewise, fewer visits with steady enquiries can mean something different from fewer visits and fewer suitable enquiries. Keep raw counts beside rates so a percentage does not hide a small sample.
Review a few important searches in the market and language you serve. Record the date and what appeared. Treat this as a sample, not a universal result or a forecast of future visibility.
Improve one important source page first
Choose a service, product, category, local or comparison page that affects a real decision. List the unanswered questions and supply the missing evidence. Correct inaccurate facts, broken links and the next action.
Keep legitimate routes beyond Google in view too: direct relationships, relevant publications, referrals and returning customers. Choose effort according to observed demand and your ability to help, rather than predictions about which platform will win.
Use the on-page review for the content and route, and the conversion diagnosis when relevant visitors stop before acting.