AI Overviews for SMEs: Adjusting Your Content When Google Answers for You

Google's AI Overviews now sit at the top of a large share of informational searches. When one appears, the user reads the answer and often does not click anything. For a business with ten to ninety-nine people, that shows up as a slow bleed in your organic numbers that nobody can quite explain, and eventually as a question from the owner or finance about why the content budget is still there. This post is written for the person who has to answer that question: a marketing generalist or a small team, working with a real but bounded budget. It covers where the data lives in Search Console, how to read it without panicking, and the repeatable process that keeps your organic program defensible instead of reactive. The playbook it is built on lives in the AI Overviews data guide.

#What is actually happening, in plain terms

When someone searches an informational question, Google increasingly generates a short synthesized answer at the top of the page and pulls from a handful of sources to build it. The user gets what they came for. They do not scroll to your result, and they do not click. Your page can still be ranking well and appearing in the panel as a cited source, and your clicks still drop, because the answer was already delivered above your link. This is not a penalty and it is not something you did wrong. It is a change in how the results page works, and it hits informational content hardest.

The distinction that matters for a small team is informational versus transactional. Informational searches ask what something is or how it works. Transactional searches carry buying intent: a service plus a city, a cost, a comparison, someone ready to hire or buy. AI Overviews fire heavily on the first group and rarely on the second. That single fact shapes your whole response, so hold onto it.

#Where the data lives in Search Console

You do not need a new tool or a new budget line to see this. Google Search Console already has it, and Search Console is free. Open the Performance report. It shows clicks, impressions, click-through rate, and average position for every query and page. The Performance report has a SERP-feature filter, and "AI Overview" is now one of the options in it, alongside things like Top stories and Image pack.

1. Open the Performance report and set the date range to sixteen months so you can see before-and-after, not just the last month.
2. Apply the SERP-feature filter for AI Overview. This shows the queries where one of your pages appeared inside an AI Overview panel.
3. Note which of those queries are informational and which carry buying intent. The informational ones are where the click loss concentrates.
4. Export the list. This becomes the working sheet your process runs against every month.

#How to read it without over-reacting

The most common mistake a small team makes here is to see a click dip, assume the whole SEO program is broken, and either cut the budget or thrash the strategy. Neither is right. Read three signals instead, and read them against the sixteen-month view so you are looking at trend rather than a noisy week.

• Impressions holding steady but clicks down on informational queries. This is the AI Overview signature. Google still shows your page, but the answer above it is absorbing the click. This is the pattern to respond to with content, not with panic.
• Impressions and clicks both down on the same queries. This is a ranking problem, not an AI Overview problem. Different fix, usually a competitor who improved or a page that went stale.
• Transactional queries flat or up. This is the good news you should be leading your internal report with. The searches that produce real leads are largely untouched.

For the interpretive traps that catch small teams, the source guide walks the full picture. The short version: separate informational from transactional before you conclude anything, because the same click drop means two completely different things depending on which bucket it sits in.

#The numbers from real client data

Across the Frontend Horizon client book, the pattern has been consistent enough to plan against. Use these as directional benchmarks for your own read, not as promises.

• AI Overviews appear on roughly 28 percent of the informational queries our clients rank for.
• They appear on under 2 percent of transactional queries such as a service plus a city or the cost of something in a place.
• When an AI Overview is present, average click-through on the underlying organic result drops by 40 to 60 percent, even at position one.
• Blog content built around broad informational queries has seen the steepest click loss.

The math a small team should carry into the budget conversation is this: if about 28 percent of your organic traffic came from informational queries and a chunk of those now trigger AI Overviews, you are looking at somewhere in the range of a 12 to 17 percent hit on total clicks if you do nothing. That is a real number, and it is also a manageable one, because the response is a content shift you can run in-house rather than a tooling purchase.

#Build a repeatable process, not a one-time cleanup

This is the part that separates a small team that stays defensible from one that keeps getting surprised. The response to AI Overviews is not a single project. It is a monthly loop your generalist can run in a couple of hours. Write it down so it survives the person who set it up.

1. Pull the AI Overview query list from Search Console and mark each query informational or transactional.
2. On the informational queries losing clicks, decide per page: convert it, or leave it. Convert the ones tied to your buyers; leave the pure-definition pages alone.
3. For each page you convert, add the things an AI Overview cannot easily synthesize: your own numbers, your own examples, local specifics, a real point of view, a comparison the buyer needs.
4. Stop commissioning new generic explainer content. Redirect that budget toward transactional, local, and provider-specific pages.
5. Record the click change on the pages you touched. That before-and-after is your ROI evidence for finance.

#What to write more of, and what to stop writing

The strategic shift is simple to state and disciplined to hold. Move your content weight toward the query types AI Overviews barely touch, and stop spending on the ones they now answer for you.

#Write more of this

• Local content. Pages about the specific cities, neighborhoods, and regional concerns you serve. AI Overviews do not synthesize local well.
• Transactional content. Cost guides, pricing pages, comparison pages, and how-to-hire guides where the reader is making a decision, not just learning a definition.
• Provider-specific content. Case studies, examples of your actual work, and your process. An AI Overview can summarize a topic; it cannot cite your experience.
• Opinionated content. Take a position. The synthesized answer gives the consensus. A clear point of view still earns the click because the reader wants a human take.

#Stop writing this

Stop commissioning generic informational content that competes directly with the AI Overview. Pages like "what is X" or "how does Y work" now get a synthesized answer at the top, and clicks are minimal even when you rank first. For a small team, the cost here is not just the wasted effort. It is the opportunity cost: those writing hours and that budget could have gone to a transactional page that still earns visits and leads.

#Do not delete your old content, convert it

If you already have a library of informational posts that has lost clicks, resist the urge to delete it. Deletion throws away rankings, internal links, and the authority those pages carry. The better move for most pages is conversion: keep the topic, and add the layers an AI Overview cannot reproduce. Add your own data. Add local specificity. Add a comparison framework. Add a real opinion. The same general subject with a strong provider point of view often keeps winning the click, because the reader wants the expert take rather than the averaged one.

Be selective about which pages get the effort. A finite budget means you cannot convert everything, so rank the list by which pages sit closest to your buyers. A losing informational page that is one step from a service you sell is worth converting. A losing definition page nobody buys from is fine to leave alone.

#In-house or outsource this?

This is the decision an SME actually has to make, and the honest answer is: mostly in-house, with a narrow case for outside help. The monthly Search Console loop and the content conversions are well within reach of a competent marketing generalist. You do not need to hire against this, and you should be wary of any vendor selling AI Overview recovery as a large managed retainer. The work is real but it is bounded.

Where outside help earns its cost is the initial setup and the strategic reframe: getting Search Console reading cleanly, deciding which content to convert versus retire, and building the repeatable process so your team can run it without a specialist. That is a one-quarter engagement, not an open-ended one. It is the same audit-to-process shape we use on our own solutions, and it is deliberately scoped so it hands back to your team.

#Defending the spend to the owner or finance

The whole point of running a real process is that it produces evidence. When the owner asks why the content budget survives in a world where Google answers for free, you want numbers, not a shrug. Bring four things to that conversation, and you turn a defensive question into a confident answer.

• The transactional-versus-informational split. Show that the searches producing your leads are largely untouched by AI Overviews, and that this is the traffic you are protecting.
• The conversion evidence. For the pages you rewrote, show the click number before and after. This is your proof the process works, not a theory.
• The budget reallocation. Show that you stopped spending on generic explainers and moved that money to transactional and local pages that still earn visits. You did not ask for more; you redirected what you had.
• The lead tie-back. Connect organic landing pages to actual leads, so the report is about revenue, not vanity traffic. A page with fewer clicks that produces qualified leads beats a high-traffic page that produces none.

#The part that is genuinely good news

It is easy to read all of this as loss, but for a business doing real work there is a real upside. AI Overviews clear thin, generic informational content out of the results page. That was the content mass-produced by competitors who never had anything specific to say. As it loses its clicks, the room it occupied opens up for content with actual expertise, real local relevance, and a genuine point of view. The bar moved from writing enough content to be present to writing content good enough to earn the click anyway. A small team already doing the real work is well positioned to clear that bar, and many larger competitors coasting on generic volume are not.

#Where this fits with the rest of your SEO

None of this replaces your core SEO. AI Overviews sit on top of the results page you were already optimizing for, and the transactional, local, and provider-specific pages you are shifting toward are the same pages that carry your leads. If your business is a considered-purchase service, the buyer research is where the durable advantage lives, which is why we anchor engagements around professional services and the questions real buyers ask. The content shift and the buyer knowledge reinforce each other.

The same change reads differently at other sizes. If you run an agency delivering this across a client book, the agency version covers explaining the loss to clients and winning the work back. If you are an owner-operator doing it yourself with no marketing staff, the micro-business version covers the cheapest high-return steps for this week. If you sit in a larger organization with stakeholders and a martech stack, the mid-market version covers governing this at scale.

#A short checklist to run this quarter

1. Confirm Search Console is set up and the Performance report has sixteen months of history.
2. Pull the AI Overview query list and split it into informational and transactional.
3. Rank your losing informational pages by closeness to your buyers, and pick the top handful to convert.
4. Convert those pages by adding your data, your examples, local detail, and a point of view.
5. Stop the generic-explainer commissions and redirect that budget to transactional and local pages.
6. Write the monthly loop down as a process, assign an owner, and put the first review on the calendar.
7. Build the four-part budget-defense report so the spend conversation is evidence, not opinion.

Google's own documentation on how it discovers and ranks content is the primary reference for the technical side of this, and it is worth reading once so your team is working from the source rather than secondhand advice (Google Search Central). For the market picture on how answer engines are reshaping search behavior, HubSpot's answer-engine research is a solid, current read.

If your content plan still leans heavily on generic informational content and you want the shift handled once and handed back, run the estimator and we will scope the realignment for your team and budget. Or book a consultation and we will read your Search Console data with you and mark the pages worth converting first.