Local Directories Are Back – Now They’re Feeding AI
AI search isn’t showing customers a list to click from anymore — It’s choosing for you. In 2026, 68% of Google searches end without a click, and Alberta small businesses are discovering that their directory listings, not their websites, are what the algorithm checks first.
For about a decade, the standard advice handed to Alberta small businesses went something like this: put your money into the website. That’s where customers land, that’s what converts, and everything else is secondary.
Your Yelp page was something you dealt with when a bad review showed up. Your GBP was something a nephew set up in 2019 because he was good with computers, and then nobody looked at it again.
That whole model has come apart. Quietly at first, and now faster than most people realize.
Where the Clicks Went
In the first four months of 2026, 68% of Google searches in the United States ended without a single click. Not a click to your website — a click to anything. The figure comes from a June 2026 SparkToro study using Similarweb clickstream data, and it’s the steepest two-year jump in that metric since anyone started tracking it. In 2024 the number was 60%. Ten years before that, around 45%.
What changed is Google’s AI Overviews — the generative summaries that now sit at the top of results pages and answer the question before the links even register. When an Overview appears, click-through rates fall by nearly 60%. The search resolves inside Google. Nobody goes anywhere.
Canada follows the U.S. closely enough here that the direction holds. A Calgary homeowner looking for a licensed gas fitter on a Tuesday evening isn’t tabbing through six websites. She’s reading a synthesized paragraph, maybe tapping a map pin, and calling whoever the AI surfaced. The businesses that didn’t make that answer don’t get a second chance at that particular moment.
Bain & Company tracked what this costs commercially: organic web traffic is down 15% to 25% across multiple sectors. For a trades operation in Fort McMurray or a physiotherapy clinic that lives on booked appointments, that isn’t an abstraction. It’s volume that existed before and doesn’t anymore.
One Answer, Not a List
The part that tends to surprise people when they actually hear it: AI assistants aren’t giving customers options. They’re making a selection.
Google gives you a list — The Local 3-pack followed by the SEO 10 – pick one, you decide.
ChatGPT, Gemini and Perplexity don’t work that way. They synthesize what they can find, reach a conclusion, and present a few options. The shopper sees just 1 to 3 results of a decision that happened without them.
SOCi’s 2026 Local Visibility Index put hard numbers on how selective that process is. The study looked at more than 350,000 business locations across nearly 2,800 brands. ChatGPT recommended 1.2% of them. Gemini was more generous — 11%. Perplexity came in at 7.4%. Google’s local 3-pack, for comparison, surfaces around 36% of businesses in any given category.
So a Calgary electrical contractor with a decent Google Maps presence and solid word-of-mouth can be entirely absent from AI recommendations. Not buried. Not penalized for something. Just not there.
The SOCi research gets at the reason. These platforms aren’t ranking web pages. They’re evaluating whether they trust a business enough to stake a recommendation on it — and they establish that trust by cross-referencing identity data across platforms. Google Business Profile, Apple Maps, Yelp, whatever industry directories exist.
An Edmonton accounting firm that updated its address on its website when it moved off Whyte Ave but left the old address sitting on three directory platforms it forgot about looks, to the AI, like something that doesn’t fully add up. So it doesn’t get recommended.
The Signals That Actually Matter
Consistent name, address, and phone number across every platform is the starting point, and it sounds almost too simple to be the thing that’s costing people business. But a misspelled street name on one listing, a phone number that forwards to a line nobody checks, hours that reflect a pre-pandemic schedule — any of it registers as a data conflict.
AI is looking for corroboration across sources. When the sources don’t agree, confidence drops.
Reviews are the second piece, though not quite in the way most business owners assume. A massage therapist in Leduc sitting on 94 reviews that say “lovely atmosphere, very relaxing” is at a real disadvantage compared to a competitor with 40 reviews where customers got specific — named the treatment, described what they came in for, said what changed.
AI systems scan review language for what researchers call micro-intent keywords. Someone asking for a physio in St. Albert who does dry needling isn’t just getting category matches. The system is looking for reviews where an actual patient confirmed that specific thing happened at a specific clinic.
And then there’s recency. A Lethbridge roofing company with 200 reviews earned between 2018 and 2022 and eight since reads as dormant. Fresh reviews signal that a business is actively operating. That matters more than people expect.
The Website Became the Back Office
The customer journey Alberta businesses spent years optimizing — get found on Google, land on the site, convert — has compressed in ways that make a lot of that investment beside the point now.
Pricing, wait times, booking links, hours, promotions: these increasingly live inside directory integrations, not on company websites. Someone asking ChatGPT for a walk-in clinic in Calgary’s northeast gets a phone number, current hours, a star rating, and a map link returned in the response itself. The clinic’s website plays no role in that transaction.
The signals the AI is reading — how a business responded to a negative Yelp review last month, whether the Google Business Profile hours are current, how recent the photos are — these things used to be reputation management tasks that lived at the edge of the marketing budget. They’ve moved to the center of whether a business shows up at all.
Alberta’s Best follows how local businesses surface in AI-generated search results, and it keeps coming back to the same trust points. The ones getting recommended aren’t always the highest-rated or the longest-established. They’re the ones whose directory presence is clean and who’s storefronts look active online.
So for the esthetician in Grande Prairie, the Roofer in Airdrie, or the Vet in Red Deer — their antiquated listing that nobody bothered to update for four years has become the very things AI checks first.
Customers aren’t finding it. But AI is reading it, and deciding whether to send customers your way.
Sources
SparkToro / Similarweb — 2026 Zero-Click Search Study (June 2026): https://sparktoro.com/blog/in-2026-less-than-one-third-of-google-searches-still-send-a-click/
Search Engine Land — Zero-Click Searches Hit 68% in Early 2026: https://searchengineland.com/google-zero-click-searches-2026-study-479717
Bain & Company — Consumer Reliance on AI Search Results (February 2025): https://www.bain.com/insights/goodbye-clicks-hello-ai-zero-click-search-redefines-marketing/
SOCi — 2026 Local Visibility Index: https://www.soci.ai/blog/the-challenge-of-ai-visibility-for-brands-part-1/
Search Engine Land — AI Local Visibility Report 2026: https://searchengineland.com/ai-local-visibility-report-2026-468085
