Meridian15 / Research

The data, not just the finding.

Every original study we publish ships the corpus it was drawn from. No email, no form, no summary standing in for the file. Take the CSV, check our arithmetic, and tell us if we got it wrong.

1. We Analyzed 330 Results in the AI Search Tools Category. Half of It Is Vendors Ranking Themselves.

   We pulled 20 US Google SERPs across the AI search tools category and classified all 330 organic results. Every query carries an AI Overview. 51.5% of results come from vendors selling a tool in the category they rank for.

   1. 100%Queries in the category that carry an AI Overview
   2. 51.5%Organic results published by a company selling a tool in the category it ranks for
   3. 53.2%Share of the top ten positions held by those vendors
   4. 172Distinct domains across 330 results
   5. 66.3%Domains that appear exactly once

   Sample

   330 organic results classified

   Collected

   August 1, 2026

   Source

   DataForSEO SERP API

   Dataset

   330 rows × 4 columns

   metrics.csv 32 KBmethod.jsonRead the study

2. AI Overviews Took Over One Category Completely and Left Another Untouched. We Measured 38 Search Results Pages.

   We pulled 38 Google results pages on one day. AI search tool queries returned an AI Overview 20 out of 20 times. Local service queries returned one zero out of 18 times, in both the US and Canada.

   1. 20 of 20AI search tool queries that returned an AI Overview
   2. 0 of 18Local service queries that returned an AI Overview, in both the United States and Canada
   3. 17 of 18Local service queries that returned a local pack instead
   4. 2Categories measured on one day with opposite result architectures and no overlap

   Sample

   38 search pages measured

   Collected

   August 1, 2026

   Source

   DataForSEO SERP API

   Dataset

   38 rows × 5 columns

   metrics.csv 2 KBmethod.jsonRead the study

3. Wix vs Shopify vs WordPress vs Webflow. What Each Website Builder Actually Costs a Calgary Business.

   We joined 5 audit datasets across 350 Calgary business homepages and cross-tabbed by builder. Every platform makes you pay somewhere. Here is exactly where.

   1. 50%Calgary business homepages built on WordPress
   2. 100%Wix sites with valid schema, at a median 185KB of HTML and 63 scripts per homepage
   3. 26%Squarespace sites with a tappable phone number
   4. 37%Custom-built sites that ship any schema, the leanest pages in the sample
   5. 5Audit datasets joined per homepage, speed, schema, contact, AI access and builder

   Sample

   403 homepages audited

   Collected

   July 21, 2026

   Source

   First-party crawl

   Dataset

   350 rows × 13 columns

   All sites hashed; this study reports platform-level medians and percentages only.

   metrics.csv 28 KBmethod.jsonRead the study

4. 41% of Calgary Businesses Cannot Be Phoned in One Tap. Every Dentist Can. Almost No Cafe Can.

   We parsed 403 Calgary homepages for tel: links, forms, and booking widgets. 41% cannot be phoned in one tap from mobile. Dentists lead. Cafes bottom.

   1. 41%Calgary business homepages with no tel: link, so the phone number cannot be tapped
   2. 97%Dental practices with a tappable phone number, the best vertical
   3. 12% to 17%Cafes and retail with a tappable phone number, the worst verticals
   4. 3.4%Sites that ship a booking widget
   5. 6Sites with a tel: link, a form, a booking widget and a viewport tag, every one of them a med-spa

   Sample

   403 homepages audited

   Collected

   July 21, 2026

   Source

   First-party crawl

   Dataset

   350 rows × 9 columns

   Winners (tel + form + booking + viewport all present on homepage) are named. All other sites are hashed.

   metrics.csv 21 KBmethod.jsonRead the study

5. 72% of Calgary Sites Have Structured Data. Only 61% Told Search Engines They're a Business.

   We parsed JSON-LD on 403 Calgary business homepages. 72% ship structured data. Only 61% marked themselves as a business. 91% skip FAQPage.

   1. 72%Calgary business homepages that ship any JSON-LD structured data
   2. 61%Declare a LocalBusiness or business subtype
   3. 15%Ship JSON-LD that only says the page is a WebPage
   4. 91%Have no FAQPage schema with a populated mainEntity
   5. 15Sites with fully valid schema, a business type and sameAs links, named in the study

   Sample

   403 homepages audited

   Collected

   July 21, 2026

   Source

   First-party crawl

   Dataset

   352 rows × 9 columns

   Winners (fully valid schema + LocalBusiness/Organization + sameAs, ranked by types count) are named. All other sites are hashed.

   metrics.csv 46 KBmethod.jsonRead the study

6. The AI-Blocking Myth: 403 Calgary Sites Later, Almost No One Blocks AI (and the Ones Who Do Picked Wrong)

   We fetched robots.txt for 403 Calgary business homepages. Only 4% block any AI crawler. Most blocks target training bots, which do not stop live AI citation.

   1. 4.2%Calgary business homepages that block any AI crawler in robots.txt
   2. 2%Block a live-retrieval bot, the kind that fetches a page to answer a question
   3. 6 sitesBlock every crawler by accident with a wildcard rule
   4. 70 of 403Have no robots.txt at all
   5. 2 of 2Popular blocking rules that do not do what the advice claims. GPTBot does not remove a site from ChatGPT answers and Google-Extended does not remove it from AI Overviews, per each operator’s own documentation

   Sample

   403 homepages audited

   Collected

   July 21, 2026

   Source

   First-party crawl

   Dataset

   333 rows × 9 columns

   This post has no winner ranking. Every site is listed anonymized: block posture is reported without naming the business, because blocking or not blocking is a defensible business choice either way.

   metrics.csv 20 KBmethod.jsonRead the study

Reviews of other people's research are taggedResearch on the blog. Only studies where we collected the data ourselves appear here, because those are the only ones with a corpus of ours to hand over.