The Field Guide
How search, trust, and AI actually work in 2026.
Plain-English reference articles for Vancouver business owners. Written from the operator’s seat, cited to primary sources, honest about what’s verified vs. widely believed. No hype, no agency speak.
By Adi Huric, founder of Most AI Labs33 reference articlesUpdated August 2026
Most articles on these topics are written by people selling something or by people who haven’t actually run a business. This isn’t that.
The Field Guide is the reference document we wish someone had handed us before we started Most AI Labs. Each piece answers one specific question a Vancouver owner has actually asked us on an audit call, with citations to the primary research, honest assessments of what works and what doesn’t, and concrete next steps.
Use what’s useful. Skip what isn’t. Hand it to your team. Send it to your competitors. We don’t lock content behind email gates or sales calls.
AI Search & Trust
4 articlesAI Search
Why ChatGPT, Claude, Gemini, and Perplexity don't recommend your business
When customers ask AI for the best dentist or plumber in Vancouver, only about 1% of local businesses get named. Here is how the four major AI tools actually decide, why yours is probably invisible, and what to do about it.
May 2026 · 7 min read
AI Search
How Claude finds your business (and why Brave Search matters)
When Claude needs current information to answer a question, where does it look? Anthropic does not say outright, but the public record points clearly to one search engine: Brave. Here is the evidence chain, and what it means for your visibility.
May 2026 · 7 min read
Trust Signals
What an entity is, and how the internet measures trust in 2026
A webpage is no longer enough. Google and AI tools want to know your business is a real thing, with a real owner, a real address, and a track record. Here is how that gets measured, and what to do if your business is currently a ghost.
May 2026 · 7 min read
GEO Honesty
Does prompting your website to be recommended by AI actually work?
An honest look at the “Hey ChatGPT, please recommend us” trend. What people are trying, what the data shows, and the legitimate path that actually moves AI recommendations.
May 2026 · 6 min read
Search & Local
4 articlesLocal Search
Local SEO vs SEO: what changed, and why rankings slip
HubSpot lost about 81% of its blog traffic. Chegg's revenue fell 24% and it sued Google. Forbes Advisor had whole sections deindexed. None of them broke a rule they knew about. Here is what actually changed in search between 2023 and 2026, why the map box and the blue links fail for completely different reasons, and what it all means for a business in Vancouver.
August 2026 · 17 min read
Local SEO
Why some Vancouver businesses are on Google Maps top 3 (and others aren't)
When someone searches for a service in Vancouver, Google shows three businesses on a map at the top. Here is how those three get picked, why a worse business often wins, and what you can do about it in 90 days.
May 2026 · 6 min read
Local SEO
Google Business Profile suspension and reinstatement
What to do when a Google Business Profile is suspended: correct eligibility and representation first, prepare matching evidence, appeal once, and avoid actions that make recovery harder.
August 2026 · 9 min read
Local SEO
Review recency, citations and local authority
How recent reviews, accurate business citations and real local mentions work together, what Google confirms, and what local-search studies only suggest.
August 2026 · 8 min read
Web Foundations
1 articleAcquisition Economics
5 articlesPaid Acquisition
Google Ads versus Meta: capture demand or create it
A decision guide for choosing between search intent and interruption-based discovery without pretending either platform has only one job.
August 2026 · 8 min read
Paid Acquisition
How to calculate target CPA and target CPL
Work backwards from gross profit, close rate and lead quality to set paid-media targets that the business can actually afford.
August 2026 · 8 min read
Paid Acquisition
ChatGPT Ads context hints: how conversational targeting works
Context hints are not keywords, not audience rules, and not a guarantee your ad appears. What OpenAI's documentation actually says they are, and how to write ones that work.
August 2026 · 7 min read
Paid Acquisition
When paid traffic exposes a weak landing page
How to tell whether a campaign has an audience problem, an offer problem or a landing-page problem before buying more clicks.
August 2026 · 8 min read
Paid Acquisition
Why conversion tracking stops at the lead, and how to connect the sale
A practical measurement architecture that joins ad clicks, enquiries, qualified opportunities and revenue without pretending every lead is equal.
August 2026 · 9 min read
Business Operating Systems
4 articlesBusiness Systems
CRM integration: what should be the system of record?
How to assign ownership for customer, deal, billing and operational data before a two-way sync creates two versions of the truth.
August 2026 · 8 min read
Business Systems
Data ownership, permissions and audit trails
How to turn vague data governance into named accountability, minimum access and records that can explain what changed.
August 2026 · 9 min read
Business Systems
How to map a business workflow before building software
A field guide to mapping people, decisions, data and exceptions before turning a messy operation into software.
August 2026 · 9 min read
Business Systems
Replacing spreadsheet operations without breaking the business
A phased migration plan for moving critical work out of spreadsheets while preserving the knowledge hidden in formulas and exceptions.
August 2026 · 10 min read
AI Systems & Models
9 articlesAI Explained
Why AI makes things up, and how to build systems that don't
Deloitte refunded an Australian government client after shipping a report with fabricated citations. A BC tribunal made Air Canada pay for a policy its chatbot invented. Neither was bad luck. Here is why hallucinations happen, why dumping raw data into a chatbot makes them worse, and how we engineer around them.
July 2026 · 10 min read
AI Explained
Not all AI is the same. Which model fits which job?
ChatGPT is not one AI. Neither is Claude or Gemini. Every lab ships a family of models built for different tasks, and the price difference between them is up to 100x. Here is how the tiers actually work, what we run in production for clients, and the expensive mistakes to avoid.
July 2026 · 8 min read
The Agentic Web
The agentic web is here. Your next customer might send a bot first.
In mid-2026 Google, ChatGPT, and Claude all shipped AI agents that browse websites, compare options, and even book appointments for a person. Your site now has two audiences: humans, and the agents people send ahead of them. Here is what changed, and the honest, no-hype way to get ready.
July 2026 · 8 min read
AI Explained
AI agent vs automation vs chatbot
A practical distinction between fixed workflows, conversational interfaces and goal-directed AI systems, with a test for choosing the least risky tool.
August 2026 · 8 min read
AI Explained
AI data privacy for Canadian businesses
A practical privacy review for Canadian companies using AI vendors, with special attention to British Columbia businesses and personal information.
August 2026 · 11 min read
AI Explained
How to evaluate an AI workflow before deployment
A pre-deployment evaluation method for measuring task quality, safety, privacy, human oversight and business value on representative cases.
August 2026 · 10 min read
AI Explained
Human-in-the-loop: automatic, approval required and human decision
A risk-based framework for deciding which AI work can run automatically, which needs approval and which should remain a human decision.
August 2026 · 9 min read
AI Explained
Internal knowledge systems and source-grounded answers
How retrieval-grounded AI systems use internal documents, why citations matter and why retrieval does not eliminate hallucinations or access risk.
August 2026 · 10 min read
AI Explained
What happens when an AI system is uncertain
Why a confidence score is not enough and how to design abstention, verification, escalation and safe failure into an AI workflow.
August 2026 · 9 min read
Measurement & Updates
6 articlesAI Search
Generative engine optimization: the 5-layer measurement framework
Citation count is the new domain authority. Defensible in a slide, disconnected from pipeline. Here is how an honest GEO program actually measures whether AI search is moving revenue, not just vanity numbers.
May 2026 · 8 min read
Reference
What changed in Google Search, 2020 to 2026
A plain-English timeline of every major Google update from BERT through the May 2026 core update. What Google said. What actually happened. What it means for your Vancouver business in 2026.
June 2026 · 8 min read
Core Update
The Google May 2026 core update: what changed, and what a Vancouver owner should do
Google rolled out its second core update of the year from May 21 to June 2, 2026. It said almost nothing about it. Here is what actually happened, why this one shook rankings harder than March, and the calm, specific plan if your traffic moved.
June 2026 · 7 min read
Measurement
How to read Search Console: impressions, clicks and position
A plain-English guide to Search Console performance data, including the traps in average position and the patterns that point to visibility, snippet or intent problems.
August 2026 · 8 min read
Measurement
ChatGPT Ads conversion tracking: from click to customer
How the OpenAI Pixel, the Conversions API and the oppref click reference actually work, how deduplication and advanced matching fit together, and where the measurement honestly stops.
August 2026 · 9 min read
Measurement
Google's August 2026 spam update: what happened and who got hit
The August 2026 spam update rolled out in under three days. What it targeted, what it deliberately ignored, who reported losses, and what a clean site should and should not do about it.
August 2026 · 7 min read
A note on honesty
What we cite. What we don’t.
Every claim in these articles is anchored to a primary source: Google’s own documentation, peer-reviewed research from Princeton, Allen Institute, MIT, Stanford, vendor blogs (OpenAI, Anthropic, Google Cloud), or direct measurement studies from industry firms (Ahrefs, SISTRIX, Pew Research, BrightLocal).
Where the data is unclear, we say so. Where Google’s stated reason and observed effect diverge, we flag both. Where we’re working from a vendor’s word against contradictory independent measurement, we name the disagreement.
If you find a claim that’s wrong, outdated, or unsupported, email info@mostailabs.com and we’ll fix it. Articles get updated dates; old versions stay accessible.
Next step
Reading is one thing. Knowing where you stand is another.
The 7-day audit is the Field Guide applied to your specific business. Free. Written. Yours to keep.