Twenty years of SEO has taught me more about human behaviour than any marketing course. Not because SEO is a psychology discipline — it isn’t — but because organic visibility is a direct reflection of how real people make decisions under real conditions. Pay close enough attention to that for long enough, and patterns emerge that no crawler can surface.
The argument in this piece is simple: good SEO strategy requires human judgement about context, not just machine-facing technical fixes. A consultant who looks only at crawlability and rankings can produce technically correct recommendations that are commercially wrong. Understanding why enquiries dropped, why conversions stalled, or why a well-ranked business is losing to a weaker competitor often has nothing to do with what is happening on the site — and everything to do with what is happening in the world, in the market, or in the mind of the person doing the searching.
Before diagnosing any visibility or performance problem, check four contexts first: trust, emotional, market, and platform. The rest of this article explains what each of those means in practice.
The signals your business sends before anyone reads a word
Before a potential client reads your homepage, checks your services, or evaluates your expertise — they have already made a preliminary judgement based on the signals your online presence sends in aggregate.
Consider a scenario most people recognise immediately when they hear it. A business has been operating for twenty years. It has genuine expertise. It has won industry awards. It has a professionally built website with strong content. It ranks well for its core terms. It has seven Google reviews and four Trustpilot entries.
That dissonance — between the claimed authority and the visible evidence of it — creates a credibility problem that no technical fix addresses. The person evaluating the business asks, consciously or not: if this firm has been trading for two decades and is genuinely excellent, why haven’t more clients said so publicly? The absence of proof becomes a form of proof. Just not the kind you want.
This is not a conversion issue in isolation. It is also a visibility issue. A business with twenty years of claims and seven reviews is not just sending a weak signal to users. It is sending a weak corroboration signal to systems — search engines and AI platforms — trying to determine whether it is a reliable entity worth surfacing. The mechanism differs. The principle is the same.
The same pattern appears elsewhere. A consultancy that claims enterprise experience but names no clients. A software company that talks about scale but publishes no usage data. The claim may be true. The absence of evidence makes it unverifiable.
The psychology is not complicated. People use external validation to reduce the risk of a decision. In the absence of it, they assume risk is higher. Seven reviews after twenty years does not read as humility. It reads as either not asking, or not receiving answers worth publishing.
What to do: Build review acquisition into every client off-boarding process. Not a bulk email to the database — a specific, personal request at the moment a client has just had a positive outcome. That is when the emotional willingness to reciprocate is highest. Make it frictionless: a direct link, a single sentence. Volume matters for algorithms. Recency matters for humans. Both require a systematic process rather than a one-off campaign.
The trust context audit: how many Google reviews do you have, and how recent are the most recent twenty? How does that compare to the competitors a potential client is most likely to evaluate alongside you? Are there sector-specific platforms — legal directories, Clutch for B2B technology, Trustpilot for consumer services — where the same gap exists? Does the volume of case studies and named client references match the claimed scale and longevity of the business?
The goal is not to manufacture authority. It is to ensure that what is genuinely true about the business is visible at every point where a potential client is forming a judgement.
The moment of need looks nothing like daily behaviour
Audience research maps average behaviour. Demographics, platform preferences, typical search queries. All useful. All incomplete — because the version of your audience that actually converts is often behaving nothing like their average.
A 24-year-old who uses TikTok daily, has ChatGPT on their phone, and generally avoids Google gets arrested for drink driving at 1am. In that moment they call a parent. The parent — 55, slightly panicked — opens Google and types “drink driving solicitor Southampton.” They want a phone number. They want to speak to a human being immediately. They are not asking Perplexity to synthesise a comparison of local criminal defence firms.
The sophisticated platform preference of the person who actually needs the solicitor is irrelevant. The behaviour that matters is the parent’s, in a high-stakes, time-pressured, emotionally loaded moment. And that behaviour is exactly the traditional local search behaviour that some agencies are declaring obsolete in the AI transition.
Platform preference collapses under pressure. Behaviour reverts to what feels fastest, safest, and most certain.
The emotional context principle: map not just your audience’s platform preferences but the emotional state they are likely to be in when they genuinely need you. A criminal defence solicitor’s clients are often in crisis. A financial adviser’s clients are often anxious. A plumber’s clients have water coming through the ceiling. Those emotional states determine where they go and what they need to find when they get there — and it is often simpler, more direct, and more traditional than a strategy built on average behaviour would suggest.
The external context checklist: what the data cannot see
Analytics tells you what happened. It cannot tell you why. That silence produces one of the most common errors in SEO consultancy: attributing to a technical problem something that was caused by the world outside the browser. Most SEO reporting ignores this layer entirely.
Here are the market context factors that regularly affect enquiry volume — none of which appear in a crawl report.
Weather. Consistently underestimated for any business with a relationship to the outdoors, construction, home improvement, or leisure. An outdoor living company will not convert garden room enquiries at the same rate in February as in April — not because their SEO is worse, but because the person who buys garden rooms in April is not thinking about their garden in February. Conversely, a roofing or drainage company spikes during and after prolonged wet weather. Because the problem just became visible.
Budget announcements and fiscal events. The Autumn Budget and Spring Statement reliably pause business spending. Capital gains tax speculation, employer NIC changes, business rate uncertainty — each creates a window where decision-makers hold position until the implications are clear. That pause shows in enquiry data within days and typically lasts two to four weeks. This is rational behaviour, not a visibility problem.
Year-end and January. B2B companies see pipeline stall in the last two weeks of December without exception. The people with authority to sign off are on leave or unwilling to start something they cannot complete before year-end. January recovers, but slowly — the first three weeks are typically suppressed before pent-up demand releases. The pattern is consistent enough to plan for.
Elections and political uncertainty. General elections create spending hesitation at both consumer and business level. For clients in policy-sensitive sectors — legal services, financial advice, property, healthcare — the effect is pronounced. A family law firm sees enquiry patterns shift around elections because the people considering divorce are also running financial calculations that depend on the tax environment. This is not a search problem.
Interest rates and economic mood. Bank of England rate announcements directly affect mortgage-dependent services and businesses whose clients are borrowing to fund purchases. When three clients in different sectors all see reduced enquiries in the same two-week window without any site changes, look at what was happening in the world first.
Wars, major events, and news saturation. Sustained news-cycle anxiety contracts discretionary attention. Emergency services — legal, medical, urgent trades — are largely unaffected or see increased demand. Discretionary services — premium consultancy, luxury products, non-urgent home improvements — typically see suppressed enquiries during heavy news cycles.
School holidays. Predictable, consistent, and reliably forgotten. Map your client’s enquiry data against the school holiday calendar for the previous two years. The pattern will almost always be there.
The diagnostic habit: before opening a crawl report when enquiries drop, ask — was there a Budget or fiscal event in the past three weeks? What is the weather doing relative to this sector’s seasonal norm? Are we in a school holiday, year-end window, or early January? Was there an election or heavy news cycle? Did multiple clients in different sectors see the same pattern simultaneously — if yes, external cause is almost certain.
Half the time, the answer to “why did enquiries drop” is in the news, not in the site. The other half, you rule it out in two minutes and move to the technical investigation faster.
The midnight panic problem
When Google rolled out AI Overviews, several clients contacted me at genuinely unreasonable hours convinced something catastrophic had happened. In most cases, their visibility was materially fine. The underlying signals were intact. What had changed was the SERP presentation — and with it, the perception of loss.
The dangerous moment is not the algorithm change. It is the panic response: the impulse to restructure, overhaul, or abandon a strategy that was working because something visible has shifted and the instinct is that action is required. This is where poor SEO advice does the most damage — not through inaction, but through unnecessary change. A restructured site that was ranking well can take six months to recover. A content strategy abandoned mid-cycle loses the compound value of everything built to that point.
The most important skill in consultancy is sometimes not technical. It is knowing when to say: your visibility is fine. The market shifted. The world paused. The SERP changed shape. Give it time. Do not break what is working.
What to tell clients: set expectations early in every engagement that enquiry volume will vary with context — seasonally, economically, and in response to external events — and that variation is normal rather than evidence of failure. Build reporting that contextualises the data: what was happening in the world during this period, what would we expect given the season, what does the trend look like over twelve months rather than four weeks? This reframes performance conversations from panic to pattern recognition.
The platform context: fragmentation is not new
British television had four channels. Then Channel 5. Then Freeview, Sky, Netflix, YouTube, TikTok. The advertiser who bought a prime-time ITV slot in 1995 was buying a captive audience. The same slot today reaches a fraction of that audience because attention has distributed across platforms that did not exist then.
Search has followed the same trajectory. People who once had a single behaviour — go to Google — now have a stack of behaviours that activates by context and intent. I use Claude for complex analytical tasks, ChatGPT for mid-level research, Gemini for quick factual lookups. Not brand loyalty — situational. Three platforms, based on what the task requires and what the habit is for that type of task.
“Where is your audience?” has become the wrong question. “Where do they go when they actually need you?” is the one that matters. The answer is often more fragmented than a single-channel strategy accommodates, and often more traditional than an AI-first strategy predicts.
The practical position: do not migrate your visibility strategy from traditional search to AI platforms. Extend it. Maintain Google rankings. Build Bing infrastructure — it feeds ChatGPT Search and Copilot. Optimise for AI citation through structured content and entity signals. Keep the Google Business Profile current and the review volume growing. Keep the phone number prominent. That is not hedging. It is an accurate reading of how fragmented, context-dependent human attention actually works.
Why this matters more in the AI era
This is the shift most SEO thinking has not caught up with.
AI systems do not reward the best-written content. They retrieve the most defensible, attributable, and contextually grounded source. When ten sites say the same thing, they are interchangeable — to AI, and unconvincing to humans. When one site contains the data, the real-world experience, the specific judgement, or the outcome that no one else can replicate — that is the one that gets named.
The businesses that will be cited, recommended, and surfaced are not necessarily the ones with the most content. They are the ones whose content cannot be replicated from somewhere else: specific outcomes with named clients, original observations from years of pattern recognition, frameworks that carry provenance, proof that exists nowhere else.
That is the same reason a business with twenty years of genuine track record and seven reviews loses to a newer competitor with forty-eight recent ones. The track record exists. The visible evidence of it does not. The systems — human and machine — looking for reasons to trust you cannot find what they need.
The practical summary
Good SEO is technical infrastructure, entity signals, structured content, and AI citation readiness. All of it matters and none of it should be skipped.
But it sits inside a context that no crawler can measure: what is happening in the world this week, what emotional state your potential client is in at the moment they search, what the aggregate signal of your online presence communicates before anyone reads a word, and which platform they are actually on when they need you specifically.
Check the four contexts before diagnosing a problem. Trust — does the visible evidence of your authority match the claimed scale of your business? Emotional — what state is your audience in at the moment they genuinely need you? Market — is there an external reason for the pattern you are seeing in the data? Platform — are you present where your audience actually goes under the specific conditions that precede the decision to enquire?
The job is not just to make a business readable by search engines. It is to ensure that what is genuinely true about it — the expertise, the track record, the proof — is visible, accessible, and present at every point where a real person is making a real decision under real conditions.
That requires judgement, not just tools. And judgement — unlike schema markup — cannot be automated. That is where the advantage sits.
