Marketing Signals

Marketing Signals We Make Sure Your Brand Gets Found - Through Search & AI.

The discipline that makes a GEO scorecard useful vs. decorative: every metric maps to a specific decision.Here's the map...
24/06/2026

The discipline that makes a GEO scorecard useful vs. decorative: every metric maps to a specific decision.
Here's the mapping:

SoAIV overall - Are we winning or losing the channel? Total GEO budget.
SoAIV by funnel stage - Where to invest content - top, middle, or bottom?
Citation rate - Your own content vs. external editorial: where to weight spend?
QoQ change - Continue strategy or pivot?
Editorial volume - Scale Digital PR up or down?
Competitive SoAIV - How fast to move to hold or gain relative position?
Sentiment frame - PR offence (more mentions) or PR defence (fix narrative)?
Engine variance - Which engine to prioritise next quarter?

If your scorecard has 20 metrics, nobody will use it.

Eight is enough to surface every important pattern without drowning the reader. Resist adding more.

The common scorecard mistakes we see - including the one that's most expensive (changing prompts every quarter, which destroys comparability):

The framework to install, matched to what you actually do:𝗜𝗻𝘃𝗲𝘀𝘁𝗶𝗻𝗴 / 𝗰𝗮𝗽𝗶𝘁𝗮𝗹 𝗮𝗹𝗹𝗼𝗰𝗮𝘁𝗶𝗼𝗻: Warren Buffett first. Charlie ...
23/06/2026

The framework to install, matched to what you actually do:

𝗜𝗻𝘃𝗲𝘀𝘁𝗶𝗻𝗴 / 𝗰𝗮𝗽𝗶𝘁𝗮𝗹 𝗮𝗹𝗹𝗼𝗰𝗮𝘁𝗶𝗼𝗻: Warren Buffett first. Charlie Munger second. Then Peter Thiel if you're growth-oriented, Ray Dalio if macro.

𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗼𝗿 𝘀𝗵𝗶𝗽𝗽𝗶𝗻𝗴 𝗮 𝗽𝗿𝗼𝗱𝘂𝗰𝘁: Steve Jobs for product taste. Clayton Christensen for understanding how you'll eventually be disrupted (worth knowing early). Paul Graham if pre-PMF.

𝗟𝗲𝗮𝗱𝗶𝗻𝗴 𝗮 𝘁𝗲𝗮𝗺 𝗼𝗿 𝗰𝗼𝗺𝗽𝗮𝗻𝘆: Jim Collins if you're stable and scaling. Ben Horowitz if you're in the hard part. Andy Grove if your team has slipped into OKR theatre without noticing.

𝗪𝗿𝗶𝘁𝗶𝗻𝗴 𝗮𝗻𝘆𝘁𝗵𝗶𝗻𝗴: Paul Graham for form. George Orwell for cutting what doesn't belong. David Ogilvy if it needs to persuade.

𝗚𝗲𝗻𝗲𝗿𝗮𝗹 𝘀𝗵𝗮𝗿𝗽𝗲𝗿 𝘁𝗵𝗶𝗻𝗸𝗶𝗻𝗴: Munger + Feynman + Kahneman. Three frameworks, no contradictions, covers most ways smart people reason badly.

𝗡𝗲𝗴𝗼𝘁𝗶𝗮𝘁𝗶𝗻𝗴 𝗼𝗿 𝘀𝗲𝗹𝗹𝗶𝗻𝗴: Chris Voss first, full stop. Never Split the Difference as a Claude Skill is the best productivity upgrade available for anyone who sells.

These are recommendations based on what the frameworks actually cover, not the celebrity ranking of the person.

The full decision guide is at the link.

500 people for an hour, or 15 senior leaders for half a day. The leverage is different.Keynotes are visibility. Workshop...
23/06/2026

500 people for an hour, or 15 senior leaders for half a day. The leverage is different.
Keynotes are visibility. Workshops are impact. I'm doing more of the second now.

The economics are different. The preparation is different. The follow-up is different. And the actual change you can drive inside a business is in a different category entirely.

I wrote up why.

Read it → https://garethhoyle.com/writing/corporate-workshops-over-conference-keynotes

Of the five layers in the new search measurement stack, the one most teams skip over is the one that causes everything e...
22/06/2026

Of the five layers in the new search measurement stack, the one most teams skip over is the one that causes everything else to work or fail.

Layer 3: retrieval performance.

Retrieval performance is whether AI engines can actually access and extract your content when they retrieve it. robots.txt allows their crawlers. Key content renders in initial HTML, not only after a JavaScript load. Factual claims appear at the start of paragraphs, not buried after setup. Structured data marks up the highest-value content.

When retrieval performance is weak, the work done in every other layer pays off less. Strong editorial coverage? The AI tries to retrieve your site as a source, fails to extract cleanly, falls back to competitors. Great comparison content? Blocked by a misconfigured robots.txt. Strong category presence? Being cited for framing from third-party sources because your own pages aren't extractable.

Retrieval performance is the layer that brands think about least because it doesn't show up in headline metrics - until you start correlating citation rate against retrieval health. Then the pattern is obvious: the brands with clean retrieval get cited; the brands without it get mentioned but not credited.

What to measure: AI crawler accessibility, schema validity rate on priority page types, server-side rendering coverage, content extractability on a sample of key pages.
Full five-layer framework:

Fintech sits at the most aggressive end of YMYL.Google's quality systems and the major LLMs are extremely cautious about...
20/06/2026

Fintech sits at the most aggressive end of YMYL.

Google's quality systems and the major LLMs are extremely cautious about citing financial brands directly. Author credentials, regulatory standing, content accuracy and source transparency now carry weight they never did before.

Generic blog content under "editorial team" bylines is no longer neutral — it's actively damaging.

Part 7 of our 12-part Visibility Playbook Series.

This week: fintech brands.

Inside the playbook:

→ Why FCA registration, FSCS protection and named credentialed authors are now SEO signals — not just compliance overhead
→ The product page anatomy that gets cited by Perplexity in financial decision queries
→ How to compete with MoneySuperMarket, Compare the Market and NerdWallet in 2026
→ A measurement approach that handles regulatory constraint AND visibility ambition

3,500 words for fintech marketing leaders who need both.

→ Link to the full playbook in the first comment ↓

The thing that makes Claude Skills different from every previous method of customising AI: You don't have to remember to...
19/06/2026

The thing that makes Claude Skills different from every previous method of customising AI: You don't have to remember to use them.

Load a Warren Buffett thinking framework as a skill. Two weeks later, ask Claude a completely different question - "should I take on this new client at reduced margin?" Claude reads your question, recognises it's a capital-allocation decision, auto-invokes the Buffett framework, and answers through the lens of moats, margin of safety, and circle of competence.

You didn't type a command. You didn't paste anything. Claude just knew.

That's a qualitatively different thing from prompting. Prompting requires remembering what to say. A skill is installed expertise - it fires when relevant, stays quiet when not.

Anthropic launched Claude Skills in October 2025. Most people still think of them as "custom instructions with a new name." They're closer to installing a specific kind of thinker in your AI and letting it invoke that thinker when the situation calls for it.

authority(.)md sells 400+ thinking frameworks - Buffett, Munger, Jobs, Feynman, Voss, Collins and more - packaged as native Claude Skills.

The plain-English explainer of what they are, how they work, and how to install one in 60 seconds:

The most common pattern in failed GEO programmes:Month 1: Excitement. Strategy decks. Initial content published. Month 2...
18/06/2026

The most common pattern in failed GEO programmes:

Month 1: Excitement. Strategy decks. Initial content published.
Month 2: First measurement. Some live-retrieval wins, but headline SoAIV is flat.
Month 3: Doubt. "Is this working?"
Month 4: Budget reallocated. Team blames the discipline.
Month 7: Major model update ships. Numbers would have moved. Nobody's measuring anymore.
Month 12: Competitor who didn't pull back is now the AI's default recommendation in the category.

Month 4 is the worst possible time to quit. The live-retrieval work has been building. The editorial pipeline has momentum. The training-data shift is imminent. Pulling back at exactly the moment the first model update was about to reflect the work is the single most expensive mistake in GEO.

The fix isn't working harder in months 2–4. It's accurate expectation-setting in month 0.

What you should see at month 3: technical fixes deployed, citations appearing in Perplexity, editorial pipeline running, Wikipedia work underway. What you should not expect: dramatic ChatGPT SoAIV growth. That's gated by model updates on a different clock.

Full timeline, with leading indicators to track quarterly:

Three years of four-day-week data. Like-for-like. Revenue per head, gross margin, attrition, hiring cost, cost per billa...
18/06/2026

Three years of four-day-week data. Like-for-like. Revenue per head, gross margin, attrition, hiring cost, cost per billable hour.
Most four-day-week content is opinion. This is mine with numbers attached.

Year one: bumpy. Year two: net positive. Year three: clearly the right model.

Full breakdown on the writing page.

Read it → https://garethhoyle.com/writing/economics-four-day-week-real-numbers

The four objections that come up most often when presenting a Search Visibility Audit to a CFO, and how to handle each."...
17/06/2026

The four objections that come up most often when presenting a Search Visibility Audit to a CFO, and how to handle each.

"𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗸𝗻𝗼𝘄 𝗫% 𝗼𝗳 𝗯𝘂𝘆𝗲𝗿𝘀 𝘂𝘀𝗲 𝗔𝗜?"
You don't know - yet. But the estimate is backed by buyer surveys across multiple categories, and we can validate by adding one question to our post-purchase survey. More importantly: we can run 5 spot-check queries right now and see whether we appear or not.

"𝗖𝗼𝘂𝗹𝗱𝗻'𝘁 𝘄𝗲 𝗷𝘂𝘀𝘁 𝗿𝘂𝗻 𝘁𝗵𝗲 𝘀𝗽𝗼𝘁 𝗰𝗵𝗲𝗰𝗸𝘀 𝗼𝘂𝗿𝘀𝗲𝗹𝘃𝗲𝘀 𝗮𝗻𝗱 𝘀𝗸𝗶𝗽 𝘁𝗵𝗲 𝗮𝘂𝗱𝗶𝘁?"
Spot checks give you a sample of 5. An audit gives you 50–150 commercial queries, run 3× each, across 3–5 engines, cross-referenced against SEO authority. The difference is the difference between a blood pressure reading and a health check.

"𝗪𝗵𝘆 𝗻𝗼𝘄? 𝗪𝗵𝘆 𝗻𝗼𝘁 𝘄𝗮𝗶𝘁?"
Every quarter that passes, competitors who are paying attention compound their AI visibility. The cost of catching up scales. Six months from now, the audit reveals a bigger gap at higher cost to close.

"𝗪𝗵𝗮𝘁 𝗶𝗳 𝘄𝗲 𝗷𝘂𝘀𝘁 𝘀𝗽𝗲𝗻𝗱 𝘁𝗵𝗲 𝗺𝗼𝗻𝗲𝘆 𝗼𝗻 𝗺𝗼𝗿𝗲 𝗦𝗘𝗢?"
GEO and SEO share 60% of their work. The audit tells you which of your current SEO investment carries over to GEO and which doesn't. Without that, you're allocating blind. The audit makes the existing SEO spend more efficient, not just the new GEO investment.

The full financial framework - including the worked ROI model and the 60-second version that gets sign-off:

Three stacks I'd actually recommend, for three types of work. Any four frameworks from authority(.)md is $14.99.𝗧𝗵𝗲 𝗚𝗲𝗻𝗲...
16/06/2026

Three stacks I'd actually recommend, for three types of work. Any four frameworks from authority(.)md is $14.99.

𝗧𝗵𝗲 𝗚𝗲𝗻𝗲𝗿𝗮𝗹𝗶𝘀𝘁: Charlie Munger + Richard Feynman + Daniel Kahneman + one specialist for your field. What it does: Mental models and inversion + first principles + cognitive bias awareness. Catches most ways smart people reason badly.
𝗧𝗵𝗲 𝗯𝗲𝘀𝘁 𝗮𝗹𝗹-𝗽𝘂𝗿𝗽𝗼𝘀𝗲 𝘀𝘁𝗮𝗰𝗸 𝗶𝗳 𝘆𝗼𝘂 𝗼𝗻𝗹𝘆 𝗯𝘂𝗶𝗹𝗱 𝗼𝗻𝗲.

𝗧𝗵𝗲 𝗢𝗽𝗲𝗿𝗮𝘁𝗼𝗿: Jim Collins + Andy Grove + Patrick Lencioni + Kim Scott. What it does: Four management frameworks that don't contradict each other. Collins for long-term direction, Grove for high-output ex*****on, Lencioni for team dysfunction, Scott for direct feedback.
𝗕𝗲𝘀𝘁 𝗹𝗲𝗮𝗱𝗲𝗿𝘀𝗵𝗶𝗽 𝘀𝘁𝗮𝗰𝗸 𝗶𝗻 𝘁𝗵𝗲 𝗰𝗮𝘁𝗮𝗹𝗼𝗴𝘂𝗲.

𝗧𝗵𝗲 𝗠𝗮𝗸𝗲𝗿: Paul Graham + Steve Jobs + Rick Rubin + Richard Feynman. What it does: Graham for essay-clear thinking and form. Jobs for product taste and ruthless editing. Rubin for noticing what's already good. Feynman for intellectual honesty.
𝗙𝗼𝗿 𝗮𝗻𝘆𝗼𝗻𝗲 𝗺𝗮𝗸𝗶𝗻𝗴 𝘁𝗵𝗶𝗻𝗴𝘀 - 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀, 𝗰𝗼𝗻𝘁𝗲𝗻𝘁, 𝗰𝗼𝗺𝗽𝗮𝗻𝗶𝗲𝘀.

The bundle pricing exists specifically because stacks work better than individual frameworks. Four frameworks for $14.99 vs $19.96 individually.

Full stacking guide with 12 specific combinations - what works and what doesn't:

Address

Manchester

Opening Hours

Monday 9am - 5pm
Tuesday 9am - 5pm
Wednesday 9am - 5pm
Thursday 9am - 5pm
Friday 9am - 5pm

Telephone

+441619288799

Website

Alerts

Be the first to know and let us send you an email when Marketing Signals posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to Marketing Signals:

Shortcuts

Share

Category