Liam

Tweets about SEO, GEO, and growing brands | Proven results across SaaS, DTC, etc | Currently building the agency I always wanted to hire @

[Liam @LiamLytton]

Our client AetherHaus is now #1 in Google's AI Overview for "best sauna in vancouver."

Cited above the Fairmont.

We didn't optimize for AI Overviews to land there.

Here's why 🧵

[Liam @LiamLytton]

as building software becomes increasingly commoditized, i find myself thinking more and more about distribution, and where the actual signal lives.

if everyone can build, taste, trust and distribution become a lot more important.

two areas keep standing out:

1. having a brand with actual taste
2. ranking high on Google and being recommended by AI models

we’re already getting pretty allergic to AI slop.

when everyone has access to the same tools, everything starts looking pretty good... and a lot of it starts feeling the same.

which makes taste more valuable.

> Anthropic opened a Claude cafe in NYC. the merch, illustrations, “thinking” caps, the whole thing felt unusually considered.

> Palantir (of all companies lol) has basically built its own apparel brand with limited drops and custom pieces.

software companies are starting to behave more like cultural brands.

i don't think that's random.

then there’s search.

Google is still massive, and AI models are creating another discovery surface on top of it.

people are asking ChatGPT:

> where should i stay?
> what software should i buy?
> what workout should i do?
> which wedding dress should i look at?
> every one of those questions forces the model to choose.

there might be 5,000 companies selling wedding dresses.

ChatGPT recommends a handful.

Google works similarly. thousands of possible businesses, very few that actually get meaningful visibility.

being one of the companies that gets chosen is distribution.

and increasingly, companies need to earn trust in two places at once:

with people

and with the systems people use to decide what to buy.

building is getting easier; getting chosen is becoming the hard part.

[Liam @LiamLytton]

@asaio87 the compounding part is what people miss. once you've built domain authority and links, every new feature or product line you launch gets landing pages that rank way faster than they would cold. rising tide raises all ships.

[Liam @LiamLytton]

@SearchForRyan This a more anomalous case - u have to be super careful with pbns. it can tank a site especially with Googles new spam updates

[Liam @LiamLytton]

@anupamrjp claude 2.1

[Liam @LiamLytton]

Liquid Death is a $1B+ company that sells water.

Water in a tall black can that looks like a cheap beer, with "Murder Your Thirst" on the side.

The guy behind it, Mike Cessario, didn't come from the drinks world at all.

He was an ad guy who happened to love heavy metal.

The idea hit him at a punk festival.

He noticed bands on stage drinking water out of Monster energy cans.

The sponsor made them hold the product, but nobody wanted an energy drink in 98-degree heat, so they'd filled the cans with water.

That stuck with him.

Cool packaging, simple product, and nobody was doing it for real.

People pick a drink partly by what it says about them, and every healthy option on the shelf was clean, soft, and boring.

One bottle of water is the same as the next, so most brands just fight over price.

Cessario didn't bother competing on the water at all.

Instead, he built a brand that's loud and anti-everything, and wrapped it around a product as boring as bottled water.

It worked so well that people pay to wear the logo, and some tattoo it on themselves.

Go back to those bands at the festival.

They had to hold the can. Cessario just figured out how to make people want to.

Make the boring choice the cool one.

Win on identity and you never have to win on price.

[Liam @LiamLytton]

@Berry_cooool auto

[Liam @LiamLytton]

AI search is the new word of mouth marketing.

Someone shopping for project management software asks ChatGPT, "what's the best tool for a small remote team?" and it hands them three names.

I've helped over 40+ companies become one of those names, and it always comes down to the same three things, in order:

→ Build the base.

Your site has to show up on Google, load fast, and be easy to read. It needs other sites linking to it. The same basics that get you found on Google are what get you found by AI.

→ Be genuinely useful.

You need real pages that clearly answer what buyers are asking, specific to what you actually do.

→ Be trusted.

Reviews, what customers say, where your name shows up. Everyone obsesses over this part, but it holds more weight when the first two are in place.

Get the order right and AI starts naming you.

Word of mouth always went to the businesses that earned it. AI just does the recommending now.

[Liam @LiamLytton]

I built a Claude Code prompt that audits any website the way an AI agent reads it.

You drop in your URL, it runs Google's new Lighthouse check, and hands you the results in plain English.

Google just shipped this test to score whether an AI agent can actually use your site.

Not just read it.

Move through your pages and take action, the way ChatGPT or an AI assistant would when someone points it there.

Most sites are built for people, so they fail.

It looks at four things.

For every one you fail, it shows you the exact element and the one-line fix.

Comment "agent" and I'll send you the prompt.

[Liam @LiamLytton]

@asaio87 I see it all over my feed that "google is dead" and "GEO is the new SEO" and it's just people with very limited understanding trying to spread FUD so you buy their vibe coded GEO application 😂

[Liam @LiamLytton]

@harshithOG @pcshipp are you adding internal links from posts? And are you submitting to GSC after you publish it?

[Liam @LiamLytton]

Investing in SEO in 2026 is like investing in the S&P 500.

Here's what I mean.

You don't buy the index and check it the next morning.

You put money in, leave it alone, and let it compound.

Slow at the start.

Then month one builds month two, and month two builds month three.

SEO works the same way. A page that earns its place on Google keeps bringing in customers long after you built it. The meter doesn't reset.

Most people give up on SEO for the same reason most people struggle to invest:

→ See a slow start
→ Get impatient
→ Pull out right before it compounds

Here's what makes 2026 different.

The same work that gets you found on Google now gets you recommended by ChatGPT, Claude, and Gemini too.

One investment, paying out in more places than it used to.

The ones who stuck with it stopped wondering where their next customer comes from. The inbound just keeps coming.

Treat it like a campaign and you'll quit early.

Treat it like an index fund and you'll be glad you stuck with it.

[Liam @LiamLytton]

Top 20 places to get free backlinks:

1. Source of Sources
2. Featured. com
3. Help a B2B Writer
4. Qwoted
5. Reddit
6. Quora
7. G2
8. Crunchbase
9. GitHub
10. Medium
11. Substack
12. dev. to
13. Hashnode
14. Slideshare
15. Product Hunt
16. Hacker News
17. Indie Hackers
18. AlternativeTo
19. There's An AI For That
20. Capterra

[Liam @LiamLytton]

All of our clients at The 66th are getting cited and recommended by AI models.

Here are 3 things we focus on:

1. BE FINDABLE (can Google and AI actually read your website?)

→ Make sure Google can see your pages

AI tools learn from Google. If Google can't load your page, AI doesn't know you exist.

Use Google's free native tool "Google Search Console," paste a page link into the top bar, and check that the page is indexed.

→ Make your site fast

A slow or broken phone experience gets passed over before anyone reads a word.

Type your site into "Google PageSpeed Insights." If the phone score is red, get a developer to speed it up.

→ Link your own pages together
A page nothing links to looks unimportant. Treat your site like a map, every stop should connect.

Make sure every important page can be reached from your menu and one other page.

───────────────

2. CREATE AND BE SPECIFIC (the page has to exist, and it has to be detailed)

→ Build a page for each thing you do and place you serve

You can't show up for a service or a city you don't even have a page for. The page has to exist first.

Give every service and every location its own dedicated page. Don't lump them all onto one.

→ Name real places on every page

If you serve a city, don't just say "the city." Name the actual neighbourhoods and streets you work in, inside the content of the page.

→ Say things only a local would know

Local rules, weather, busy seasons, permits. The stuff a copy-paste page could never fake.

Ask yourself: could a business in another city steal this paragraph word for word? If yes, rewrite it.

→ Stop writing "thin" pages

"Thin" means a page that could describe any business anywhere. It says nothing real, so AI learns nothing about you.

Search what you do. Read the top 3 results. If your page says less, add the missing details.

───────────────

3. BE TRUSTED (what other people say about you matters more than what you say)

→ Fill out your Google Business Profile completely

That free business listing on Google Maps is one of the things AI grabs about a local company.

Open it. Fill in every box. Photos, hours, services, description. Leave nothing blank.

→ Get mentioned on sites people already trust

Being listed in "best [your service] in [your city]" articles puts you where AI is looking.

Search that exact phrase, find the articles that rank, and ask to be added.

→ Pay attention to your reviews and reputation
What customers say about you shapes how AI describes you to the next customer.

Ask ChatGPT: "Is [your business name] any good?" That answer is the impression you're giving off.

[Liam @LiamLytton]

That's it. No guesswork, no expensive tools.

Two searches, one look, and Google tells you whether you're building one page or two.

Follow for more SEO breakdowns like this.

[Liam @LiamLytton]

Two of your own landing pages can fight for the same Google search.

Instead of one ranking high, both get stuck.

You don't have to guess when it's happening. Google already tells you, you just have to read the results.

Here's the 2-minute check 👇

[Liam @LiamLytton]

This is the whole call: are the same sites showing up for both?

Mostly the same sites → Google is reading one intent. Build ONE page. Two pages here just split your own strength in half.

Mostly different sites → two intents. Each topic has earned its own page.

[Liam @LiamLytton]

I took a look at the your backend.

you're basically starting from zero. I would recommend creating blogs posts from your own experience (dont just have ai generate random slop), and bottom of funnel landing pages.

The mental model is to build useful content. Quality over quantity.

[Liam @LiamLytton]

A Fraser Valley contractor came to us with a 5-page Squarespace site and a pipeline that relied on ad spend.

10 weeks later he's pulling 4x the leads, and most of them are coming from Google, ChatGPT, Claude, and Gemini.

We built him a new client acquisition channel from scratch.

Here's what we did:

→ Rebuilt the site for conversion and SEO
→ Built 100+ pages targeting keywords across the full funnel
→ Earned niche-specific backlinks

What we learned along the way:

→ Buyers burned by competitors were searching "[competitor] alternative" in ChatGPT and Google
→ People were searching at the neighborhood level, not just city-wide

What we are doing next:

→ Comparison pages for buyers typing "X vs Y" or "[competitor] alternative"
→ More neighborhood-level pages
→ Expanding geographical reach into Vancouver Island

What this means for you:

→ Different buyers live in different tools: Google, ChatGPT, Claude, Gemini. Optimize for search and you open all four. AI tools cite the same pages that rank on Google.
→ The highest-intent buyers are typing queries your competitors ignore. "[Competitor] alternative." "[Neighborhood] [service]." Easy wins if you show up.
→ Search compounds. Pages you build today still produce leads next year.

Landing pages are salespeople that work for you 24/7.

He went from a team of 5 to a team of 100+.