SEO On-Air

What AI SEO Still Cannot Do For You

Can AI SEO replace an agency, or does it just do whatever you ask it faster, right or wrong?

In this episode of SEO On-Air, host Koa Kauwe sits down with Kyle Roof (PageOptimizer Pro) and Jake Soffer (Firm Pilot) to be honest about where doing it yourself genuinely works and where it falls apart.

The conversation examines data rather than opinion. Kyle ran a study across first time users of his on-page software and found every piece of content missed the same seven things, starting with the target keyword in the title tag and the H1, which are the two most basic placements in SEO. Worse, 99 percent of pieces came in light on contextual terms, often two standard deviations below the range Google needs to understand a page at all. His sharpest finding came from eighteen months of testing the same prompt: asking an LLM to make content SEO optimized measurably makes it worse, lowering term counts, word counts and optimization scores across the board.

They also get into Jake's moonshine analogy for what raw LLM output actually is, why AI is a force multiplier in both directions, why a model's goal is the next probable token rather than anything SEO wants, the myth that the chat sidebar is memory, why shoveling content over the fence gets expensive later, and why an agency should be judged on being more right than your competitors rather than right every time.

🎯 What You'll Learn:

• Why asking an LLM to optimize for SEO makes the output worse

• The seven things nearly every AI written page misses

• Why 99 percent of pages come in light on contextual terms

• Why execution, not strategy, is the real DIY bottleneck

• Why an LLM's goal is fundamentally different from yours

• Why AI multiplies bad SEO just as fast as good SEO

• Why the chat sidebar is not the memory people think it is

• Why scaled content becomes a cleanup bill later

• Why nobody is right 100 percent of the time in search

• Why agencies are moving further down the client journey

Whether you run an agency, brought marketing in-house, are weighing whether to keep paying one, or just want to know what AI genuinely cannot do here, this episode gives you tested numbers instead of hot takes.

🎧 Tune in to hear why prompting for better SEO is one of the fastest ways to get worse SEO.

What Google AI Mode Is Doing To Local Search

Is Google AI Mode genuinely reshaping local search, or just adding another surface to optimize for?

In this episode of SEO On-Air, host Koa Kauwe sits down with Andrew Shotland, CEO of Local SEO Guide, fresh off a conference talk built on an analysis of millions of local search results.

The conversation examines data most agencies have not seen. Across roughly 15 million local queries with AI Overviews, Andrew found that being cited put you visibly on screen only 23 percent of the time on desktop and just 8 percent on mobile, which makes citation a far weaker signal than it sounds. The three Google AI surfaces also disagree with each other on the top local recommendation around 91 percent of the time, and in four out of ten US local markets, ChatGPT's top pick never cracks the Google Maps top three.

They also get into competitor conquesting inside AI, pulling AI queries out of Search Console to shape content strategy, piping sales call transcripts into Slack as a daily FAQ feed, why ChatGPT leans on Bing, Foursquare, Yelp and TripAdvisor rather than Google reviews, why the first 200 reviews behave differently from everything after, the engagement experiments that moved rankings in a week, and why no agency he has consulted for has a testing regimen.

🎯 What You'll Learn:

- Why being cited in AI rarely means being seen

- Why only 8 percent of mobile citations are actually visible

- Why the three Google AI surfaces disagree on the top result

- Why ChatGPT's top pick often misses the Maps top three

- Where each AI system pulls its local data from

- How competitor conquesting works inside AI

- How to pull AI queries out of Search Console

- Why your first 200 reviews matter differently

- Why proximity is still the number one local factor

- Why the page your Google profile links to matters

Whether you run local SEO, own a multi location brand, work agency side, or are trying to work out whether AI has actually touched your map pack, this episode gives you data instead of opinion.

🎧 Tune in to hear why 80 percent of local situations are unique, and what that means for anyone selling a fixed package.

How AI Automation Works In Healthcare And Legal Marketing

Can AI automation work in regulated verticals like healthcare and legal, or does compliance rule it out?

In this episode of SEO On-Air, host Koa Kauwe sits down with Brian Hong, who runs InfintechDesigns and a stack of companies across SEO, AI automation, and data, serving home services, medical and legal clients from three thousand a month to two million.

The conversation examines how far the automation actually goes. Brian's answer on regulated industries is closed loop systems, HIPAA compliant messaging built for med spas, plastic surgeons and law firms, including hardware that sits on site so patient information never touches a public model. His rule on the automation itself is a human in the loop at every high level decision, a live transfer option on every call, and always disclosing that it is AI. His framing is blunter than most: the goal is not to eliminate humans, it is to eliminate the ones you did not want working for you anyway.

They also get into why he scrapped fixed SEO packages entirely, the five buckets most clients think are one, why SEO is a salesman's game more than a strategist's, building the hurricane pack before hurricane season, how he got his team past the terminal, the team member with no web design experience whose Claude built sites keep beating his actual design department, and what to do in your first 90 days if you have built none of this.

🎯 What You'll Learn:

- How AI automation stays HIPAA compliant in regulated verticals

- Why every automated call needs a live transfer option

- Why fixed SEO packages stopped working entirely

- The five buckets of SEO most clients think are one

- Why clients pay for revenue, not links or rankings

- Why the SEO was working and the follow up was not

- How to build seasonal campaigns before the season hits

- Why high ability beats high skill when hiring

- What to build in your first 90 days with AI

- Why he would not start a web design agency today

Whether you run an agency, market in a regulated industry, are weighing AI automation, or are stuck selling fulfillment, this episode gives you an operator who rebuilt everything around the assumption that fulfillment is now commoditized.

🎧 Tune in to hear why a team member with no design training keeps winning pitches against the design department.

How Agencies Are Actually Using AI In Marketing

Is AI replacing agency headcount, or just raising the baseline everyone now has to clear?

In this episode of SEO On-Air, host Koa Kauwe sits down with Jason Ciment (Get Visible), Andrew Melchior (Avalaunch), and Adam Taylor (Splendor) to find out what AI is genuinely doing inside agencies with decades of operating history between them.

The conversation examines three honest answers. Jason says headcount has not moved, they are simply getting more done, and his team has built three platforms since February including an audit that would take fifty people to produce by hand. Andrew's framing is that AI is the first draft, useful for research, reporting narrative, QA passes and data reconciliation, but it never supplied the strategic judgment about what the draft should be about. Adam adds the hiring angle, that A players gravitate toward new tools, pointing to an art director and former typography professor whose assembly work dropped from five days to an hour.

They also get into why brands stopped buying deliverables and started buying accountability, the client who fired their agency for AI tools and came back saying the site was annihilated, why Jason no longer lists detailed deliverables in proposals, why the sales to fulfillment handoff is where quality breaks, the hiring pendulum swinging back from foosball tables, and the conversation that ends with telling a client they are undercapitalised by 80 percent.

🎯 What You'll Learn:

- Why AI has raised output without reducing headcount

- Why AI is the first draft and never the strategic judgment

- The audit platform that replaced fifty people of manual work

- Why entry level roles feel this first

- Why A players adopt new tools fastest

- Why brands stopped buying deliverables and started buying impact

- Why one agency stopped listing deliverables in proposals

- Where quality actually breaks down as an agency scales

- Why the hiring pendulum is swinging back

- How to tell a client they are undercapitalised

Whether you run an agency, work inside one, buy marketing services, or are deciding how far to lean on AI, this episode gives you three operators with very different models landing on the same conclusion.

🎧 Tune in to hear why the client who replaced their agency with AI tools was back within weeks.

Why White Hat SEO Was Always The Right Answer

Was white hat SEO always the right answer, or just the slow answer that happened to outlast everything else?

In this episode of SEO On-Air, host Koa Kauwe sits down with Jack Shepler (Ayokay), Bruce Rowe (Sebo Marketing), and Adrian J. Reynolds (Peak Activity) for a part two about why doing it properly stopped being optional.

The conversation examines a claim Adrian makes plainly: nothing that mattered a decade ago has stopped mattering. Keyword meta tags and invisible text fell away, but those were always gray or black hat, so anyone who stuck to white hat strategies has been doing the right thing the whole time. Jack frames trust as the foundation of brand and says the newest trust gap is agencies built on AI output, which he can now recognize on sight. Bruce brings the horror stories, including the client paying ten thousand a month on a three year contract who had zero of fifty target keywords in the top fifty after eighteen months, and was quoted 180,000 dollars to leave.

They also get into the CEO who demanded every link below DR 30 be disavowed and watched the rankings crash, eight salespeople drop to two, and the company fold, why Google made 450 algorithm changes in 2007 alone, the people deliberately poisoning LLM answers by seeding Reddit threads, why if link building is not hard you are probably doing it wrong, and the three questions a client should ask any agency.

🎯 What You'll Learn:

- Why the white hat strategies were always the durable ones

- Why a natural link profile needs low quality links in it

- What happened when a CEO disavowed every link below DR 30

- Why trust is the foundation of brand, not a soft metric

- How to spot an agency built on AI output

- Why people are deliberately poisoning LLM answers about SEO

- Why doing link building at all is still non negotiable

- The three questions to ask before hiring an agency

- Why you cannot deposit traffic in the bank

- The SEO myths these three wish would finally die

Whether you run an agency, buy SEO, build links, or are weighing a long contract you are not sure about, this episode gives you three practitioners who never met before agreeing on exactly the same answer.

🎧 Tune in to hear how cleaning up a link profile destroyed the company that ordered it.

The Red Flags To Watch For Before You Hire An SEO Agency

Can a client spot a bad SEO agency before signing, or do the red flags only show up months later?

In this episode of SEO On-Air, host Koa Kauwe sits down with Aaron Hockel (AltaVista Strategic Partners), Kevin Clinton (DEPOPAS), and Sandro Taliana (Interwetten Group) to name the warning signs plainly, from three very different corners of the industry.

The conversation examines an uncomfortable pattern. Sandro points out the agency that promises everything is usually the one that wins the contract, even though nobody can guarantee results, and he has watched identical white label brands perform completely differently with the same marketing behind them. Aaron names two specific red flags: any guaranteed ranking position, and anything that goes against a platform's own terms of service, such as spinning up extra Google Business Profiles. Kevin adds blog stuffing, and counters it with a client who published one good post a week and tripled their traffic in six weeks.

They also get into the client who fired an agency after closing one lead out of 65, the commercial painter ranking for bedroom questions he does not answer, why ROI tracking usually breaks on the client's side rather than the agency's, the prospect who insisted Claude had captured their brand voice, and what to do when clients start running your monthly report through an AI to generate hard questions.

🎯 What You'll Learn:

- Why any guarantee of rankings is an immediate red flag

- Why the agency promising everything usually wins the contract

- Why extra Google Business Profiles can get you penalised

- Why seven blog posts a week stopped working entirely

- The one blog a week that tripled a client's traffic

- What happens when an agency does not understand your business

- Why your own systems decide whether ROI can be proved

- Why lifetime value is the number most clients ignore

- Why agencies are going deeper rather than wider

- What to do when a client runs your report through an AI

Whether you're about to hire an agency, unhappy with your current one, running an agency yourself, or handling marketing in-house, this episode gives you three practitioners naming the warning signs they would walk away from.

🎧 Tune in to hear why nobody ever asks their AI what their agency is doing well.

How Marketing Operations Is Changing Agency Work

Is marketing operations now part of an agency's job, or does that line still belong to the client?

In this episode of SEO On-Air, host Koa Kauwe sits down with Kymber Heinze (Growth99) and Brandon Kidd (DeltaV Digital) to work through a shift plenty of agencies are feeling but few are naming.

The conversation examines where the handoff stopped working. Kymber says marketing used to end at delivering traffic and leads, and now her team sits deeper in a client's daily business than ever, because staff hours and appointment availability are what actually run the marketing. Brandon frames it as a value problem, since marketing is usually the first budget cut, so proving value means getting a hand into whether the phone even gets picked up. Both land on the same uncomfortable question: who owns the conversion now.

They also get into why marketers are becoming overseers rather than executors, the AI made marketers sending ChatGPT answers at midnight asking why we are not doing this, the Claude landing page mockup that lands in your inbox on a Monday, why both agencies keep AI tools out of client hands, the two levels of conversion and why only one really counts, and the operations mistake agencies keep repeating as they scale.

🎯 What You'll Learn:

- Why conversion is becoming the agency's responsibility

- Why marketing is usually the first budget to get cut

- Why marketers are becoming overseers instead of executors

- How to handle clients who arrive with ChatGPT answers

- Why both agencies keep AI tools out of client hands

- The risk of clients having access to their own ad accounts

- Why owning the website changes what you can track

- The two levels of conversion and why only one counts

- The operations mistake agencies keep repeating

- Why what started your agency will not keep it going

Whether you run an agency, lead operations, manage client accounts, or are watching your scope creep past the handoff, this episode gives you two operations people explaining what the job actually became.

🎧 Tune in to hear why the people who built your agency are rarely the people who can scale it.

How To Write SEO Content For Humans And Bots

Should you write SEO content for the humans reading it or the bots crawling it, when the bots now read more of it?

In this episode of SEO On-Air, host Koa Kauwe sits down with Kelsey Mast (Third and Grove), Marc Brookland (SEO Locale), and Drew Harden (Blue Compass) to work out who your content is actually for now.

The conversation examines three answers that converge. Kelsey writes for completeness against the query fan out model, since a search for the best moisturizer for dry skin breaks into questions about ingredients, products, and price, and her stance is that content unhelpful to a human will not help an LLM either. Marc pushes back that writing purely for humans is not enough, since a three sentence answer may serve the reader and still never rank. Drew lands in the middle, arguing good structure is simultaneously good for people and good for crawlers.

They also get into why reporting is the thing that actually needs breaking, why clicks falling does not mean the work stopped working, the Perplexity and Reddit fallout that wiped out anyone over invested in one platform, why you can no longer fake authority, author bios and accreditations, and Drew's advice to celebritize your team.

🎯 What You'll Learn:

- Why writing for humans alone is not quite enough

- How the query fan out model changes what completeness means

- Why structure that helps crawlers also helps readers

- Why clicks falling does not mean your SEO stopped working

- Why LLM traffic converts at a higher rate

- What the Perplexity and Reddit fallout taught about single platform bets

- Why you can no longer fake authority on a page

- Why author bios and accreditations now matter

- Why mentions matter even without a link attached

- Why celebritizing your team beats corporate content

Whether you write content, run an agency, own a search program, or are deciding how much to trust AI with your copy, this episode gives you three practitioners arguing out the same question from different sides and landing somewhere useful.

🎧 Tune in to hear why the best content strategy still starts with a human, and why that is not the whole answer.

Why Digital PR Is The New SEO Strategy

Is digital PR genuinely becoming an SEO strategy, or just PR wearing an SEO label?

In this episode of SEO On-Air, host Koa Kauwe sits down with Chris Ferris (Pierpont Communications), Tyler Jacobson (4B Marketing), and Lindsey Carnett (Marketing Maven) to work out why PR and SEO started merging, and what AI had to do with it.

The conversation examines a shift all three are living through from different sides. Lindsey says her PR first agency has taken more referrals from SEO agencies in the last nine months than in the previous seventeen years, because those agencies can no longer do it alone. Chris brings the measurement view, tracking correlation between earned media and branded search, and argues visibility rather than ranking is the right word for AI. Tyler adds that PR became the last grain of sand that moved niche clients into position.

They also get into whether the industry is overcompensating with watered down pay to play placements, why industry publications often outperform a Wall Street Journal hit, why AI companies are signing content deals with publishers, and the five factors that separate a newsworthy pitch from a blacklisted one.
🎯 What You'll Learn:

- Why SEO agencies are now referring clients to PR firms

- Why visibility, not ranking, is the right word for AI

- Why industry publications can beat a tier one placement

- Why AI models favor fact checked journalism over Reddit

- Why AI referred visitors convert at a higher rate

- The attribution gap where Google takes credit for AI

- Why pay to play PR is the new PBN problem

- Why journalists now ask you to confirm a pitch is not AI

- The five factors that make a pitch newsworthy

- Why a press release and a media pitch are different things

Whether you run an agency, own PR or SEO, buy either service, or are trying to understand why the two keep converging, this episode gives you three practitioners who came at the same shift from opposite directions.

🎧 Tune in to hear the host get coached through the pitch he sent that probably got him blacklisted.

AI Search Optimization: How Much Has The SEO Job Changed?

Has AI search optimization genuinely changed the SEO job, or just added another surface to cover?

In this episode of SEO On-Air, host Koa Kauwe sits down with David Reske (Nowspeed) and Lizzie Lewington (Amstrad Digital) to work out how much of the job actually changed once LLMs entered the picture.

The conversation examines a split worth arguing about. David treats GEO and SEO as separate practices, tracking keywords for Google and queries for AI platforms, with link building on one side and entity authority on the other. Lizzie notes the industry is genuinely divided on that, and says either way the measurement layer is what lets you refresh a strategy honestly. Both agree the part AI cannot replace is strategy, and that most agencies never learned what the word means, since hire an SEO agency has historically meant a few blogs, some links, and a meta tag fix.

They also get into why keyword packages stopped working as retainers, which industries AI Overviews hit hardest, why SEO traffic is labor rather than free, David's champagne taste on a beer budget analogy, cannibalization between a Google strategy and an LLM strategy, the questions a client should ask to find out if they are getting value, and why a monthly report is not the same thing as the work.

🎯 What You'll Learn:

- Why GEO and SEO may be two different practices

- Why keyword packages stopped working as retainers

- Why strategy is the part AI cannot automate away

- Why SEO traffic is labor, not free

- Which industries AI Overviews hit hardest

- Why cookie cutter strategy is the clearest red flag

- The questions to ask before hiring an SEO agency

- Why a monthly report is not the same as the work

- When project work still makes sense over a retainer

- How involved a client actually needs to be

Whether you run an agency, buy SEO services, sell retainers, or are trying to judge whether your current provider earns the fee, this episode gives you two practitioners naming what separates strategy from fulfillment.

🎧 Tune in to hear the one question that catches most agencies off guard when a client asks it.