{ad_logo_primary}

Hi Marketing Bestie,

You know AI is spreading like wildfire.

When your mother in law says, "YOU ARE WRONG. CHATGPT SAID I SHOULD DO THIS." This happened to me yesterday.

Ok Chat stop being a know it all sometimes.

What do you use AI for in your personal life? Reply to me I want to know. Yes. I read the inbox to this newsletter. And yes you are allowed to reply to a newsletter. I look forward to chatting with you.

Was this email forwarded to you?

Subscribe Here

{ad_content_primary}

SPARKNOTES FROM THE POD

5 WAYS TO WIN AI SEARCH

Mike Roberts founded SpyFu. You know, the tool that lets you type in any website and see every keyword they rank for and every ad they've ever run. He basically invented the "spy on your competitors" category.

Before SpyFu, there was no way to see what your competitors were doing. Mike was bootstrapping a software company with $3,000 in his bank account. If his ads didn't work, he'd have to get a real job again. So he built a tool to see what everyone else was ranking for.

Now he's CEO of Rival Flow, which uses AI to analyze what questions your competitors answer that you don't. Then fills those gaps automatically.

The problem: Google's going zero-click. AI overviews answer questions WITHOUT sending people to your site. Your branded traffic? Being answered by AI instead of sending clicks to you.

Fun times in SEO land.

Listen to the full episode here where Mike breaks down why page 4 rankings suddenly matter, what local businesses are screwing up on every page, and why you need a hypothesis before looking at ANY data.

Or keep reading for the key takeaways.

1️⃣. Focus On What You Can Actually Control

Mike's Take: "I would personally focus on the things that you can truly control. And those things are the things that are on your site. And those things are about answering the questions that are being asked... closing the gaps between anybody that's beating you and making sure that you're answering more questions than they are."

Here's what's happening: Google's been moving toward zero-click for years. Remember when definitions started showing up IN the search results? Or weather showing up in autocomplete before you even search?

That's been happening forever. AI overviews are just the next evolution.
But here's the thing everyone's freaking out about: If someone searches your brand name, AI might answer their question WITHOUT sending them to your site. Your branded traffic gets answered by a robot.

Fantastic. And terrifying.

So what do you do? Stop chasing hacks. Reddit was hot for a minute. By the time you hear about a hack, it's probably too late. That's how the internet works.

Mike's advice: Answer more questions than your competitors. On YOUR site. That's what you control.

RivalFlow does this automatically. It looks at pages outranking you and asks: What questions are they answering that you're not? Then uses AI to fill those gaps in your voice.

Because here's the reality: Google still uses traditional search indexing FIRST (TF-IDF if you're nerdy). They return the top 1,000 results. THEN they re-rank with AI.
So traditional SEO still matters. A lot. You need to show up in that initial pool before AI even touches your content.

Takeaway: Stop chasing the latest hack (Reddit mentions, whatever). Focus on YOUR site. Pull up your top 10 ranking pages. Ask: What questions am I NOT answering that competitors are? Look at their content. Find the gaps. Fill them.

Use AI to help if you want, but make sure you're covering more ground than anyone else. That's what gets you in the initial pool. And that's what gets surfaced in AI overviews.

2️⃣. Page 4 Rankings Actually Matter Now.

Mike's Take: "Showing up on page four is actually more valuable now than it used to be. It used to be that if you're on page four, you're absolutely buried. But the AI will find that... AI will find your page four content and surface it into the AI overview."

Used to be, if you ranked on page 4, you were dead. Nobody's clicking to page 4. You're basically invisible.

Not anymore.

AI models dig DEEPER than page 1. They'll find your page 4 article if it answers the question well. And they'll surface it in the AI overview.

ChatGPT. Perplexity. Google's AI overviews. They all do this.

So that deep-dive article you wrote about Jordan brand marketing? If it genuinely answers a question, it can pop up in AI search even if it's not ranking #1.

This changes the game. You don't need to be #1 for every keyword anymore. You need to answer questions THOROUGHLY.

Deep research models help with this. ChatGPT Pro. Claude. Gemini. They dig through more content to find better answers.

And if YOU write that deep content, you're helping everyone. Google doesn't want to run expensive AI on every search. But if you use YOUR ChatGPT credits to create something deep, then publish it, Google can surface it later without the compute cost.

Win-win.

Takeaway: Don't obsess over ranking #1 for everything. Focus on writing DEEP content that actually answers questions thoroughly. That article ranking on page 4?

It might get more visibility NOW through AI than it did before through traditional search. Use AI tools (ChatGPT, Perplexity, Gemini) to research and reference high-quality sources in your content. Then publish comprehensive pieces. AI search will find them and surface them even if they're not page 1 rankings.

3️⃣. Local Businesses Are Still Screwing Up The Basics.

Mike's Take: "They lack a lot of just typical local SEO CRO... there's not like mentions of the local area or landmarks. There's not mentions of weather patterns... There's also often no CTA above the fold... That's kind of a pretty common, unfortunately, really common thing on the internet."

You'd think big companies have this figured out. They don't.

Mike sees this constantly: Local service businesses write perfectly good content about their service. But they forget the BASICS.

No local mentions. No landmarks. No neighborhood references. Just generic content that could be for any city.

Google needs those local signals to rank you for local searches.

Also: No CTA above the fold. You're driving traffic to your site, then making people SCROLL to find your phone number or contact form.

Why?

These are stupid-simple fixes. But they make a HUGE difference.

Mike built this into Rival Flow because it was so common. The tool now detects missing local terms and adds them automatically. It also checks for CTA placement.

Think about it: If you're getting FEWER clicks (because of AI overviews), the clicks you DO get need to convert BETTER. Improving conversion rate is how you survive.

And most businesses are leaving easy wins on the table.

Takeaway: Open your top local service pages right now.

Check:

(1️⃣) Do you mention your neighborhood, nearby landmarks, or local references? If not, add them. Google uses these signals for local rankings.

(2️⃣) Is your CTA (phone number, contact form, booking button) visible ABOVE the fold? If not, move it up. You're getting fewer clicks now. Make the ones you get COUNT. These aren't sexy tactics. But they work. And most businesses still aren't doing them.

4️⃣. Improve Existing Content Before Creating New Stuff.

Mike's Take: "The best, most effective way to improve your traffic is to improve your existing content... The Rival Flow process only focuses on existing stuff. All it's gonna do is juice your existing stuff, get you two to three x more clicks on your stuff that already ranks."

Everyone wants to create more content. More blog posts. More pages. More more more.

Mike's approach: Fix what you already have FIRST.

Why? Because you're probably already ranking for stuff. Maybe page 2 or page 4. If you improve those pages, you can 2-3X your traffic from content that's ALREADY indexed.

Creating new content is expensive. Time-consuming. And it takes months to rank.

Improving existing content is faster. Cheaper. And you see results immediately.

RivalFlow's whole process: Find pages that rank. See what questions competitors answer that you don't. Fill those gaps. Publish. Track results in Google Search Console.

Mike has data from 20,000 customers. He says nobody drops when they make these improvements. When someone does drop, it's usually because the page went 404 (not something they did).

He literally does live videos where he uses a random number generator to pick customer accounts, then shows their charts. It just works.

The pattern: Answer more questions thoroughly on pages that already rank. Traffic goes up.

Simple.

Takeaway: Before creating new content, audit your existing pages. Pull up Google Search Console. Find pages ranking in positions 5-20. These are your low-hanging fruit. For each page, ask: What questions do competitors answer that I don't? Use tools (RivalFlow, or manually analyze top-ranking competitors).

Fill those gaps. Add sections that answer those questions. Publish. Track results. Do this for 10 existing pages before creating 10 new ones. You'll see faster results with less effort.

5️⃣. Come Up With Your Hypothesis BEFORE Looking At The Data.

Mike's Take: "You have to come up with a hypothesis before you look at the data... Everybody tends to look at the data after the fact and then come up with a story after the fact. But you have to figure out how to separate the thing where you can come up with a story after the fact."

This is Mike's marketing hill to die on. And it's brilliant.

Here's the problem: Humans are AMAZING at pattern recognition. Too amazing.
At SpyFu, they do this in interviews. They generate a completely random chart with random axes. Then ask candidates to explain it.

Good candidates can tell a compelling story about the chart. Even though it's completely random. No meaning whatsoever.

That's a strength in storytelling. But it's a weakness in analysis.

We can look at ANY data and create a narrative that makes sense. Even when there's no actual signal.

Vanity metrics are the obvious example. But there's deeper stuff. Cause and effect. "This campaign worked because..." when really you're just making up a story to fit the data.

Mike's rule: Write your hypothesis BEFORE you look at the data. What do you think will happen? Why?

THEN look at the results. Did your hypothesis hold? If not, why not?

This is the scientific method. And it's how actual testing works.

Otherwise you're just storytelling with numbers. And marketers are GREAT storytellers. Which means we can convince ourselves (and our bosses) that anything worked.

Also: AI has the same problem. Give ChatGPT data and ask what it means? It'll make up a story. Just like humans.

So if you're using AI to analyze marketing data, you need to be even MORE skeptical.

Takeaway: Before your next campaign, write down: What do I think will happen? What metric will move? By how much? Why? Put it in a doc. Date it.

THEN run the campaign. THEN look at results. Compare to your hypothesis.

This forces you to think clearly about what you're testing and why. It also prevents you from retroactively justifying anything that happens. Do this for every test. Every campaign.

Every major initiative. It's the only way to actually LEARN instead of just telling stories.

{ad_content_secondary}

IN A MEME