The Punch List: May 2026
By Tim Dini | May 2026 | Issue #3
Five things happened in April that should change how you think about AI and your business. Not all of them are obvious. None of them came from a tools-roundup newsletter. Most of the AI content circulating right now is still selling you the same productivity hacks you saw in February, while the actual ground is shifting underneath.
This month: a new model release that quietly reset how often you can afford to ignore this stuff, a Google feature that’s about to hit your phone whether you’re ready for it or not, hard data showing AI search is roughly thirty times harder to win than Google local, a productivity stat that doubles as a permission slip, and a position from one AI company that’s worth flagging because every other one is going the opposite way.
No filler. No hype. Let’s get into it.
The Punch List
1. OpenAI shipped GPT-5.5 six weeks after GPT-5.4. That cadence is the story.
On April 23, OpenAI released GPT-5.5, their most capable model yet. It powers Codex (their agentic coding tool, now used by more than 4 million people a week) and is rolling out to ChatGPT Plus, Pro, Business, and Enterprise. Anthropic released Claude Opus 4.7 around the same window. Google has been pushing Gemini 2.5 Pro into AI Mode.
None of that is the Punch List item. The Punch List item is the cadence.
GPT-5.4 shipped March 5. GPT-5.5 shipped April 23. That’s seven weeks. The labs are now releasing major frontier models on a roughly six-to-eight-week cycle, and they’re doing it because they’re racing to lock in enterprise procurement before competitors do. That’s the working theory in the analyst coverage and it’s consistent with what you can see from the outside: faster releases, bigger capability jumps, more aggressive marketing.
Why it matters for your business.
If you’ve been telling yourself “I’ll figure out AI search next year,” the math just got worse. The version of ChatGPT recommending plumbers in Phoenix or dentists in Charlotte today is meaningfully different from the one that was doing it in February. By the time you get around to thinking about AEO, the underlying systems will have iterated four or five major versions past where they are now.
This doesn’t mean you need to chase every release. It means the window where ignoring this was a defensible business decision is closing fast. The advantage isn’t going to people who picked the right tool. It’s going to people who started paying attention while everyone else was still arguing about whether AI was real.
2. Google’s AI is about to start calling your business.
Announced April 22: Google Search now includes an agentic feature that calls local businesses on a user’s behalf to get pricing and availability. Search for “pet groomer near me” and you’ll see a button that says “Have AI check pricing.” Click it, submit your request, and Google’s AI calls businesses, gathers the information, and presents you with options.
The feature started rolling out to Google AI Pro and AI Ultra subscribers in the US through the Labs experiment. It’s going to expand.
What this means for the phone ringing on your front desk.
If you run a service business, you’re going to start getting AI-generated phone calls. Sometimes you’ll know. Sometimes you won’t. The voice will be polite, the questions will be specific (price ranges, availability, service area), and the call will end without booking anything because the AI is gathering information for someone who hasn’t decided yet.
Three things to think about now:
First, your front desk staff needs to know this is happening. Otherwise they’ll spend ten minutes giving detailed pricing to what they think is a customer and what is actually a Google data-gathering call. Train them to ask early: “Are you calling on behalf of someone, or are you the customer?” That’s not rude. It’s how the world works now.
Second, your published prices and service descriptions matter more than they did last month. If Google’s AI can call you and confirm pricing that matches what’s on your site, your business is more likely to be the one Google’s AI recommends. If your website says one thing and your phone answers say another, you’re going to lose to the competitor whose information matches across both.
Third, this is one more reason your Google Business Profile needs to be airtight. The AI is grounding its answers in your GBP data first and your phone call second. If the GBP is wrong, the AI is starting from wrong and you’re playing catch-up on the phone.
3. Verified data: AI search is up to 30x harder than Google local.
SOCi published their 2026 Local Visibility Index this spring, and the numbers are the most useful AEO data anyone has put out so far. They analyzed nearly 350,000 locations across 2,751 multi-location brands and tracked how often those locations were surfaced or recommended by AI assistants.
The headline numbers:
In Google’s local 3-pack, brands appeared 35.9% of the time. That’s the traditional local SEO baseline.
In AI search results, that drops dramatically. Gemini surfaced locations 11% of the time. Perplexity, 7.4%. ChatGPT, 1.2%.
SOCi’s own framing: AI visibility is three to thirty times harder to achieve than ranking well in traditional local search.
Why this is actually useful and not just scary.
If you’re a local business owner reading that data and thinking it sounds bad, here’s the reframe. AI search isn’t ten more spots on a search results page. It’s one or two recommendations, sometimes none. That sounds like worse odds, and statistically it is. But it also means the businesses that do show up in AI search aren’t competing with twenty others. They’re competing with one or two.
The same SOCi data showed AI systems are picky about who they recommend. They favor businesses with above-average review sentiment (locations recommended by ChatGPT averaged 4.3 stars), accurate cross-platform data, and complete profiles. AI doesn’t distribute attention evenly. It picks the sources it trusts most and cites them.
That’s not a bug. That’s an opportunity. The bar to win in AI search is higher than the bar to win in traditional local search. But once you clear it, you’re not sharing the spotlight with as many competitors. The plumbers and dentists who get this right early are going to compound their advantage every month while the ones who wait fall further behind.
4. The productivity stat nobody is framing correctly.
Goldman Sachs research released earlier this year reported that workers using AI are saving roughly an hour a day and seeing productivity gains in the 23% range from academic studies, with company anecdotes pointing to gains as high as 33%. Those numbers got a lot of coverage. The Goldman analysts have been on financial news segments framing it as the productivity story of the decade.
Here’s the part everybody is leaving out. According to the same research, roughly 80% of companies aren’t using AI yet.
What you should do with that information.
Eighty percent. That’s not a number that says you’re behind. That’s a number that says almost nobody is ahead. The narrative of “if you’re not using AI you’re already losing” is being pushed hard by people selling AI tools and AI consulting. It’s not how the actual business landscape looks right now.
The window to be early is still open. Not for another five years, probably. But for another 12 to 24 months, the businesses in your market that put genuine effort into figuring this out are going to have a structural advantage that doesn’t yet exist in most local industries. Most of your competition is still doing what they did in 2023, with maybe a Canva subscription added on.
This isn’t an excuse to keep ignoring it. It’s the opposite. It’s the reason to start now, when the bar is still low and the upside is still uncrowded. By the time the 80% number flips to 50%, the easy wins will be gone.
5. One AI company just promised not to put ads in front of you.
On April 28, Anthropic published a piece announcing that Claude will remain ad-free. Their reasoning was straightforward: advertising incentives are incompatible with a genuinely helpful assistant. If an AI is paid to recommend, its recommendations are no longer trustworthy. They’ve explicitly committed to expanding access without compromising user trust.
This matters less because of what Anthropic said and more because of what it implies about everyone else.
AI search is going to get monetized. Here’s what that looks like.
Google has built an entire empire on advertising. Meta runs on it. OpenAI has explored it and reportedly continues to. Microsoft already serves sponsored content through Copilot. The economic gravity here is enormous. AI assistants are expensive to run, and at some point most of them are going to need to monetize the recommendations they make.
When that happens (and it’s a question of when, not if), AI Overviews and ChatGPT search results are going to start including paid placements. The plumber who paid more is going to get recommended over the plumber who has better reviews. The dentist who bid for the keyword is going to show up in front of the dentist who didn’t. We’ve seen this movie before. We’re going to see it again on a new screen.
None of this is a reason to panic. It is a reason to take AEO seriously now, while AI recommendations are still mostly merit-based. The businesses that build genuine authority in AI search before the systems get monetized are going to have a much easier time staying visible afterward. The businesses that don’t are going to wake up one morning and find out they’ve been quietly priced out of recommendations they used to earn for free.
Anthropic’s position is worth flagging because it’s the rare company committing publicly to keeping the channel honest. Whether they hold that line over the long run is a separate question. But the asymmetry is real: most of the AI you use is on a track that ends in advertising, and the time to build durable visibility is before that happens.
That’s the List
Five items, all verified, all pointed at things that are actually shifting. The model cadence is real. The AI calling feature is real. The SOCi data is published and reviewable. The Goldman numbers are sourced. The Anthropic post is dated and signed.
If you only take one thing from this issue, take this: the businesses that win the next two years aren’t going to be the ones with the best AI tool stack. They’re going to be the ones who started thinking about AI as part of how customers find them, not just as a productivity widget. That’s the entire premise of this site, and it gets more true every month.
See you in June.
Sources for this issue: OpenAI launch announcement (April 23, 2026); Vellum and NVIDIA analyses of GPT-5.5; Google blog post on AI-powered calling (April 22, 2026); SOCi 2026 Local Visibility Index via Search Engine Land; Goldman Sachs AI productivity research; Anthropic news (April 28, 2026).
