AI news & training for business owners & operators. One email. One clear next step.

THIS WEEK IN AI

This was the week the AI bill came due. Microsoft, Alphabet, and Meta all reported earnings, and the question every analyst kept asking was the same one most business owners are quietly asking themselves: when does any of this actually pay back. The answers were uneven. The pressure to find ROI is now everywhere.

What happened:

On April 29, Alphabet, Microsoft, Meta, and Amazon reported earnings on the same evening. Combined AI-related capital expenditures across hyperscalers will exceed $600 billion in 2026, with Alphabet alone raising its full-year capex guidance to $180–190 billion. Markets rewarded Google, whose Cloud revenue grew 63% year-over-year to $20 billion, and punished the others when they could not show the same return. Microsoft's CFO said the company expects to stay capacity-constrained on AI compute through 2026.

Why it matters to your business:

The companies selling you AI tools are under pressure to show their own boards the spending is paying back. That pressure flows downhill. Expect more aggressive sales motions, more bundled features added to plans you already pay for, and more "limited-time" pricing on enterprise tiers over the next two quarters. None of that changes the question you should be asking. Is this tool fixing work I can name, or am I buying it because someone in a webinar made me feel behind?

Source: Fortune

What happened:

On its Q1 earnings call this week, Meta said its business AI tools are now handling about 10 million conversations per week, up from 1 million at the start of the year. The product is free, available across Facebook, Instagram, and WhatsApp, and aimed at small businesses without a customer service team. Mark Zuckerberg said monetization "may change in the near future."

Why it matters to your business:

If you sell to consumers and you live on Meta's platforms, this is the first AI tool worth a real test. It is free, it lives where your customers already message you, and the bar for "good enough" customer response is lower than most owners think. The catch is the timing. Meta is scaling adoption now and will start charging later. Get value out of the free window. And document what works so you can decide whether to keep paying for it when the price arrives.

Source: TechCrunch

What happened:

At Google Cloud Next this week in Las Vegas, CEO Thomas Kurian rolled out the "Agentic Enterprise" strategy, framing AI as a "system of action" rather than a "system of intelligence." The headline product is the Gemini Enterprise Agent Platform, which lets businesses deploy what Google is calling "fleets" of agents that can plan, reason, and run multi-step workflows for days at a time. Google announced over 260 product updates at the event. Paid monthly users of Gemini Enterprise grew 40% last quarter.

Why it matters to your business:

"Fleets of agents" sounds impressive in a keynote and intimidating in your inbox. The honest read is simpler. Google, Microsoft, OpenAI, and Anthropic are now selling the same thing in slightly different wrappers, and the product is no longer a chatbot. It is a worker. Whether you need one yet depends entirely on whether you have workflows clear enough to hand off. Most 10-to-200 person businesses do not. That is not a failure. It is the actual starting line.

Source: devFlokers

This week, four browser tabs from four different AI companies pitched me the same thing. An agent that runs your business while you sleep. Pull from your CRM. Draft your follow-ups. Update your project tracker. Send the invoice. Approve the expense. All of it, autonomously, for a credit-based fee that nobody can explain in plain English yet.

The demos are good. The product is real. And for most businesses I work with, it is the wrong purchase right now.

Here is the pattern I keep seeing. An owner watches the demo. Gets excited. Buys seats. Picks one workflow to point the agent at. Sits down to write the instructions. And freezes. Because the workflow only lives in one person's head. Janet does the estimating. Mike does the follow-up. Janet and Mike do not write things down. They just know how it works. The agent cannot read minds. Neither can the next person you hire. So the agent purchase quietly exposes a problem that was already there. The work has no owner outside the person doing it.

This is the part the AI sales pitch skips. An agent is a system that runs a workflow. If the workflow does not exist on paper, the agent has nothing to run.

I am not saying agents are useless. I am saying they are the second purchase, not the first. The first purchase is clarity. Sit down with the people doing the work. Watch them do it once. Write down what they did, in order, with the decisions they made and why. Now you have a workflow. Now an AI tool, agent or otherwise, has something to grip.

The companies winning with AI right now are not the ones with the most expensive tools. They are the ones who took a process that lived in three people's heads and put it on paper before they automated anything. A contractor I worked with last year cut estimating from three hours to twenty minutes. We did not start with an agent. We started with a question. What does the estimate actually need to say, and where does the information come from? Two hours of conversation. One template. One prompt. The AI was the last twenty percent of the work. The first eighty was clarity.

And there is one more reason to slow down. Every hyperscaler reported this week that they are spending hundreds of billions on AI infrastructure. That money has to come back somehow. Expect prices on agent platforms to rise, free tiers to shrink, and pricing models to get more confusing before they get clearer. The owners who built clarity into their operations this year will be in a position to choose. The ones who bought the demo will be locked into whatever pricing arrives next.

If you are an owner staring at the agent announcements wondering whether you are falling behind, here is a fairer question to ask. If the one person who knows how it works stepped away tomorrow, what breaks first? Fix that. Then add AI.

Speed is not your advantage. Clarity is.

Before you upgrade any subscription, hire a consultant, or sign up for a new agent platform this month, run the tool past these four questions. If you cannot answer all four, the tool is not the problem. The work underneath it is.

Forward this to the next person on your team who tells you they want to "try AI for something." Have them answer the four questions before they spend a dollar.

If one of those four questions stopped you cold, that is the signal. Book a free 20-minute fit call and we will talk through it. No pitch, no slides, just an honest look at what is actually costing you.

Book the call

Thanks for reading,

Andrew Keener

Operations & AI Strategist

Keen Alliance Consulting