Meta committed between $115 billion and $135 billion to AI spending in 2026. That is more than most countries spend on their entire technology budgets. And their flagship AI model, codenamed Avocado, still is not good enough to ship.
Internal testing showed Avocado trailing Google, OpenAI, and Anthropic in reasoning, coding, and writing. This was supposed to be the model that proved Meta could compete at the frontier of AI. It could not. Reports suggest Meta's leadership discussed temporarily licensing Google's Gemini to fill the gap. Think about that for a second. The company that has been championing open-source AI for years considered renting a competitor's model because their own was not ready.
Meanwhile, on Friday, Reuters reported that Meta is weighing layoffs that could affect 20% of its workforce. That is roughly 16,000 people. The reason? To offset the cost of AI infrastructure.
Spending more does not mean getting more
Meta is spending more on AI than almost any company on the planet. Their model still trails the competition. And they are cutting a fifth of their people to pay for it. If that does not prove that money alone does not solve AI adoption, I am not sure what would.
I bring this up because I hear a version of this thinking from business owners all the time. "We need a bigger budget for AI tools." "We need to invest more in the technology." "Once we can afford the enterprise tier of [whatever platform], things will click." And it almost never works that way.
The businesses getting real value from AI right now are not the ones spending the most. They are the ones being the most deliberate about where they apply it. They pick a specific workflow, rebuild it properly, measure the result, and then move to the next one. That discipline matters far more than the budget line.
Focus beats spending every single time
An agency I worked with had a monthly tool spend of maybe a few hundred pounds on AI. Nothing fancy. Claude Pro for a handful of team members. A couple of automation tools. Total AI budget probably less than what Meta burns in about two seconds.
But they used it with precision. We identified the three workflows that were eating the most time, rebuilt each one with AI at the core, documented the process, trained the team, and measured the impact. Within two months, they had freed up 10 hours per person per week. Same team, same headcount. They took on two new retainer clients without hiring.
Compare that to Meta. Billions in spend. Model delayed. Workforce cut. Partnership deals falling through. The money did not solve the problem because money was never the problem.
What actually drives results
The pattern I see in every business that gets real value from AI is the same. They start small. They pick one workflow that actually matters. They rebuild it properly, with the team involved so the knowledge sticks. They measure it. Then they do the next one.
No massive upfront investment. No "AI transformation initiative" with a steering committee and a consultant's report. Just someone sitting inside the business, watching how work actually gets done, and identifying where AI creates genuine leverage. The rest is execution.
Meta's situation is an extreme example, but the lesson scales all the way down. If you are telling yourself that you need to spend more on AI before you will see results, you are probably avoiding the harder question: do you know exactly where in your business AI would make the biggest difference? Because until you answer that, more budget just means more dabbling at a higher price point.
I help businesses of 5 to 25 people find that focus and build from it. If you want results without the Meta-sized budget, here is how the fractional AI engagement works.
