Apollo Sees No AI Payoff Outside Big Tech. Can PE Prove It Wrong?

🧭 Apollo Just Said the Quiet Part Out Loud

One of private equity's biggest names has looked for the AI payoff outside Big Tech.

It didn't find one.

In a Daily Spark note published October 24, Apollo chief economist Torsten Slok argued that revenue per worker and profit margins are rising for the Magnificent Seven, but staying flat for the other 493 companies in the S&P 500. Productivity data, he wrote, shows no sign of an AI boost either.

The sharper line: mainstream companies are "starting to view token spending not as a source of growth but as a cost center that weighs on margins."

The note drew pushback online, and it echoes what CEOs told PwC this year: 56% of more than 4,400 surveyed saw neither higher revenue nor lower costs from AI. Only 12% saw both.

Here's why it matters to you. Your portfolio companies aren't the Mag 7. They look a lot more like the S&P 493.

A note from the team behind this newsletter: we are building OpenMap, The Agent Network, the connective tissue for the world's AI agents, starting with every person and every local business. We are raising a £5M seed. Meet us at openmap.ai/investors.

💡 From AI Spend to AI Return

For two years, the question in every board pack was: are we using AI?

The question now is: what did it earn us?

That's a much harder question. Usage shows up in a vendor invoice. Return has to show up in revenue, margin or cost, inside a P&L someone can audit.

The token bill arrives every month.

The EBITDA line hasn't moved.

That gap is exactly where a value creation plan goes to die.

🔍 Why the Payoff Keeps Missing the Portfolio

Big Tech has an unfair advantage. Its data already lives in one place, defined one way, owned by people who build models for a living.

A typical portfolio company has none of that.

Customer data sits in the CRM. Cost data sits in the ERP. Contracts sit in a data room. Every CFO defines "gross margin" slightly differently, and the AI pilot was fed a spreadsheet export from last quarter.

So the model writes a clever summary. Nobody can tie it to a dollar.

The pilot gets praised. The payoff never gets proven.

This isn't a model problem. It's a context problem.

AI can't move a margin it can't see.

⚙️ How It Works: Connect the Portfolio, Then Measure the Payoff

PrivateEquities.AI gives your fund one connected view of the portfolio, so every AI use case runs on real numbers and can be measured against them.

1. Connect, Don't Collect
   The platform plugs into the systems your portfolio companies already run: ERPs, CRMs, data rooms, warehouses and BI tools. No data migration. No integration project per company.

2. One Context Graph
   Customers, costs, contracts and KPIs are linked across the portfolio, so "margin" means the same thing in every company and every board deck.

3. Agents on Top
   ProfileAI, CleanseAI and other in-house agents validate data and harmonize formats before a figure reaches a model, a forecast or an LP update.

4. Always Live
   When a source system changes, the graph updates. You track AI's impact on today's numbers, not last quarter's export.

📊 The S&P 493 Problem: Before and After Context

The AI is the same. The data underneath it is what changes the result.

💸 When Tokens Become a Cost Center

Apollo's point about token spending deserves a second look.

In a separate note on October 20, Slok argued that slowing growth in corporate token spend suggests demand is highly price-sensitive.

Translation: companies will use AI, but only at the right price.

Much of that price comes down to which model each task runs on. The PrivateEquities.AI Inference Cost Calculator shows up to a 35x price gap between frontier and open-weight model classes for the same workload.

A frontier model reviewing a complex contract may earn its keep. The same model tagging invoices probably doesn't.

The capital is committed. The overpay is optional.

🧩 A Day in the Life of an Operating Partner Using PrivateEquities.AI

8:00 AM - The Payoff Question
The investment committee wants to know which AI projects are actually earning money. Every portfolio company's revenue, margin and cost data is already in the context graph, on one definition.

10:30 AM - Finding the Real Win
A pricing assistant at one industrial company is lifting gross margin on quoted jobs. A support chatbot at another shows heavy usage and no change in cost to serve. Now you know which is which.

2:00 PM - Cutting the Token Bill
You review model classes for each workflow. The pricing logic stays on a frontier model. Document tagging moves to a cheaper class.

5:00 PM - Rolling It Out
The pricing use case is queued for three sister companies. Same definitions, same graph, no new integration project.

That's how an AI pilot becomes an AI return.

Most PE-backed companies look far more like the 493 than the Mag 7. That makes private equity the place where the AI story gets proven, or doesn't.

🚀 Ready to Prove Your AI Payoff?

If your portfolio data still lives in a dozen disconnected systems, every AI project starts on guesswork and ends without a number.

PrivateEquities.AI helps funds turn portfolio chaos into operating clarity.

👉 Book a Demo
See how connected context changes your value creation plan.

👉 Run the Inference Cost Calculator
Find out what your portfolio is overpaying for AI.

🔔 Key Takeaways

1. The Payoff Is Narrow. Apollo sees AI gains in Big Tech, not yet in the rest of the S&P 500.

2. Your Portfolio Looks Like the 493. Ordinary operating companies are where AI's real test happens.

3. Context Comes First. AI can't lift a margin built on disconnected data.

4. Measure Return, Not Usage. Tie every AI project to revenue, margin or cost.

5. Price Each Task. Model class is one of the biggest levers on token spend.

🧠 Final Thought

The funds that win the next cycle won't be the ones that spent the most on AI. They'll be the ones that can show, company by company, exactly what it earned.

PrivateEquities.AI is making that possible.

Warm regards,

PrivateEquities.AI Team