Your AI Narrative is Not a Moat

Hyperscaler wins aren't strategy. It's time to build a business that lasts beyond the cycle.

"And you may ask yourself, 'Well… how did I get here?'" — Talking Heads, “Once in a Lifetime”

If your multiple just doubled, congratulations — but now what?

The AI boom has made you a winner. But unless you use this moment to transform, you'll be yesterday's news when the music stops. This article is not about riding the wave. It's about building the raft.

One week you're trading like a boring industrial. The next, a hyperscaler gives you a nod, and suddenly you're AI infrastructure. Your multiple doubles. Your investor deck gets a new title slide. You're a hero.

And maybe you deserve it. But don't confuse market sentiment with a durable shift in your business model.

Most of today's AI winners in the industrial world got pulled into this moment by one or two design wins. That's not a strategy — that's timing. That's not scale — that's luck.

If your entire company is rebranding itself around a single RFP win, ask yourself: is this conviction, or is this cosplay?

Fragile exposure: One customer, one product, one cycle 🔍

AI is a huge market, but it's also a concentrated one. Most of the dollars are coming from five or six hyperscalers. A single win — or a single product line — can elevate your profile, but it's still just that: exposure, not entrenchment.

I once watched a supplier's entire market cap swing on a single line in a hyperscaler's earnings call. The CEO mentioned them in passing, and boom — stock up 20%. The next quarter? No mention. Down 18%. That's the danger of living in someone else's narrative.

Many of these companies now valued as AI darlings are really just hyperscaler-dependent mono-product businesses. If your total exposure to AI is one SKU in one rack at one datacenter — you're not strategic. You're a variable cost line item.

You need a second engine before the first one sputters. You need a story that outlasts your current cycle.

And don't fool yourself into thinking program wins are moats. The only moat you may have is how well you integrate with a hyperscaler's procurement playbook. The second they want your margin — or your IP — they will build it themselves. Just ask the vendors who thought they had AWS locked in until the in-house cooling division got funded.

Don't drink your own Kool-Aid — make others believe it while you stay ruthless about what's real.

Valuation is a weapon — use it 💰

We all know there's a football field of valuation methods — DCF, precedent transactions, comparables, sum-of-the-parts. But when it comes to how capital gets allocated in moments like this, it's almost always:

Valuation = EPS × Multiple

EPS is hard-earned. But the multiple? That's magic. It's belief. It can double in a quarter.

Which means this: When your multiple is high — that's when you act.

Raise capital. Make acquisitions. Issue paper. Even cash deals make sense when your equity is trading like a rocket ship. Buying something at 7× EBITDA with cash when you're at 16× is value creation 101.

One deal we looked at didn't pencil at 10×. But at 15×, it was a slam dunk. Same target. Different cost of capital.

This is your window. If you wait until your stock comes back to earth, you're playing defense. The bold move is building while you're overvalued — because that's when it costs the least.

Stop pretending AI is a moat 🧠

AI is not a moat. Your relationship with hyperscalers might be. Your engineering quality might be. Your global deployment capabilities might be. But your AI label? That's just what's fashionable this quarter.

Vertiv cratered when AWS built its own liquid cooling. Broadcom got called out on custom chip concentration. Nvidia keeps moving closer to the value chain's core. This is not a safe neighborhood. It's a war zone — and you don't own the land.

Too many companies have convinced themselves that an AI adjacency is a structural advantage. It's not. It's an opportunity. And if you're not converting that opportunity into something lasting, you're just a temporary markup on someone else's spreadsheet.

Frame yourself more broadly. You're not an AI company. You're a scaled technology platform that enables compute, power, cooling, systems. AI is a tailwind. But without a keel, you're still drifting.

What to do differently — for real ⚖️

Here's how to convert hype into resilience:

1. Present a broader investor narrative. Not just "we win in AI." Show a future where you thrive without it. Highlight parts of your business that generate recurring cash flows or strategic control. Make it credible. Make it long-term.

2. Consider a merger of equals. Find a peer with similar froth and different exposure. A hardware player + a cooling integrator. A rack builder + a deployment team. Both overpriced? Great. Trade stock. Combine scale. Hedge risk. Create synergy in supply chain, customers, and capabilities.

3. Use cash for tuck-ins. No one wants your expensive equity? Fine. Use cash. The return math still works. Buy something at 7×, bring it into your 16× world, and watch the arbitrage re-rate your balance sheet.

4. Grow organically. Yes, it's the hardest path. Legacy business units are tired. Internal orgs resist change. Product teams are underfunded. But this is when you invest — when the capital allocation police are distracted by buybacks and AI slides.

Reignite the roadmap. Open new geos. Build the muscle that keeps you alive after the hype.

Why divestiture won't save you ❌

Selling off your non-AI assets sounds clean. But the market won't pay your AI multiple for a legacy business. The arbitrage destroys value. And worse — you might need that cash cow when the AI capex train slows down.

The right move isn't to dump your past. It's to reinvest selectively. Think like a portfolio manager, not a founder exiting.

Learn from flameouts — and the ones who evolved 🔥

3D Systems chased additive manufacturing hype. GoPro thought action cams were a category. Several AI-labeled startups in India raised at 40× revenue — and disappeared in 18 months.

Meanwhile, Flex and Jabil — once derided as low-margin EMS players — evolved. They leaned into power, racks, cooling, integration. They built complexity. Now they're core to the AI stack — but not dependent on it.

One built a product platform. The other built systems expertise. Neither bet the farm on a single SKU.

Final thought: Don’t wake up asking “how did I get here?” 🎵

"Same as it ever was." That’s the Talking Heads lyric. A warning against complacency. A signal that inertia is dangerous.

This market is rewarding you for being in the right place at the right time. But that won’t last. Multiples fade. Hype ends. And the companies that survive will be the ones who saw it coming.

Use your inflated valuation to build. Diversify. Acquire. Reinvest.

Because when the music stops — you’ll either be leading the next wave… or watching your stock chart become someone else’s case study.

Alok K. Agrawal is the Managing Director and CEO of Agrawal Capital, LLC. He has served as Chief Strategy Officer at three companies across multiple industries, and now advises and invests in early-stage ventures.