Nobody knows which AI companies are going to win long term. But there are three that are showing strengthening signs in all vital fields. Three companies that adapt to the model shift and come out more important on the other side, not less. Cloudflare. Palantir. Shopify.

Everyone else gets commoditized, partnered-into-irrelevance, or buys their way out of a product gap they should have built. These three will not.

That is the thesis. Now the cold lens. Same ten dimensions we run on every name in the FCF Tracker. Honest scores beat high scores. Here is what the work says.

Cloudflare is the edge of the internet. 300+ cities, a network closer to the end user than anyone else's, and a developer platform — Workers — that turns that network into programmable compute. Workers AI runs inference on the same edge that delivers your website. That is not a demo. That is architecture.

The moat is the network. Matthew Prince has said for years that most of their competitive advantage reduces to one number: how many cities they are inside single-digit milliseconds from the end user. Every year that number widens. You cannot buy Cloudflare's network. You have to build it. And nobody is spending the capex to catch up.

Revenue around $1.7 billion growing 30%+ a year, gross margins around 77%, the top line is compounding. The issue is lower down: GAAP losses, thin FCF, reinvestment near 100%. Full land-grab mode.

Then the price. Twenty-five to thirty times sales depending on the week. That is priced for perfection. Any top-line wobble gets punished 50%.

Palantir's Foundry was enterprise software. AIP — their AI Platform — is something else. It is the piping that lets a Fortune 500 deploy LLMs against their own proprietary data without turning every workflow into a two-year consulting project. The bootcamp motion (fly in, deploy in five days, land a contract) is converting at a rate I did not think was possible twelve months ago.

US commercial revenue is growing 50%+. Government revenue compounds at glacial switching speeds. Net margins are expanding fast. Rule of 40 is off the charts. Balance sheet is a fortress: $4 billion+ in cash, zero debt.

The problem is the price. Seventy to eighty times sales. The bull case requires AIP becoming the standard enterprise AI layer and revenue compounding 40%+ for five more years. That might happen. The market is already pricing it in. If the ramp decelerates even a little, the multiple resets painfully.

Shopify is what happens when you build infrastructure for every SMB with a product to sell and then stay focused for fifteen years. Eight million active merchants, global. Shop Pay pulling one-click conversion rates competitors cannot match. Shopify Magic weaving generative AI into every merchant workflow — product descriptions, images, email campaigns, ad copy.

Divesting Deliverr in 2023 was the moment this thesis got boring in the best possible way. They stopped trying to be Amazon. They went back to being the best platform for everyone who is not Amazon.

Revenue around $8.9 billion growing 25%, FCF roughly $1.6 billion, net margins back in double digits, ROIC climbing. Balance sheet: $5 billion+ in cash, minimal debt. Capital returns are light — occasional buybacks, no dividend — but reinvestment is now disciplined, not desperate. Valuation around 15 to 17 times sales is not cheap, but it is the most reasonable of the three.

Same ten dimensions, scored one to ten. Sum out of 100.

1. Essential or nice-to-have? Cloudflare 9. Palantir 9. Shopify 9. All three sit under something real. Internet plumbing, enterprise AI, SMB commerce. None is optional for the thing it enables.

2. Current moats? Cloudflare 8. Palantir 9. Shopify 8. Cloudflare's network scale. Palantir's entrenched government codebase and Foundry stickiness. Shopify's merchant lock-in and Shop Pay flywheel.

3. Moats expanding? Cloudflare 9. Palantir 10. Shopify 8. Palantir wins. AIP is turning services revenue into scalable product revenue faster than I have seen a platform pivot before.

4. Balance sheet strength? Cloudflare 7. Palantir 10. Shopify 9. Palantir is a fortress. Shopify is strong. Cloudflare carries more leverage.

5. EPS accelerating? Cloudflare 5. Palantir 10. Shopify 9. Palantir is inflecting sharply. Shopify is past the trough. Cloudflare is still building.

6. Net margins increasing? Cloudflare 6. Palantir 9. Shopify 9. Same story.

7. ROIC profile? Cloudflare 4. Palantir 7. Shopify 7. Two of the three are above our 15% bar. Cloudflare is still reinvesting so hard the returns have not shown up yet.

8. Reinvestment rate? Cloudflare 10. Palantir 7. Shopify 7. Cloudflare is all gas, all the time. The other two are now more disciplined.

9. Capital return? Cloudflare 2. Palantir 3. Shopify 3. All three are weak here. They reinvest instead. Fine at this stage.

10. Valuation (reverse DCF)? Cloudflare 3. Palantir 1. Shopify 4. None is cheap. Shopify is the least expensive. Palantir is priced for the best-case scenario. Cloudflare is priced for perfection.

Totals:

Palantir wins the scorecard by a nose. Best balance sheet, best moat expansion, best earnings acceleration. Loses on price. Shopify comes in second with the most durable business at the most reasonable valuation. Cloudflare is the infrastructure pick with the longest runway, the thinnest near-term margin story, and the most expensive ticket to ride.

If I had to own one today for the next decade, it would be Shopify. Reasonable price, strong balance sheet, AI embedded in every SMB workflow, and the discipline to stop chasing fights they cannot win.

If I had to own one if the multiple halved, it would be Palantir. Fortress cash, fastest earnings acceleration I can find, and a product motion that genuinely keeps expanding its moat.

If I had to own one if the sector sold off 40%, it would be Cloudflare. A fifty-cents-on-the-dollar Cloudflare is a generational buy on the most important infrastructure layer of the AI decade.

Three AI winners. Three different entry points. One scorecard that tells you which to buy on which day.

Every thesis needs its bear case written down before you buy. Cloudflare's risk is margin discipline — if they cannot stop burning cash before the network moat needs defending, the valuation cracks. Palantir's risk is deceleration — AIP bootcamps are converting today; if the enterprise pipeline ever blinks, 70x sales turns into 25x sales very fast. Shopify's risk is competitive narrowing — Amazon's seller tools, TikTok Shop, and Stripe's merchant push all circle the same SMB. Merchant lock-in is deep but not permanent.

Buy none of these without writing your exit trigger next to the entry price. That is the discipline. The scorecard tells you the what. Your own conviction has to tell you the when.

Want to build your own repeatable investment process? Free tools and frameworks → strategysprints.com/tools

Also worth reading: The Sales Show — B2B sales acceleration for founders. 800+ episodes.
Subscribe → https://thesalesshow.beehiiv.com