Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

Managed $5Bn AUM | Early stage VC and 3x founder | Invested $770MM in 60+ companies, seeded 9 unicorns, and generated $10Bn in exits. jonathantower.wordpress.com San Francisco Joined December 2007

Tweets

3K
Followers

738
Following

277
Likes

485

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

tinyurl.com/4stuv5eb

0 0 0 2 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

9/9 Some AI companies are building highly scalable software platforms. Others are building sophisticated technology-enabled services businesses. Both can be valuable. But they deserve very different assumptions about future economics. The next phase of AI investing will be defined by who can generate the most scalable revenue. Link to full article here 👉 tinyurl.com/4stuv5eb

0 0 0 5 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

8/9 Investors should spend less time debating model benchmarks and more time understanding business-model quality. Questions that matter: 🔹 How much onboarding is automated? 🔹 How much customization is required? 🔹 How are margins evolving? 🔹 Is implementation shrinking over time?

0 0 0 3 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

7/9 The strongest AI companies will gradually convert labor into software. Custom workflows become product features. Implementation becomes automation. Human effort becomes repeatable product functionality. That's where durable value gets created.

0 0 1 6 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

6/9 This is where revenue quality becomes more important than revenue quantity. The key question isn't: "How much revenue does this company generate?" It's: "Does each new customer make the business more scalable?"

0 0 0 6 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

5/9 Two AI companies can each report $50M ARR. One deploys customers in days. The other requires months of implementation and ongoing human support. Same ARR. Very different economics.

0 0 0 16 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

4/9 As a result, many AI companies rely on: ▶️ Solution architects ▶️ Forward-deployed engineers ▶️ Customer success teams ▶️ Implementation specialists ▶️ Workflow consultants None of this is necessarily bad. But it matters.

0 0 0 4 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

3/9 The challenge is that many enterprise AI deployments don't look like traditional software deployments. They require: 👉 Workflow redesign 👉 Data integration 👉 Governance 👉 Security reviews 👉 Employee training 👉 Custom implementation Customers aren't just buying software. They're buying transformation.

0 0 0 4 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

2/9 For two decades, investors have rewarded companies that build a product once and sell it repeatedly. Why? Because software scales. Revenue grows faster than costs. Margins expand. Operating leverage compounds. That's what software multiples are supposed to capture.

0 0 0 5 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

2 days ago

1/9 Many of today's highest-valued AI companies are being valued like software businesses. A surprising number actually look much closer to services businesses. The market is increasingly confusing revenue growth with product scalability. 🧵

0 0 0 6 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

tinyurl.com/mu4njt3v

0 0 0 9 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

12/12 Much of the AI industry is still pricing intelligence as though it were infrastructure while customers increasingly experience it as labor. That distinction may seem subtle today. Over time, it may determine who captures the economics of the AI revolution. Full article here 👉 tinyurl.com/mu4njt3v

0 0 0 8 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

11/12 The coming battle won't be about who generates the most intelligence. It will be about who captures the greatest share of the value that intelligence creates. That's a very different competition.

0 0 0 5 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

10/12 That's great for customers. But what benefits customers doesn't always benefit shareholders. As intelligence becomes cheaper and more abundant, pricing power can come under pressure.

0 0 0 3 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

9/12 We've seen this movie before. ✅ Storage got cheaper. ✅ Bandwidth got cheaper. ✅ Compute got cheaper. There's growing evidence that intelligence is following the same path.

0 0 0 5 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

8/12 We don't pay employees based on emails sent, hours at a desk, or calls made. We pay them based on outcomes. So why are so many AI companies still charging by the token?

0 0 0 2 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

7/12 More importantly, AI is increasingly becoming labor. Traditional software made workers more productive. AI increasingly performs the work itself. That's a fundamentally different proposition.

0 0 0 4 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

6/12 But AI is increasingly governed by a different economic reality. Customers aren't purchasing intelligence as a raw input. They're purchasing what that intelligence enables them to accomplish.

0 0 0 2 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

5/12 Historically, usage-based pricing made perfect sense. ➡️ Cloud providers sell compute. ➡️ Data platforms sell storage. ➡️ Telecom companies sell bandwidth. The infrastructure itself was the product.

0 0 0 2 0

View Details

Jonathan Tower 🦄🦄🦄🦄🦄🦄 @jonathan_tower

5 days ago

4/12 The CFO doesn't report token consumption to the board. Investors don't value businesses based on inference volumes. Customers care about economic impact. That's what they pay for.