Databricks and Snowflake — the two companies that have defined the modern data stack — are now both at just about $5 billion in ARR. Databricks just announced a $4.8 B run‑rate growing >55 %. Snowflake’s annualized Q3 puts them at roughly the same number, growing 29 %.

Same destination. Very different paths. And where they go from here? That’s where it gets fascinating.

---

The Tale of the Tape

Two companies. Same scale. One valued at nearly 2× the other. Growth is king in the Age of AI.

---

How Did We Get Here?

Snowflake’s Path:

Snowflake was the golden child, and in many ways, still is. The largest software IPO ever. Warren Buffett’s first IPO investment. 158 % NRR at listing. They separated compute from storage, made data warehousing actually work in the cloud, and grew like a weed. For years, they were the name in cloud data.

Databricks’ Path:

Databricks took the longer road. Born from the Apache Spark project at Berkeley, they were the “data science” company when Snowflake was the “data warehouse” company. They built MLflow, Delta Lake, and bet early that data + AI would converge. They stayed private, kept raising, kept building.

Now they’ve converged at the same revenue. But the growth vectors are pointing in very different directions.

---

The Growth Gap Is The Whole Story

Let’s be clear about what >55 % vs 29 % growth means at $5 B scale.

Databricks at 55 % growth (~$5 B ARR)

• $5 B → $7.75 B next year

• $7.75 B → $12 B the year after

• Path to $15 B+ ARR by 2027

Snowflake at 29 % growth ($5 B ARR)

• $5 B → $6.45 B next year

• $6.45 B → $8.3 B the year after

• Path to $10 B ARR by 2027

That’s a $5 billion gap in just two years if current trajectories hold. No wonder the valuation multiple is so different.

The question isn’t whether 29 % growth is good (it’s excellent at scale). The question is whether Snowflake can continue to re‑accelerate — and whether Databricks can sustain.

Snowflake in fact has re‑accelerated. Just not to Databricks’ level of growth.

---

The Net Retention Divergence

This is the metric that should keep Snowflake’s board up at night.

• Databricks: >140 % NRR – existing customers are spending 40 %+ more year‑over‑year. At $5 B scale, that’s extraordinary and signals expanding use cases.

• Snowflake: 125 % NRR … today – still great, but it was 158 % at IPO, then 171 % at peak, then 135 %, then 127 %, now 125 %. The trend line matters. Declining NRR each quarter points to product velocity or competitive problems.

Snowflake’s management says NRR has “stabilized” at 125 %. Maybe. But stabilizing at a lower level isn’t the same as re‑accelerating.

---

The AI Revenue Chasm

| Company | AI Revenue Run‑Rate |

|---------|---------------------|

| Databricks | > $1 B |

| Snowflake | $100 M |

That’s a 10× gap on the metric that matters most for the next decade.

Databricks’ AI products crossed $1 B because they’ve been building for AI since inception (MLflow, Feature Store, Model serving, Vector search, and now Agent Bricks – a platform for building AI agents on enterprise data).

Snowflake is playing catch‑up. Their Cortex AI family is showing “significant adoption” and AI is “linked to roughly 50 % of new bookings.” Those are good leading indicators, but $100 M vs $1 B+ is a canyon. Enterprises are making AI platform bets right now; whoever wins the initial deployment often wins the expansion. Databricks has a massive head start.

---

The Data Warehousing Counterattack

Databricks’ Data Warehousing business crossed $1 B in revenue run‑rate – Snowflake’s home turf – and is taking share.

The “Lakehouse” architecture (combining data‑lake flexibility with data‑warehouse performance) is winning enterprises that don’t want two systems. Why pay for a warehouse and a lake when you can have one platform that does both?

Snowflake’s response: embrace open formats (Iceberg tables) and build out AI capabilities. They’re now fighting a two‑front war—defending warehousing while attacking AI. Databricks is also fighting two fronts—but they’re attacking on both.

---

Where Does Snowflake Go From Here?

Advantages

1. Ease of use – “the easiest and most cost‑effective enterprise data platform.” SQL analysts love Snowflake.

2. Installed base – 688 customers paying $1 M+, 766 Forbes Global 2000 customers.

3. Partnerships – deep integrations with SAP, Anthropic, Google Cloud.

4. Cash – $4.4 B in cash and investments; can buy its way into adjacencies.

Bull case: AI products hit an inflection point in 2025‑2026 (Cortex AI, Snowflake Intelligence). NRR climbs back above 130 %. The “AI Data Cloud” narrative becomes reality, not aspiration.

Bear case: Growth decelerates toward 20 %, NRR slides to 120 %, Databricks takes material warehousing share while extending its AI lead. Valuation multiple compresses further.

What to watch: Q4 guidance (27 % growth) and whether AI revenue can 5‑10× in 2026. If Snowflake reaches $500 M+ AI revenue by FY27, the narrative shifts.

---

Where Does Databricks Go From Here?

Key considerations

1. The IPO question – Employees and early investors need liquidity; Series L includes employee liquidity, but the clock is ticking.

2. Growth sustainability – 55 % at $5 B is remarkable; 55 % at $10 B would be historic.

3. Margin expansion – FCF‑positive, but public markets will demand GAAP profitability at scale.

4. Competition – Not just Snowflake but hyperscalers (AWS, Azure, GCP) all want this workload.

Bull case: IPO in 2026 at $150 B+ valuation; AI revenue hits $3 B+; becomes the default enterprise AI platform. The “data intelligence” category belongs to Databricks.

Bear case: Growth slows to 35‑40 % as the law of large numbers kicks in; public debut disappoints relative to private valuation; competition intensifies from both Snowflake and hyperscalers.

What to watch: IPO timing, AI revenue trajectory, and whether they can maintain >50 % growth through 2025.

---

The $200 Billion Question

Together, these two companies represent over $200 billion in combined value and have defined how enterprises think about data for the past decade.

Take‑aways for founders and operators

1. At $5 B ARR, growth rate is still the primary value driver. 55 % growth at scale commands nearly 2× the multiple of 29 % growth. Don’t let up on the gas.

2. NRR is the canary in the coal mine. Snowflake’s decline from 158 % to 125 % foreshadowed everything else. If your best customers aren’t expanding, you have a problem.

3. Timing market shifts is everything. Databricks bet on AI convergence years before it was obvious. That head start is now a $1 B+ revenue stream.

4. Private markets are extraordinarily generous — for now. Databricks at 28× revenue while private; Snowflake at 15× while public. That gap will close when Databricks goes public.

5. Both can win, but one will win more. The enterprise data market is enormous, but platform consolidation is real. Whoever wins the AI workload likely wins the data workload too.

---

Two Adjacent Leaders. Same Revenue, Today. Completely Different Stories.

Snowflake is a great company with decelerating growth trying to pivot into AI. Databricks is a great company with accelerating growth that bet on AI years ago.

At $5 B ARR, the decisions both companies make over the next 12‑18 months will determine whether the gap widens or closes.

My guess? Databricks extends the lead. Their AI revenue advantage is structural, not cyclical. Their architecture is purpose‑built for what enterprises need next, and their execution has been flawless.

But Snowflake has surprised us before. And $74 B companies with 29 % growth and massive installed bases don’t go away quietly.

Place your bets.

---

Data from Snowflake Q3 FY26 earnings (December 3 2025) and Databricks Series L announcement (December 16 2025).