How Databricks Broke into the Enterprise on Microsoft’s Paper

How Databricks Broke into the Enterprise on Microsoft’s Paper

Databricks today: $7B+ revenue run-rate, $190B valuation. Yet for its first 3 years, it struggled to sell to enterprises. It broke into the enterprise on Microsoft’s playbook.

Databricks today: $7B+ revenue run-rate, $190B valuation. Yet for its first 3 years, it struggled to sell to enterprises. It broke into the enterprise on Microsoft’s playbook.

Now Databricks runs the same playbook for others, and still leans on all three hyperscalers.

A strong product enterprises couldn't buy

Databricks was founded in 2013 by the team behind Apache Spark, an open-source data engine. The plan: sell the best managed Spark and let the technology win.

It didn't work that way. For roughly the first three years, about 90% of sales came from digital natives and smaller companies.

Co-founder Arsalan Tavakoli explained why at TechCrunch Disrupt.

Enterprise deals stalled on people and process. The buyer had often bet his career on a competitor. Legal and security had to sign off. UBS alone had 13 CIOs.

The Microsoft turning point

In 2017, Microsoft made Databricks a first-party Azure service. Customers launched it from the Azure portal, and Microsoft handled the billing.

That removed the contract problem:

"It was on Microsoft paper… customers who had Microsoft agreements could just use it without needing to sign another paper."

It removed the access problem too.

The UBS meeting that had been so hard to get happened the next week. A Microsoft rep brought Databricks to the bank's global CIO.

The growth that followed

Demand arrived faster than Databricks could staff it. Microsoft asked for two Thai-speaking solution architects in Singapore. Databricks had none.

"We started getting pulled everywhere," Arsalan said.

ARR passed $100M in 2018. Microsoft invested the following year. Today the run-rate is above $7B, ~70x in 8 years.

Now Databricks runs the same playbook

Databricks now offers startups what Microsoft once offered it. Customers can spend up to 10% of their Databricks commits on eligible partner products, and Databricks pays its sellers for helping close those deals.

Hyperscalers are still the engine

Microsoft extended the partnership into the 2030s this July.

On Amazon Web Services (AWS), the Databricks business passed a $1B run-rate in 2024, with AWS Marketplace contract value doubling annually over two years.

Last December, AWS gave Databricks 7 Partner of the Year awards.

Google Cloud named it a Partner of the Year in 2024 and 2025.

Lessons for alliance leaders

  1. Get on paper the customer already signed. For most ISVs, that's a cloud marketplace listing that counts against the customer's commit.

  2. Staff for co-sell. A hyperscaler's field can pull you into accounts faster than you can hire.

  3. Follow the seller pay. A commit makes you eligible to be bought. Seller compensation gets you sold.

Which commits can your customers spend on you today, and whose sellers get paid when they do?

Scale to $100M+
via Cloud Marketplaces

Join 5,000 GTM leaders

Weekly Newsletter

Scale to $100M+
via Cloud Marketplaces

Join 5,000 GTM leaders

Weekly Newsletter

Scale to $100M+
via Cloud Marketplaces

Join 5,000 GTM leaders

Weekly Newsletter

Join 5,000 GTM leaders

Weekly Newsletter

Scale to $100M+ via Cloud Marketplaces

© 2026 Partner Insight

Join 5,000 GTM leaders

Weekly Newsletter

Scale to $100M+ via Cloud Marketplaces

© 2026 Partner Insight

Join 5,000 GTM leaders

Weekly Newsletter

Scale to $100M+ via Cloud Marketplaces

© 2026 Partner Insight