How Atlan powers its GTM foundation with Sumble data

About Atlan

Atlan is a leading active metadata platform that helps data teams collaborate, govern, and activate their data assets at scale. Built for the modern data stack, Atlan sits at the center of a company's data ecosystem — integrating with tools like Snowflake, Databricks, dbt, Looker, and 200+ other platforms.

After raising a $105M Series C at a $750M valuation, Atlan — the "control plane" for modern data teams — has been rapidly scaling from product-led growth to field-driven enterprise sales. The challenge? Selling data governance software to data people means facing buyers with high standards for data quality.

We sat down with Jake, who runs revenue operations at Atlan, to understand how they built an account scoring system that now powers nearly every GTM decision they make.

Tech presence ≠ depth (when every signal looks the same)

Atlan's buyers are data leaders at enterprises where data infrastructure is a major investment. These technical, time-pressed professionals demand tools that provide immediate value and support hands-on exploration rather than passive discovery calls.

Better data gives you an edge in GTM. There are tons of vendors that tell you that a company uses X technology. But that was valuable in 2010, not 2026. For instance, it is not helpful to know a company uses Snowflake (they have a free-tier).

Top GTM teams like Atlan crave deeper account intelligence to prioritize their efforts. They want insights on:

• Is Snowflake a side project at an account, or the backbone of their analytics stack?
• If an account uses both Snowflake & Databricks (happens often), which teams use what?
• Is their data team 5 people or 50? Growing or stagnant?
• Is the data team centralized, or scattered across business units?

Seeing a core data team using a tool versus a downstream analytics team completely changes your approach — from messaging and positioning to who you engage and when.

The problem wasn't a lack of data. It was that the data couldn't tell them anything meaningful about whether an account was actually a good fit.

The solution: using Sumble to build a "Fit Score"

Jake started by enriching all of Atlan's 70k target accounts with Sumble data directly in Salesforce. But rather than just storing surface-level technographic flags (like uses_snowflake = true), he focused on fields that could reveal depth of penetration of a technology within a company.

Sumble data in Salesforce gets piped to Clay through Snowflake using reverse ETL tools. For each account, Jake built a complex scoring model in Clay that weighs Sumble data to produce a fit score — then uses LLM prompts to generate a qualitative summary explaining why that account scored the way it did.

Clay writes this data back into Salesforce, where it becomes immediately useful for the entire team.

Turning Sumble numbers into narrative for sales teams

The power of Sumble account intelligence isn't in the raw numbers — it's in what those numbers reveal about a company's data maturity and trajectory.

Jake gave an example using a leading retail company: "They have 1.42% of their entire company headcount in data engineering. We think anything above 0.35% is significant. Their governance team is just two people, but their BI team is forty. That tells a story about where they are in their maturity journey — and where they're heading."

This kind of insight is only possible because Sumble breaks down technology usage by team, not just by company. When multiple teams are using a data warehouse, that signals the data team is embedded across business units rather than centrally managed — signaling both complexity and opportunity.

Implementation: from data to decisions

Jake spent his first three months at Atlan building out this scoring infrastructure. The technical setup involves several moving parts:

1. Salesforce is the initial data store, receiving enrichment directly from Sumble
2. Snowflake is the data warehouse, receiving Salesforce data via sync
3. Sumble data gets sent to Clay for running the scoring logic and LLM prompts
4. Clay writes back to Salesforce

The result: reps never leave Salesforce. Every account has a fit score, a qualitative summary explaining why, and deep links to explore further in Sumble. No tab-switching or manual research for reps.

Pre-call intelligence

Beyond the fit score, Jake built a "pre-intel brief" system. Reps enter an account ID and receive automated research in Slack that includes:

• A people readiness score based on team composition
• Suggested contacts to reach out to, with direct links to explore team structure
• A qualitative assessment of the account's data maturity

The Sumble data is woven throughout: "We can say, 'They have 41 governance team members,' and then we suggest the rep go check out who these people are — and we take them directly to that view."

Results: one score, multiple decisions

The same data forms the underlying source of truth for three different teams.

Why this approach works

It changes what reps say in the first 30 seconds of a call.

That kind of insight shows you understand their pain — and how to solve it. And it's only possible because Sumble shows depth, not just presence.