Chaos, a global 3D-visualization software company, had a huge e-commerce customer base and no way to tell which accounts were worth working. We used Clay to enrich all of it, then turned that into a real-time pipeline that keeps every new customer sales-ready.
Official Clay PartnerRevOps Shop is a Clay partner, and a proud Clay customer.
Workspace ~#9,000We've been building in Clay for years, since it was a much smaller platform.
In nearly every engagementClay runs inside almost every active RevOps Shop client.
170,000 accounts · one Salesforce org
A global software leader that couldn't see its own customers.
Chaos sells 3D-visualization software through e-commerce and 350 reseller partners. Many buyers purchased with a personal email, so nobody could tell which accounts were real business opportunities. Reps cold-sourced new leads instead of working the customers they already had.
The ask: identify the business behind every account, size its potential, and keep doing it automatically as about 15,000 new accounts arrive each month.
More than a third of accounts were just a personal email address, with no company attached.
170kcustomer accounts in one Salesforce org
350reseller partners worldwide
~15knew accounts every month
37%signed up with a personal email
What we built
A data foundation, built in Clay.
Every account traced to a real company, enriched, scored and handed to sales. First as a 170,000-record backfill, then as a real-time pipeline.
1.Find
Trace each personal email to a LinkedIn profile, and from there to the company behind the purchase.
2.Enrich
11 to 14 data providers, cheapest first, with Claygents and a custom AI scraper filling the gaps.
3.Score and hand off
Every account graded High, Medium or Low and written to Salesforce, so reps start with the best ones.
Proven on 1,000 records first
99.1%industries mapped
97%company LinkedIn pages found
80.4%personal LinkedIn profiles found
19enrichment columns, 5 of them AI agents
What each account looked like after enrichment (illustrative)
Account email
LinkedIn match
Company domain
Industry
Employees
Seat potential
Confidence
j.moreau@gmail.com
Found
atelier-ln.fr
Architecture
11–50
Medium (5–20)
High
k.ito@studio-kumo.jp
Skipped (business email)
studio-kumo.jp
Media & entertainment
51–200
Large (20+)
High
r.patel@outlook.com
Found
brightline-id.com
Interior design
1–10
Small (1–4)
Medium
m.schulz@web.de
No match
–
–
–
–
Low
Example rows with made-up data. Highlighted cells are what Clay added, and blanks stay visible because what we couldn't find tells sales something too.
Each new order triggers a webhook to Clay, runs the same waterfall in 12–14 minutes, and lands back in Salesforce as a scored account. Live since February 11, 2026, at about 800 records a day.
Webhook in, write-back out
Salesforce sends each new order to Clay, and Clay updates Salesforce directly. No manual exports.
Three-field monitoring
Requested date, completed date and an open-request flag feed a failure dashboard. It has already caught a sync gap.
90-day cooldown
Accounts aren't re-enriched more than once every 90 days, so credits go where they matter.
Original data protected
Company data writes to dedicated enriched fields, so customers' billing details are never overwritten.
The curveball
Days before the deadline, a key data vendor changed its API.
The LinkedIn provider at the heart of the waterfall pushed an unannounced change that cut accuracy and volume, and for personal-email accounts LinkedIn was the only route to the company. We brought options to Chaos's leadership right away, adapted the workflow, and still delivered the core backfill on the deadline.
Results
On deadline, above the accuracy target.
Industry match rate across all 170,000 accounts
Target
70%
Delivered
70.8%
Including personal-email accounts with no company signal at all. Delivered 45 days after terms were agreed.
Phone coverage on qualified contacts
EMEA
63%
APAC
57%
Reserved for the accounts with the most seat potential.
170kcustomer accounts enriched in Clay
~800new records enriched every day, automatically
12–14 minfrom new order to a scored account in Salesforce
From anonymous to actionable.
What the data revealed
It changed how leadership saw the business.
Leadership assumed about 90% of customers were architecture, engineering and construction firms. The enriched data showed it was about half. It also showed reseller partners sell 50–100% more seats per customer than direct sales.
The company went on to split its sales team into Inside (5–20 seats) and Strategic (20+ seats) tiers based on the enriched seat potential, and Clay now fills a matching segment on every new account.
Same engine, new revenue
Turning unlicensed use into warm accounts.
The company's piracy-detection software spots unlicensed use but can't name the company behind most flagged users. We pointed the same Clay engine at that gap, from a raw email to a validated company and the right decision-makers.
~2/3of flagged users had no identifiable company
36kflagged records in the pilot
~4.6kcompanies identified
~13kcontacts found at those companies
“Huge dream.”
That's how Chaos's head of data and analytics described the results, adding that they “expected less” given how hard the dataset was.
Why Clay
Why Clay was at the center of this project.
It scaled to enterprise volume
170,000 records, split into workbooks to stay under the 50k row limit
A shared company table so no company was paid for twice
A real-time webhook pipeline on top of the backfill
Waterfalls kept costs in check
Many providers in one table, cheapest first
Claygents and a custom scraper only where vendors came up empty, including firms under 30 employees
Phone lookups only for the best accounts
Sales could trust the output
A 1 to 10 confidence score on every AI answer, shown as High, Medium or Low
Checked against Chaos's own 250-record spot-check
Built in Chaos's own workspace, so the data stays theirs
RevOps that turns your stack into predictable pipeline.
Data, operations and enablement in one team, with Clay at the center of the stack.
Case study · Clay Partner
