How BILL Traded Email Blasts for Real-Time Campaigns with Inflection.io
Not long ago, a significant share of BILL’s marketing week went into a single motion, run 20 to 30 times over: build a one-time email blast, wait on an analyst to hand-pull the product data, upload the list, and send — often to data that was already two weeks old. As Kaden Crutchfield, who runs product-led-growth operations at BILL, put it: they were “scaling through labor, not logic.”
The problem wasn’t effort. It was that every send started from a manual list and none of it could see what customers were actually doing inside the product.
The Challenge: labor, not logic
The old motion was a manual loop that baked in staleness at every step:
An analyst hand-pulls product-attribution data from the warehouse.
Someone uploads that list into the sending tool.
Copy and design to build a single-time blast — sent once, then thrown away.
It ships, often on a two-week delay → the wrong message, at the wrong time, to a stale list.
At 20-30 emails a week, the team’s effort didn’t match its return, because the stack couldn’t measure the product data it needed.
A legacy MAP that couldn’t hold the data. With Marketo, BILL’s product database was too big for the tool’s limited custom-object slots, causing the team to build workaround after workaround.
Usage blindness. In Kaden’s words: “We had no idea whether somebody converted in the product. We’d send them an email and pray it did what it was supposed to do.”
Vanity metrics. Clicks and opens — “metrics that didn’t reflect real customer behavior” — never in-product engagement or conversion.
Manual bottlenecks. Only Marketo specialists could build: submit a ticket → someone picks it up → build → QA → launch. The majority of campaign work was manual.
The Solution: Connect the product data, then act on it.
BILL rebuilt the foundation so marketing runs on live product behavior instead of hand-pulled lists.
One play, end to end. BILL’s product data — spanning both its core AP/AR product and Divvy’s spend & expense — lands in an enterprise data warehouse and event tracking layer. From there a connector reads that dataset straight into Inflection. The result is that a customer action becomes a marketable moment almost immediately: order a virtual spend card, and Inflection knows within seconds — so the next touch can teach the product instead of guessing.
That shift — from list-pulling to signal-reading — is the whole point:
“We’re not looking at clicks and open rates or even form fills. We’re looking at product success rates within our own product.”
— Kaden Crutchfield, product-led growth operations, BILL
On that foundation, BILL runs 50 specialized agents and tools inside Inflection against hundreds of millions of product insights, and it has opened campaign-building beyond the power users through intuitive, agent-assisted building – where a marketer who doesn’t live in the platform can talk to an agent and go from idea to execution, natively in Inflection or through Claude.
The Results
Marketing no longer waits on an analyst to hand-pull a list for every send — the data comes live from the warehouse.
The team measures in-product success rates instead of clicks, opens, and form-fills.
Campaign data went from two-week-old exports to a 30-minute refresh. As Kaden put it: “we could never have done that before.”
BILL’s marketing team used Inflection to trade manual, one-time email blasts for lifecycle marketing that runs on live product data. By connecting its product warehouses to Inflection through a single go-to-market source of truth, BILL replaced analyst-pulled lists and two-week-old data with product signals that refresh every 30 minutes — so marketers act on what customers are actually doing in the product, at the right time, instead of chasing clicks and opens.
See how Inflection can power the same outcome for your team. Request a demo → https://www.inflection.io/demo