Heidi takes $240m of its $340m round as revenue-share financing instead of equity

Heidi has raised US$436.6m in total ^[9]. Subtract General Catalyst's tranche and $196.6m sits behind the $900m mark ^[30]. The revenue-share money is about 71 percent of what was announced on Tuesday ^[29]. Against the $50m of annual recurring revenue Heidi reached in April, $240m is 4.8 times revenue committed to a single budget line ^[10][31].

Olivia Poh reported for Bloomberg that the Customer Value Fund is a debt-like instrument, and that General Catalyst holds no stock and no warrants in Heidi ^[6]. The fund pays for sales and marketing, gets its principal back, then takes a capped share of the revenue that spending produced ^[7]. Pranav Singhvi, who cofounded the fund, said it exists so companies with good unit economics can grow without funding that spend from their own balance sheet ^[8]. The announcement does not disclose the cap, the share rate or the term, so the cost of this capital cannot be compared with the dilution it replaced.

Weekly patient interactions went from about two million at the Series B to about 2.8 million, a rise of 40 percent in a year, and the country count went from 116 to 190 ^[13][12][32]. ARR over a two-year window went from $1m to $50m ^[10]. The windows differ, so the two series cannot be set against each other to show whether the revenue came from more clinicians or from higher prices per clinician. The figure closest to a rollout question is enterprise activation, which Heidi puts at around 62 percent on its own measure ^[14]. That leaves 38 percent unactivated on a measure Heidi itself defines ^[34].

For a company weighing the same structure, two tests decide it. The first is whether payback on marketing spend is measured per channel and per cohort rather than blended. A capped revenue share is priced off that number, and a blended average hides the channels where the payback never arrives. The second is whether marketing spend is the binding constraint at all. When the bottleneck is a clinical evidence package, a regulatory submission or a model you have decided to train yourself, a fund that can only pay for go-to-market cannot reach it. Heidi says transcription and note generation, which account for most of its compute, now run almost entirely on in-house technology ^[19].

"From the day I started Heidi, the ambition was always bigger than writing doctor's notes," Kelly said on Tuesday ^[1]. Part of the round pays for clinical evidence, quality management systems and regulatory submissions, and the company says it will work with regulators on the appropriate pathway ^[21]. The Next Web reported in August that NHS AI scribes were making errors in patient records, with Healthwatch raising concerns and the MHRA involved ^[26]. In Europe the route for autonomous clinical AI is slow: Vara spent years securing CE certification for AI that reads breast screening images without a second radiologist ^[25].