Bill.com's paying customers were a fraction of its network

The record

Bill.com Holdings, Inc. filed its Form S-1 on November 15, 2019 ahead of its IPO. The S-1 discloses revenue of $64.9 million in fiscal 2018 and $108.4 million in fiscal 2019, alongside Total Payment Volume, its term for the value of transactions processed on its platform, of $49.6 billion in fiscal 2018 rising to $71.3 billion in fiscal 2019, a 44% increase. The filing defines "network members" as its paying customers plus the suppliers and clients those customers transact with on the platform, reporting over 1.8 million network members as of June 30, 2019, against a paying customer base the filing separately states exceeded 81,000 as of September 30, 2019. The filing states that approximately 80% of both Total Payment Volume and transaction count came from repeat customers in every month of fiscal 2019.

What the documents establish

The S-1 is Bill.com's own account of a network in which paying subscribers represent a small fraction of the total participants connected to the platform; the 1.8 million network-member figure and the 81,000 customer figure are both issuer-reported counts as of different, nearby dates, not the same measurement. A final amendment, the S-1/A filed December 10, 2019, restates the same network-member and payment-volume figures and sets an assumed IPO price of $20.00 per share, the midpoint of its range, confirming those network figures were unchanged at the filing closest to pricing. The filing states the company believes it can convert more network members into paying customers but does not disclose a conversion rate or target.

The operating read

A two-sided payments platform disclosing a network of unpaid participants roughly twenty times the size of its paying customer base is describing a distribution channel, not yet a fully monetized one. Editorially, a reader evaluating a similar network-effect claim should ask what fraction of the disclosed network actually pays, since a large network of non-paying participants is evidence of reach, not by itself evidence of the take-rate or conversion economics that would make that reach valuable, and Bill.com's own filing does not quantify the conversion rate it says it expects to improve.

What to check before you decide

Before treating a disclosed network size as a moat, check the following.

• What share of the disclosed network are paying customers, versus unpaid participants counted only because they transact with a customer?
• Does the filing disclose a take rate or revenue-to-payment-volume ratio directly, or does it need to be inferred from separate figures?
• Has a later filing disclosed the conversion rate from network member to paying customer that this S-1 does not quantify?

Bill.com's own figures document a large connected network and a smaller paying base as of specific 2019 dates; the filings describe that structure without quantifying how efficiently one becomes the other.