How Nomorobo grows

Entered the FTC's crowd-sourced Robocall Challenge (found via TechCrunch) while between companies. Instead of a 30-page business plan, submitted a one-to-two-page description, a working Twilio prototype using simultaneous ring (simring) to get a copy of each call, the full source code and database schema, and a demo video. Pitched distribution (going to consumers directly, around the carriers) rather than an algorithm, reading between the lines of the rules. Named it something catchy (Nomorobo, 'no more robocalls') and grabbed the domain, while the co-winner's entry had a long unmemorable name.

At the FTC press conference the co-winner declined to talk for fear of idea theft; Aaron took every interview (CNN, Fox Business, New York Times) and described the approach at a high level without technical details. The site was just the demo video and an email box ('we'll let you know when we launch'). Also used Congressional testimony (July 2013) as more media. Treated the daily inflow of emails as demand validation before any revenue.

Rather than generic national stats ('8 billion robocalls') like competitors, used carrier and honeypot data to pitch what is new: e.g. to an Austin reporter, 'here are the top scams hitting Austin right now', with recordings; later new angles such as a deepfake Biden voice promising a Publishers Clearing House win. Kept in touch with reporters who asked for a monthly update and came back when they had a gap. Recordings also lived on the landing pages.

After launch costs threatened 6 weeks of runway, negotiated with Twilio and opened a friends-and-family convertible note (mother, stepfather, a few founders with exits). Opened the first note for $500K but closed ~$50K, then $250K, extended to about $1M pre-seed, then a ~$1.25M Series A that reached profitability. No milestones attached; money went to telecom and developers, not ads or servers.

After the mobile launch, the 14-day trial plus Apple's payout lag (roughly 30 + 30 days) left a cash gap, and investors offered only very poor terms. Asked Twilio for Net 90 instead of Net 30 for 3 months (they agreed), secured a bank line of credit as backup (not drawn), and paid some bills late, rather than accept a deal that would have left the founder with nothing.

People google unknown numbers while the phone rings. Built a honeypot of 300,000+ company-owned phone lines (old 'dirty' numbers bought from Twilio and carriers) that record and transcribe robocalls, never recording consumer calls. Generated a landing page per phone number showing it is a scam, with the recording and transcription and a CTA ('block this and 7 million other robocallers') into a frictionless trial. The acquirer later planned to point paid traffic at the same pages.

Robocall blocking needs data that only comes from already blocking calls. On the FTC's suggestion, filed a FOIA request for five years of Do Not Call complaint data, received it on CD-ROM on day one, and crunched it into the initial blocklist, then gave landline blocking away to gather real-time data.

When iOS 10 (2016) allowed call-blocking apps, launched as one of the first subscription apps: $1.99/month or $20/year via Apple in-app purchase. Did the math and found Apple's fee cheaper than off-the-shelf billing run in-house at that price point, despite pushback that $2/month was too low. Ongoing per-call costs made one-time pricing impractical.

Built a two-week free trial into the mobile app from day one rather than freemium, sized so that users would receive at least one robocall during it, and told users plainly to just cancel if it didn't work for them. Kept friction as low as possible.

Landline blocking has always been free with no paid upgrade. It works as the acquisition channel and data source: every landline user's calls improve the blocklist that powers the paid mobile apps, and every user's email can be cross-sold. Chose to spend on running the free service instead of ads; never did paid acquisition. Aaron frames free/freemium as a user acquisition path, not a revenue model, and says free users must deliver some other value (data here).

Used the email addresses of free landline users (who already loved the service) to launch the mobile app: 'Hey, you love it on your landline, why don't you try it on mobile?', pointing to the two-week free trial.

Avoided big-bang launches after being burned at the 2013 launch (email servers flagged for spam, servers overloaded). Releases go out when ready; after bug fixes, emailed the reporters who had already covered robocalls ('Hey, Herb, I know you wrote about this all the last time'). Then moved straight on to the next build (Android).

Because landline blocking was free, carriers such as Time Warner (now Spectrum) came inbound asking to offer it so they could say they provided robocall blocking without paying. They added one-click activation in customer control panels that calls Nomorobo's API with the phone number and sets up the simring number automatically.

Used Apple's privacy-preserving call-blocking model (load a blocklist that iOS proxies) instead of fighting it; the basic product doesn't even need the user's phone number, and no tracking. Pitched Apple for App of the Day on the grounds that it protects people and is privacy-conscious.

As a solo founder, first carved off customer support to a Techstars acquaintance (a contractor who set up Zendesk); when he left he handed it to his brother, who later became product manager. For development, bought a Laravel author's book-plus-consultation package ($200 tier), hit it off, and contracted him for years. Never posted a job ad or ran interviews; kept everyone as contractors to stay small and focused, with Aaron keeping the developers' 'hopper' full.

Nomorobo got inbound interest weekly. An earlier investment-banker process produced only lowball offers, which were declined while the business kept throwing off cash. When a large company made an inbound offer, Aaron took it to the board and then told all previous suitors 'we're not running a process, but we got this offer' ('two is one, one is none'). Chose a smaller acquirer that fit the product better.

When STIR/SHAKEN and carrier blocking threatened the core product (~2021), built Max: blocklist plus allowlist plus call screening for unknown numbers, at $5/month or $50/year, with an $80/year family plan, answering users' requests like 'can I protect my whole family?'. The old basic tier was $1.99/month.

The only early KPI was robocalls blocked, because it showed the product worked, people trusted it and it was growing. Any spending or fundraising had to serve blocking more robocalls.