Some entrepreneurs would give their first prototype to replicate Roamz's two-year journey — $3.5 million investment, 140,000 downloads and a dozen staff — but unsatisfied founder Jonathan Barouch has killed his darling so that he can realise his dream to "engineer serendipity".
In a fortnight, he will release a new version of Roamz, which he has built from scratch. Roamz analyses social media content to recommend popular services and events in a user's proximity, eg a new coffee shop or Vivid live events.
He rushed the last release of Roamz, but over the past few months has drawn on user feedback, and the company's war chest — funds committed by ASX-listed digital marketing company Salmat, which owns 60 per cent of Roamz, a NSW Government grant and Barouch's own seed funding (from when he sold his previous start-up) — to develop the version he always wanted.
"We started from scratch," said Barouch, who will demo the new version to American investors and customers over the next fortnight.
"We had this notion that when [people] leave the house in the morning, they're looking for serendipity, but the reality is they don't. They leave the house to go to work, to look for coffee, or take the kids to the park; serendipity just happens by happenstance."
"As a developer, to try and engineer serendipity as a primary use case is pretty tough. So we want to give [users] the tools to find what they are looking for, as the primary use case, but then also surprise them with things they might not be necessarily looking for," he said.
"So maybe, on a Saturday morning, you're looking for coffee and you don't have a coffee shop, but Roamz can tell you there's something really special going on in the park nearby, or there's a fair in Paddington, and you wouldn't have known otherwise. That's where we see serendipity playing a role."
"This is the product we wanted to launch at the end of last year, but we didn't have enough time and the benefit of real-life user feedback."
The new version of the app won't only try to use social media feeds to guess what people like, but will instead ask users to find events that are relevant to them, and then suggest some other events that might be of interest.
"We previously looked very heavily at people's behaviour — things they liked/didn't like — and their past behaviour, and trying to heavily screen content to show them stuff we thought they were interested in. The reality is, it's very tough to get that right all of the time, so you'd end up having a lot of noise, and then one or two things which were actually really relevant."
"I think we're getting closer. It's very hard to do, because even if you nail it, there's so many different variables — somebody's mood; whether they've had a good night's sleep; whether they're rushing to an event — there's so many things that have to go right, and so many stars that need to align. I suppose, as a developer, we can help some of those stars to align, that's what we're trying to do."
Barouch has experienced the full start-up lifecycle, and has learned from failures.
The start-up is essentially starting from square one.
GPS-enabled smartphones combined with real-time social information provide the perfect device for people to experience something new and, most importantly, when they want, as addressed by the start-up's new angle.
Innovation may be stifled if majority stakeholder Salmat influences product development, forcing Roamz to obsess about the duty to return value to shareholders.
This is probably Barouch's last chance to make his app work on his terms, which can be the greatest motivator of all.