John Goleby Quit University, Shipped Dozens of Duds, and Found a $22 Million Question

John Goleby’s founder story begins with a pun expensive enough to become a business plan. In his final semester at Queensland University of Technology, he left before collecting the degree. Just before he did, he sold T-shirts that changed QUT into QUiT. The proceeds went into patenting a sugar-packaging solution. He turned down an early offer because, in his later telling, the thing was obviously going to take over the world. It did not.

This is a useful first chapter because it resists the usual polishing. The early invention did not contain the secret blueprint for Askable. It taught him that conviction can be admirable right up until reality sends the invoice. Goleby and university friend Andreas Zhou moved on and started Orange Digital in Brisbane in 2007. The agency designed websites, built campaigns and carried the small daily indignities of client service. It also put the founders close to problems that companies would pay to remove.

A product founder accidentally builds an agency

Five years in, Goleby noticed the joke. He had left university to create a product and built an agency instead. Orange Digital was working, but its work arrived attached to revisions, deadlines and somebody else’s roadmap. So the company made a decision few modest agencies can make casually: it rented a second office and gave staff one paid day each week to pursue their own products.

The incubator became a long apprenticeship in throwing things away. Over roughly five years, the team produced more than 30 projects, games, apps and side businesses. Most were brief visitors. The value was cumulative. Designers learned to spot recurring friction. Developers learned which prototypes survived contact with customers. The company learned that an idea could be cleverly made and still deserve a quiet burial.

In 2017, UX designer Vivien Chang brought the incubator a complaint rather than a revelation. Recruiting people for usability tests was painfully manual. Someone had to find suitable participants, screen them, negotiate calendars, remind them to appear and arrange payment. The research itself might be illuminating. Everything around it resembled administrative fog.

The least glamorous chore in research becomes the business

Chang’s frustration became Askable. Goleby, Zhou, Chang and John’s father, Scott Goleby, turned the agency’s experience into software that could recruit, schedule and pay research participants. Its early shorthand was the “Uber for user testing.” Crude as analogies go, it made the job instantly legible: request the right person and remove the machinery required to get them into the room.

The market answered quickly. Within the company’s early months, Askable had built a network of 5,000 participants and worked with Woolworths, CBS Interactive, Deloitte and Canstar. Revenue arrived from day one. The founders had not discovered a fashionable category by reading a trend report. They had spent ten years standing next to an ugly workflow and finally packaged the escape route.

The selling style was as important as the software. Goleby argued that a startup should start a conversation instead of trying to seal the deal. Keep it light. Keep it short. Do not pretend to be a massive operation. “Genuineness seems to go a lot further than highly polished communications,” he said. That approach helped a small Brisbane team speak to large companies without performing corporate dress-up.

It also shaped the ownership plan. The founders wanted to bootstrap for as long as they could and keep equity with the people doing the work. There was no moral theatre around venture capital, only a preference for earning leverage before negotiating for it. Askable remained self-funded and profitable for about six years while widening the job it performed.

The product expands around the original question

Participant recruitment was a wedge, but research teams had adjacent problems. They lacked capacity to run studies. Their software stack was fragmented. Interviews produced hours of material that still had to become a decision. Askable added certified researchers, managed delivery and a broader suite for conducting and analyzing studies. The simple marketplace started to look like research infrastructure.

Goleby’s argument was that adding another isolated tool would not clear the backlog. Teams needed participants, research expertise and software to operate as one system. His favored metaphor put researchers at the front of the bus, steering product decisions with evidence instead of being summoned near the end to validate a route already chosen. There is comedy in imagining a researcher shouting from the back while a product manager floors it toward a launch date. There is also a familiar corporate tragedy.

The company’s international growth relied on people asked to stretch beyond tidy résumés. Jonathan Peck remembered Goleby trying to recruit him as Askable’s only salesperson even though Peck was not a salesperson. He accepted. Five years and another coffee later, Goleby asked him to move around the world to open the North American office in Chicago. Trust came before the completed job description.

By late 2023, Askable had the sort of position founders often claim they will someday have: profitable, global and able to choose its capital. It completed a A$22 million Series A led by AirTree Ventures. Then it waited nearly a year to announce it. Goleby’s explanation was dry. The company had been busy building, serving its “ridiculously good customers,” and commencing its plans for world domination before sharing them.

The money funded product development, hiring and a larger US and UK presence. More revealing than the amount was the sequence. Askable had already formed its habits before the acceleration arrived. Capital could amplify a known operating rhythm rather than pay for a search party.

San Francisco, faster software and the stubborn need for truth

Goleby is now based in the San Francisco Bay Area, where Askable’s question has moved into more technical territory. AI can generate a prototype, a campaign and a convincing summary at unsettling speed. It cannot make a synthetic opinion become a lived human experience. The faster software becomes, the more expensive it can be to build the wrong thing quickly.

Askable’s current platform connects studies, a verified participant network, AI-moderated interviews and a repository of traceable findings. The point of the AI is not to invent a customer. It is to help researchers conduct more conversations and travel through the resulting evidence without losing the line back to who said what, where and when.

Askable Labs pushes the same logic toward frontier-model development. The division, with Goleby listed as founder, captures consented human work for training and evaluating AI systems. Its promise is provenance: named and paid participants, aligned recordings, auditable consent and data shaped for a partner’s pipeline. A company that began by finding five people for a usability test is now considering how human judgment enters models used by millions.

There is continuity beneath the change of scenery. The agency years taught Goleby to watch people struggle with a workflow. The incubator taught him to let weak solutions die. Askable learned to make human feedback operational. The lab is applying that machinery to a market newly worried about the quality and legitimacy of human signal.

Build a machine for noticing, not a shrine to the first idea

Goleby’s career does not offer a shortcut. It offers a structure. Work close enough to customers to see an annoyance repeat. Give a team room to test responses. Keep the cost of abandoning a weak response survivable. Speak to early buyers without pretending to possess a grandeur you have not earned. When a product finally pulls, widen it around the customer outcome rather than decorating it with unrelated features.

The QUiT shirt is funny because it turns a departure into merchandise. The sugar packet is funny because every founder has held at least one idea whose destiny seemed obvious only to its inventor. The dozens of discarded incubator projects matter because they made the eventual success less mystical. Askable was not delivered by a lightning bolt. It was recognized by people who had developed the patience to notice what kept hurting.

Now the setting is San Francisco, the nouns include models, agents and training data, and the product moves faster. The governing question remains almost old-fashioned: what do real people actually do, want and mean? For Goleby, the route from a Brisbane agency to an AI lab has been long, crooked and unusually consistent. Ideas come and go. The outcome keeps asking to be understood.