Why Jem splits its work into two lanes, and how a week replaces a six-month bet.
Most companies test a new idea by building it. By the time they find out nobody wants it, months of work are gone.
We asked Jem co-founder and CEO Simon Ellis, whose WhatsApp-based HR platform serves more than 250,000 frontline workers, how Jem decides which new ideas to pursue…
The move: run new ideas in a separate lane
Jem runs two lanes. The growth lane is the existing business, built over six years. The discovery lane is where new ideas get tested, quickly and cheaply, before anyone commits real time to them. The aim of the discovery lane is speed, in either direction.
“Shortening the time from idea to killed, or idea to moving forward.”
How to test a new product idea in a week
1. Keep the core business in its own lane
Don’t let new ideas pull people off the business that pays the bills. At Jem, the growth lane keeps running and improving the core product, while a separate group works on what might come next. That way testing new ideas doesn’t slow down the thing that already works.
2. List every idea competing for your time
Write down everything you’re tempted to build. When Simon described the discovery lane, Jem had six things it wanted to build. Seeing them side by side makes it obvious you can’t afford to spend six months on each.
3. Give each idea one week, not six months
The question for each idea is simple: how do we take a week to find out whether this is a good idea, instead of betting six months and having to kill it? A week is enough to talk to customers, test demand and see whether anyone would pay. It isn’t enough to fall in love with a half-built product.
4. Use someone else’s framework to test it
Simon doesn’t invent the test from scratch. One method Jem uses is based on Y Combinator’s office hours, where partners pressure-test founders’ ideas. Y Combinator president Garry Tan has released an open-source version that asks the questions a YC partner would: is there real demand, what do people do today instead, who needs this most, what is the smallest version worth building, and have you watched someone use it.
5. Make the call at the end of the week
When the week is up, decide: kill it or move it forward. Killing an idea after a week is a win, not a failure, because it frees the team for the next one. The ideas that survive move towards the growth lane with real evidence behind them.
The big payoff
Instead of one big bet every six months, you test six ideas in six weeks and keep only the ones with real demand. When AI makes building cheap, choosing well becomes the advantage.
You might also be interested in
Want the full story?
Simon’s full panel from The Open Letter Cape Town 2026 is available to members inside the Founder Collab, where he goes further than we could cover here:
How Jem’s engineering agent now resolves 92% of maintenance tickets on its own
Why Jem nearly killed its AI experiments too early
How to get Claude to spot your repetitive work every week
How to share context across offices in three cities
Why he wants a thousand AI-native people, not ten
You’ll also get access to 40+ masterclasses from SA founders and operators on sales, fundraising, UX, paid media and more inside The Founder Collab.
Get more SA tech and business news, tips and business-building workflows and subscribe to The Open Letter.
