Where TradeYoke began: an idea from helping run window cleaning rounds

Part 1 of 39 of Building TradeYoke, the story of how the platform was built, from January 2024 to today. Next: Five years of the same idea, and why we finally started building. See every part on the build timeline.

Quick answer: TradeYoke began as an idea around 2019, while we were helping to manage window cleaning rounds and saw the same problems on every one: the round living on paper or a spreadsheet, cleans being missed, and payments needing to be chased. Development started at the beginning of 2024, and this series follows it month by month.

• The idea came from helping manage window cleaning rounds, not from a boardroom: the problems were visible on every round.
• Window cleaning was the first trade because a round is a clear, repeatable shape: customers, frequencies, a route and a price.
• We left the idea alone for about five years and began building at the start of 2024.
• This series, Building TradeYoke, tells the build in order from January 2024 to today, with the reasons behind each step.

Around 2019 we were helping to manage window cleaning rounds. Not as observers: we were close enough to the work to see how a round actually runs day to day, and what happens when it does not run well. A window cleaning round is simple to describe. You have a list of customers, each with an address, a price and a rhythm (every four weeks, every eight), and you work through them in an order that makes sense for the van. The simplicity is deceptive. Keeping that list true, week after week, is where the trouble starts.

The same few problems turned up on every round we saw.

• The round lived on paper or in a spreadsheet. Only one person really knew where it was up to, and it could not be in two places at once.
• Cleans were missed. Not through carelessness: a wet week, a locked gate, a new customer written on the back of an envelope. Nothing told you a house had been skipped until the customer asked.
• Money needed chasing. Who has paid, who has not and for which cleans was a job in itself, and it was usually done late in the evening.
• Rounds drifted. Prices stayed at what they were three years ago, a customer moved away but stayed on the list, and the order of the day stopped matching the map.

None of this is anyone's fault. It is what happens when a business that does hundreds of small, repeated jobs runs on tools built for something else.

A round has a clear shape, which makes it a good thing to build software around. Every customer has a place, a price and a frequency. Every visit has a date, and either it happened or it did not. Money follows the visit. If software can look after that well, with the next visit worked out for you, the route in a sensible order and the payment tied to the clean it was for, then a lot of the evening admin simply goes away. That is what we wanted to build, and window cleaning was the obvious first trade.

For a long time it stayed an idea. We kept coming back to it, because the problems did not go away. At the start of 2024 we decided to stop talking about it and begin. The next article in this series, Five years of the same idea, and why we finally started building, explains that decision.

Building TradeYoke is an honest, month-by-month account of how the platform was put together: window cleaning jobs and customers first, then invoicing, quoting, surveys and inspections, maps and directions, and live monitoring, and then everything since. We will say what we built, why, and what it does today. Where something is still in pilot, we will say so.

Related: Window cleaning software · Scheduling and rounds · Running the round · How to plan a window cleaning round · Pricing · The build timeline