We were doing around 3,000 jobs a week, which meant roughly 10,000 people sitting in the back of our cars, so the idea seemed obvious: give local businesses the chance to advertise directly to them. And commercially, it worked. Businesses wanted the space, some were bidding against each other and a few even wanted exclusivity within their industry. For a little while, it genuinely felt like printing money.
The problem was that I had tested whether people wanted to buy it, but I had not tested whether the business could actually carry it. Worse, I had already sold a year's worth of advertising before we had properly tried it in the real world.
Then reality arrived. Drivers would not always turn the screens on, customers did not always want adverts playing in the back of a taxi and the technology was nowhere near as reliable as it would be now. Advertisers were understandably unhappy if they had paid for something that was not being shown, and the drivers, never exactly a selfless bunch, quite reasonably wanted to know what they were getting out of it.
The final straw was customers leaving bad reviews because they did not want to be advertised to while they were already paying us for a taxi. That mattered more than anything else because the advertisers were customers of the advertising idea, whereas the passengers were customers of the actual business. I could refund an unhappy advertiser. Lose a taxi customer and I might never see them again, and for every one angry enough to complain I assumed there were plenty more who disliked it and simply said nothing.
So I started trying to fix it. We limited the screens, tried to make sure drivers turned them on, gave passengers the option to switch them off and even put notices in the taxis explaining what was going on. All of those things were possible, but none of them made the idea worth the effort.
That is something I learnt a few times over the years: just because something can be fixed does not mean it deserves fixing.
If I wanted to mark my own homework, I could probably prove the advertising made money. There was revenue, there were advertisers and it opened doors with local businesses that I might not otherwise have spoken to. From a commercial point of view, that was not completely worthless. Having a reason to speak to another business owner without turning up with the obvious sales pitch can be useful, and in some ways it became a bit of a Trojan horse for the taxi business.
So I would not call the whole thing a disaster, but I would still mark it down as a loss because a balance sheet does not show everything. It does not show the time spent managing something awkward, the effort required to keep staff doing something they do not really believe in, the customers quietly getting irritated or the damage caused when a side project starts interfering with the thing people actually came to you for.
That is where the economics changed. The advertising made some money, but the business paid for it in other ways.
If I was doing it again, I would have trialled it properly first. A couple of local businesses, one or two taxis, free advertising, no promises and definitely no year-long commitments. Then I would have waited to see what actually happened. What did the drivers think? How did customers react? Did the technology work? Were advertisers genuinely getting what they had been promised? And, perhaps most importantly, how much effort did the whole thing take to keep running?
Only then would I have decided whether it deserved scaling.
That is the bit I got wrong. I proved demand before I proved delivery, and those are not the same thing.
A new revenue stream can look like growth. Sometimes it is. Sometimes it is just a distraction with a sales line attached to it.
So now, if someone tells me they have found another way to make money alongside their main business, the first thing I want to know is not how much it could make. It is what impact it is going to have on the core business, and how distracted they are going to become from the thing that actually pays the bills.