The story behind it

For thirty years my consulting business has taken between seventy and a hundred hours a week. Everything else came on top of that: photography, agency work, rentals, a string of online shops — around twenty of them over the years. And the consulting side had visibility needs of its own, because in the financial market you have to keep up your contacts if you want to be asked.

All of it needed marketing. None of it got any.

And it was not for lack of knowing. I could have sat down with anyone and pointed out what they should publish that week, and why. I just did not have a single hour to do it myself. So it was left undone, and the bill came the way it always does: in revenue that never arrived.

I searched hard, and found the wrong things

I spent years looking for something to buy. The problem was rarely that the products were bad — it was that each of them covered one slice of a business, while I had several businesses that looked nothing alike. What I needed always fell in the gap between two products.

In the end it dawned on me that the question itself was wrong. I was asking for one package that would cover photography, rentals, agency work and publishing all at once. Nobody builds that, and there are good reasons why. The answer had to be components that exchanged data with each other.

That realisation should have come much earlier, because it is just about the only thing I already knew how to do.

Where that confidence came from

From 1980 onwards I worked on financial systems, and back then I wrote the code myself — in languages that now belong in a museum. I was a test manager, a project manager and a database administrator. Through the nineties I moved from building individual systems to designing how systems should fit together, and after that came years as a senior executive, including in finance and as head of the consulting division of an American IT group.

One assignment says the most. For a large Danish financial group I designed an integration solution that let the machinery they already owned work as a team, rather than replacing it with something all-encompassing. The price tag was a fraction, and their annual running costs fell by around forty per cent.

So I knew perfectly well how parts should be put together. I just had not written one of them myself since the nineties.

What broke the deadlock

I do not write modern code, and I am not going to learn it now. But being able to state a task sharply enough — the data it works from, the result, and how it should behave the day something breaks — turned out to be worth more than syntax. I described the problem, tested the answer and corrected course. That is the role I had held for three decades, only with someone at the other end who replied immediately.

The very first version was spreadsheets and keyboard macros on an ordinary Mac. It did the job for me, and at the same time it was completely unsellable: a machine sitting at home clicking its way through the day cannot carry customers. So the sheets became data structures, the macros became code, and publishing became something a server could handle on its own.

Three attempts

The first version proved it could be done. When I showed it, I got a response I have benefited from ever since: people liked the idea, but did not understand what I was solving for them. The second one turned that around and was not about what the system could do, but about what you no longer have to do once you have it. That one landed.

And then I did the thing that cost the most and paid off best: I put my own solution under the microscope against the privacy rules, the marketing act and Meta’s terms, before anyone else got the chance. The conclusion was uncomfortably clear. My original idea — a system that read groups and replied automatically to people’s life events — was not permitted, and never would be.

So the hardest part of the project was not technical. It was killing off my own best idea while it still worked, without anyone asking me to. The version that came out of that is the one running today.

When it stopped being my problem

I had no plans to sell this to anyone. In time I mentioned it to people I talk to anyway — colleagues, acquaintances in other industries, and a couple of competitors. They all said the same thing: they knew what they ought to be doing, but had no capacity to do it alongside running the business itself.

When a competitor says that out loud, it is worth noticing. People dress up most things, but nobody dresses up the fact that their marketing has stood still for two years.

What it is today

GoRabbit is the engine that runs the small business: eighteen modules around one core, covering marketing, sales, online shop, customer dialogue, accounting and control. Information is entered once and used everywhere. And one rule stands above all the others, inherited from that round with the regulations: the system proposes, the human confirms.

A good deal of it is reuse. The media monitoring comes from a tool I built to follow the Spanish press, and from the report I used for my own investments over thirty years. The control module is the oldest of them all — I have been delivering setups like that to clients since the nineties. What lies here is not two years of work. It is a working life, sorted.

My first customer was myself. The first one who was not me was a small publishing house that puts out poetry in dialect — about as far from a server room as you can get, and precisely for that reason a good test. Because now the whole chain had to stand without me in the middle: sign-up with consent that could be documented, a confirmation that actually arrived and did not end up in the junk folder, a click that verified it, the delivery itself, and finally a place on the mailing list that could withstand being audited. I sat and followed the logs while it happened, and that is a different feeling from testing it yourself.

And one thing about the price

You subscribe to GoRabbit monthly, with no lock-in. That is the main way in, and it is that way because a product that needs a binding contract to keep its customers is telling you something about the product.

For anyone who cannot put money on the table before she knows that it works, there are also a few ways in where we carry the risk instead — an online shop in return for a share of sales is the most used. I know very well what it feels like to pay eighty thousand for a shop before selling a single thing.