My First Product Failure - Pt 1

The magic of beginnings

It was a Friday afternoon in the very early fall of 1997. I was standing on the seventh floor in the office of Entra Capital Solutions - the company that had hired me less than a year ago right out of university. I could barely believe my luck and the proud happiness I felt. 

The beautiful office building was located in the center of downtown Stockholm, right at the top of the light yellow building left of the tower in the picture above. It had an absolutely stunning dockside view overlooking the Nybroviken bay with its fancy hotels, the beautiful archipelago boats, the national Dramaten theater, the lush Djurgården island and the posh boulevards connecting them.

As was the case every Friday at around 3pm or so the very real company beer tap was just being opened by enthusiastic colleagues and I was about to be handed a filled glass - one of the job perks that went well with the serious and smart, but light hearted, often pranky, tight nit and friendly attitude of my colleagues.

I walked out on the balcony to join some of my friends. This day the late summer sun was shining warm over a blue sky and the yellow rays were dancing next to the boats in the calm water below. With the smooth sensation of a light breeze into my face, I felt great.

Afterwork beer at work, at a top employer with some of the best clients in the nordic financial industry, with the best night pulse in the country only a five minute walk away? Reflecting on the contrast to my recent five years of university budget lifestyle, I could not imagine how my 26 year old life could be any better than this!

I had just had a good conversation with my boss, who had hired me less than a year ago. I had been doing well, helping customers by developing custom reports and spreadsheets using the existing capital management product suits. But there was a special reason for me to feel especially joyous on this day.

For a good few months now some of my colleagues had been working on a project to build a new portfolio management product. The decision had been made that I was going to join them to work on that product.

I was very excited. This was a serious effort to build the next generation product for the future, using better tech and using the experience from previous successful products. It would be the challenge I had hoped I could become part of.

However, on that shiny day on the balcony I did not know it yet. But the new product was going to become a failure.

The product journey begins

Our team looked good. Three other guys on the team were familiar faces. They were people I already knew from university, as they were also graduates from the same computer science degree as me. To go with that we had a very sharp seasoned database wizard who had worked with capital market clients for years. On the business side we had three very experienced capital market experts, with decades of joint experience from running fund, portfolio and treasury management operations, as well as operating market trading exchanges. On top of that, some of the best and most prestigious business degrees with deep contact networks as well as experience from operating previous products. To overlook our efforts we had one of two brothers who had been part of building one of the preceding product successes for a small product company that based on their success had been acquired.

The office was crowded and busy, and to get our own space the product team was moved to street level at the quiet end of the street. In fact, it was half a floor below street level which earned it the nickname “The Cave”. We loved it, it suited us perfectly. Quiet, somewhat secluded and well suited for focused work. And it was all ours so we made it our home.

Some new technology choices had been made already. There was a starting code base with some innovative architecture. We were going with the state of the art Microsoft technology at the time, which was COM-based high productivity Visual Basic 6 and the latest version of SQL Server for high performance data access, all easily integrated into the financial market’s most favorite analytic tools - Office. We knew that no product on the market at the time had tech that could rival the performance and user experience that was possible by using these tools. And with a team solidly trained in true object oriented programming, strong database experience and deep knowledge of the industry we felt we were well set up for success.

We worked hard for a little over two years. Gradually we built out most of the functionality that we believed a professional capital management client would need. Our in-house experts taught us engineers metric tons of knowledge we needed to understand the multitude of conventions for managing financial instruments and transactions, an area developed over literally hundreds of years. They told stories from the early years of capital markets in the Nordics and beyond, explaining not just how things worked but why things had developed like they did. I felt privileged to be working with people with such deep understanding, and with my engineering colleagues we learned a lot building solutions that followed the principles and practices taught by our more experienced financial veterans.

Beyond working hard, we also developed our own vibe and team culture with small habits, such as deliberately arriving late to the office in the morning to trigger the agreed upon “punishment” of then having to bring cinnamon buns for the whole team - a fun office practice that we took to the next level. Some late Friday evenings we set up our own LAN party to play Halflife deathmatch during after hours. I even remember downloading 0.5 and 0.6 versions the early versions of Counterstrike, which then was a community mod to Halflife. We set up our server, and the whole team had a beer and played together for an hour or two, before ending the day for the weekend.

Still, we were all business and we kept working hard. And eventually the time came to start putting versions together that we could show to customers.

We had version control, but no automation. A separate machine was later set up to create a release build which was done manually. Eventually, we created a "golden" CD with a pre-release version of the product that now had gotten a name - Osiris, named after the Egyptian god of fertility. We felt that was fitting for a portfolio product, and we had ideas to create a later separate version we intended to call Isis for treasury management, after the goddess of resurrection, life and death.

I still remember the cover of the product marketing folder, it looked very good, full of egyptian symbols and themes.

The cracks grow wider

A couple of customers had been contacted, to become early users of the product. We visited with one of them, and I remember us trying to install the product and getting it to work with their data.

We could not make it work well.

There were multiple issues with the installation, things we were not prepared well to manage and it took major work to get the product to run properly on a client machine. There were also problems with performance, especially when we got more real data into the product.

On top of that, the user experience had significant issues.

Because of our innovative design that featured a very high degree of flexibility and customization, basically allowing anyone to add and modify information types, people new to the product experienced a high threshold and found the product hard to use.

The intention was to make the product much more easy to customize for different customers, a major pain and shortcoming in previous successful products - something that generated a lot of consultancy hours but made it more expensive and required additional work for customers to get the solution to work well for them. We felt that our new design would solve all those issues, and to a significant extent we had succeeded in that. But these design choices had made both performance and the user experience a significant problem instead.

On top of that, the product seemed to be lacking compared to what most clients expected. We had set off to build a fully fledged solution for most capital manager needs. Sure, we had refrained from building some features needed by treasury and fund management. But we had not targeted a small enough segment of the market for our first release. So instead of a deeper product for a smaller segment, we had built a too thinly developed product for a segment too large.

In many ways, we had just become too home-blind to see the flaws that were now uncovered quickly.

It was very uncomfortable.

Dealing with failure

Looking back today is easy, but at the time I remember I did not fully understand what was happening. I was struggling to make sense of it.

We had a good team. We had worked very hard and we had done great work. We had completed all the work that people much more experienced than me believed we needed to succeed. How was it possible that we could fail this hard?

I remember feeling confused and helpless. Blindsided. And I did not feel anybody around me had any good answers on how or why we had failed. That was hard to process, and I don't think I did it very well. The fact that we had not talked much about the risk of failure did not help.

I felt betrayed, but I did not know by whom. It made me lose trust with the whole idea that people in business actually knew what they were doing.

And that included me. I realized I had a lot to learn.

The product ended up being canceled, and several of the people who had been involved left. Including me and a few other members of the product team, who instead started our own company. We were too inexperienced to do that, but we did not know that at the time so we did it anyway. But that is another story.

In my next part two of this article, I will break down some of the reasons for why this product effort failed, and some of the things I learned personally from this early experience.