I often say that it is not so difficult to predict the future, but it is very difficult to predict the timing. Self-driving cars are a good example. We can be pretty sure they will come. But does it take two, five, ten, or twenty years?
This question is important because investment decisions depend very much on the timing, and wrong timing can kill many companies. I have also looked at my personal track record, and sometimes the timing has been good, but often it has not.
Learnings from my early career in tech
I have spent many years in the tech business. I remember many cases where we got the timing wrong. And in some cases, even when we were right about the timing, we nonetheless failed to succeed. Fortunately, there have been a few cases where the timing went right, and we were ready for the market.
In the 1990s, one of my first roles was related to special-use radio networks, like public safety and security. We were developing a new-generation digital radio network. I remember one of the senior engineers started saying that circuit switching was not a relevant solution anymore and voice would shift to packet switching networks. Many people didn’t take him very seriously. Of course, it turned out that he was right, but it started to happen approximately fifteen years later than he envisioned.
In the late 1990s, our unit had to predict ADSL market growth for a carrier. We were over-optimistic: we expected the mainstream move from modems to broadband to happen almost five years earlier than it happened.
Then, at a network equipment vendor, I predicted the fixed-line telephone-switch business would decline soon and that it would be better to shift focus to broadband. The management of the unit didn’t like my comments. A year later, the unit was closed when the top management of the company came to the same conclusion as me.
These examples show that changes happen in different parts of the value chain at different paces. Certain technology might become obsolete, but still, millions of people use services built on it, which affects the timing of the transition to the latest technology.
When we got it right in the data and fintech sectors
In 2004, we started to work on a startup that was focused on social network analytics. We named our first product for it “Social Links.” When I was on the way to an event to launch it, one of our investors called me. He asked if it wasn’t disadvantageous to use the term ‘social’ in the product name, especially when Americans associate it with something related to Communism.
Nonetheless, we launched the product with that name. In 2007, no one questioned the name anymore when Facebook and other social media companies entered the mainstream.
In the fintech sector, it was harder to predict timing. Normally, the timing depends on the technology’s quality and how ready users are to adapt to it (which also depends on the user experience). However, in fintech, it also very much depends on financial regulations.
When we started the first crowdfunding and peer-to-peer (‘P2P’) finance services, it was illegal to offer them in many countries. Fortunately, the regulations in many countries rapidly caught up with that new trend change. So we can say that five years is a short time in financial services.
We were also one of the first to offer cloud-based back-office solutions for financial services. In 2016, the chief digital officer of a bank told me he didn’t talk with any company that offers financial solutions in the cloud and refused to talk with me further. Two years later, he was no longer in that position, and several banks had started using cloud-based services.
The things we got wrong
But at the same time, we were wrong about the big changes in the finance sector. We thought that fintech companies could challenge banks in many service areas in just a few years. This has in fact happened with some niche services, and fintech has been adopted by banks too, but we are still waiting for the ‘big’ changes we expected.
Ironically, we worked with the first concept for decentralized data analytics and user-held data in 2005,and even filed a patent application. Fortunately, we realized it was very early days for what we were proposing, and didn’t invest too much in it then. Now, it looks like these things have started to emerge – that said, it is still hard to tell the exact timing and services that trigger the big changes in the consumer data market.
Wrong timing kills businesses
Timing is a very important question in business. Typically, if a startup is too early or too late, it cannot survive. Large corporations have the same dilemma, but from a slightly different angle. Typically, large corporations are hesitant to cannibalize their old business too early. But if they introduce adjustments too late, companies with new technologies and solutions are likely to take over the market.
Nokia’s failure with mobile phones is a perfect example of being too slow and too fast at the same time. Nokia was too slow to react to market changes when Apple launched the iPhone, but with its famous burning platform memo, it also killed its existing Symbian-based business too early.
Cryptos and quantum computing are two hot trends in the start-up world where the timing is uncertain – especially when they get mixed together.
Some time ago, I was sent a startup business plan for building a system to make investment decisions with AI in quantum computers for the crypto market. The founder wanted to convince me that this would be the future. I tried to tell him that it might be the future, but that was beside the point. Their business plan stated that they would get the first products into the market in 2023, start to make revenue and that the business would go up rapidly. So, if they assume to make revenue in the next year, their company wouldn’t exist when “the future” arrived.
Still, it is very important to realize that if you invest in these kinds of technologies, they can still be very long-term investments. It is also especially important that startups don’t burn money too early. One company cannot change the market alone. So, if you really want to make an AI-based quantum computing crypto hedge fund technology, you should not burn money on marketing yet – unless your actual objective is to collect investor money with hype words, take the money and run. (But that’s not what I’m talking about or even interested in.)
Changing the business
Another complex area is how certain new technologies have the potential to change the business. Looking at quantum computers, for example, we will likely see quantum computers in the future, and they will play an important role in accomplishing many tasks. But what are the areas where they could come in to change the business? Can quantum computing ever replace traditional computers?
Or when we look at cryptos, amid the buzz of Bitcoin and other currencies, many people miss that the real game-changer is the decentralization behind it. As I wrote earlier, we should focus more on decentralization than cryptocurrencies, because decentralization is a big change for the technology and the business, while cryptocurrencies are a more speculative business model.
Will decentralized technology be based on crypto or Web3 business models, or something else? It is hard to predict. Maybe one day – but again, when?
Five predictions for next five years
Predicting the future of technology is not easy – and even if you’re right, the timing is even harder to predict. But that doesn’t stop people from trying – and it doesn’t stop people from asking those who try for a living.
For myself, if I were asked to predict five areas in which the biggest changes will happen within the next five years, I would list the following:
- Health and wellness data will start to change the way health care and wellbeing services are provided to the patients
- Decentralized models will allow storage and analysis of data locally and for individual users
- Online and offline retail and related data will become more integrated and lead to the creation of new retail concepts
- We will see a continuous transition from dedicated hardware to off-the-shelf hardware, with software and data to implement actual use cases
- Better big data analytics will enable combining data from many different sources, including weak signals, to make better and more objective predictions for different purposes.
Some of those predictions may be wrong, and some may turn out to be right. But in business, you cannot just wait – you must act. But you must also remember to be aware of the possible need to adjust your course at any time as the situation changes.