It is said that software consumed hardware and now data consumes software. Although software is core to most businesses today, it is not easy to create a viable software business, or at least generate big profits by only selling software. And to make matters worse, there is currently a global shortage of software people. What should we conclude from all this? Can companies make money from software?

We see many trends in the software business:

1. software has gone to the cloud and is sold as Software-as-a-Service (SaaS);
2. data analytics (including Machine Learning and AI) have become an important part of needed software;
3. many software components have become commodities or open source;
4. more and more effective software development tools are now available;
5. proprietary software offers competitive advantages, but software companies need to offer more scalable products.
   

​When we look at these trends, it is plain to see that it is not easy to find a ready model to become a successful software company. One popular option is to sell top level software development capability, but that is not a very scalable model. Software companies must become more innovative, and we could even claim that business model competence has become more important in the software business than software competence itself – presuming you want to create something big.

Microsoft used to be representative of traditional software business. It provided software licenses mainly for PCs. Hardware vendors paid forand supplied the MS-DOS/Windows operating systems but customers paid for Office, that has now become a cloud-based SaaS. Apple is a leading hardware and software company, but after people pay for their iPhones they use services like the App Store, for their apps, music and content. Google has until now offered its Android operating system for free, although it now plans to introduce license fees, yet its main target has been to gather user data, offer ads and provide content and apps through its Google Play platform.

Then we have traditional B2B software companies like SAP, Oracle and Salesforce. Oracle has always had the most traditional license model. SAP also has license fees, but it is more a platform to develop actual applications. Salesforce, of course, is the famous SaaS pioneer, that now includes a lot of applications from third parties.

Actually, all these cases give an indication on how software businesses are developing. In each case you need to offer a very scalable system; effective channels to get it to users; but also the opportunity to build niche and tailored services and apps on top of it. This is the platform economy model to offer software and it is important to get enough services, service providers and users to the platform. However, the quality of the services can be more important than quantity.

It is hard to provide simple success factors for any software business (and even if I knew them, I probably wouldn’t publish them), but we can list components that seem to help:

1. Offer very scalable platform to use your software from cloud;
2. Open your platform to third party applications and open source components that can offer targeted apps for users;
3. Take care of the user experience and quality of the service and apps;
4. Offer tools and engines that enable customers and developers make more tailored and proprietary apps based on your system;
5. Engage and motivate developers to use your system.

Of course, many parties try to follow this route now. Quality and usability are probably the key success factor to be successful in this game. It is, for example, easy to make compromises on the quality and just focus on attracting lots of apps or developers, but this can be harmful in a longer run.

Is data eating software? It is probably a topic for its own article, and it is not so simple. I think the comment is used to mean that it is more important to own and utilize data than software. It is clear that for companies like Google and Facebook it is like this, although they have good software to use data. Maybe Amazon belongs with these companies, too.

But the latest issues with data misuse, breaches, and privacy make this more complex. Data will be important, but who owns it and who has the right to use it is becoming more complex. Maybe people will opt to get their own data back and be willing to use good software solutions to utilize it, like personal AI services.

Business models are funny things. When a strategy consultant formulates a winning concept, innovative entrepreneurs work on the next big thing that will beat the old model. If you are a good software developer, you can make good money by coding, but if you want to create something much bigger, then you must innovate and create a new business model as well.
The software business is like the music, movie or game business – you could make it big with some luck, but to build systematic success you need a platform and distribution channels to offer so much more than just luck. It is then more about orchestrating than coding.

The article first appeared on Disruptive.Asia.

We have probably passed the peak of blockchain and cryptocurrency hype. It doesn’t, however, mean that nothing is happening with the technology and market at the moment. But it does raise the question, are we now approaching real blockchain business and more stable crypto solutions? And what is the status of public and private solutions? Can we really count on the security of blockchain-based systems? Can we have a generic blockchain or is it more to have application specific solutions?

We are now hearing fewer fantasies about blockchain and crypto finance than a couple of years ago. This is a positive sign after the first peak of hype. At the same time, we can say that blockchain development hasn’t stopped. There are a lot of developments with the technology and a large number of pilot projects. 

Blockchain and crypto finance pilots include projects for supply chain, settlement, clearing, cryptocurrencies and voting models. There are also several projects including the tracking of coffee beans; smart contracts for beverage logistics; money transfers and international telco settlement. However, most of the pilots are quite limited, i.e. include a few parties that are committed to the pilot.

The real target of blockchain is to be a multi-party system without an authority. Some blockchain activists also see that real blockchain must be always an open system, but some others see that private closed blockchains can still be relevant and valuable for many needs. But whichever approach we take, it is very important to get from low number party pilots to systems that include multiple parties.

The next important target of blockchain pilots and PoCs (Proof-of-Concepts) is to get different companies and parties to work together with one system. It is technically an important step and really tests the capacity and technical compatibility of any solution. Just as important, if not more important, is that the blockchain and smart contract solutions should offer a non-authority, trustless system.

There is not enough proof of a pilot’s viability if it only includes a few parties that know and trust each other. For example, JPMorgan has talked about its own cryptocurrency, but many parties see it purely as a digital currency, because it is a private system that only works between a lender and its clients.

Many claim that it is hard, or even impossible, to have generic blockchain and smart contract solutions that can be used for all kinds of applications and contracts. It is felt that we need tailored solutions for different applications and business needs. Can we really expect blockchain solutions to take off with these limitations? 

Part of the complexity is that we have to create contract models with software, including special terms and conditions of the contracts in the software code. If we compare this to development of hardware and software solutions, one could claim that big breakthroughs happen only when it is possible to create layers for solutions, and properly isolate different layers and keep them generic. A clear layer model would also help with the security issues. If it is possible to use tested and better-known lower layers for contracts, then only the details and terms of the contracts would be case specific.

There has also been quite a lot of discussion about the security of blockchain solutions. Especially for voting and identity systems where one realistic threat is plutocracy. Parties that have enough money and resources can basically buy votes. There seems to be several ways to do this depending on the implementation of the voting system, but it still appears to be hard to prevent this with the existing solutions. This is definitely one area, where new solutions are needed.

A more acute threat now comes from 51% attacks – where a party can control enough (over half) of the computing capacity to mine a crypto coin and then basically create its own blocks. Only a few months ago this was seen mainly as a theoretical threat. A theoretical example touted was that some parties in China could coordinate such an attack in extreme situations. But now we have seen the first 51% attack on Ethereum Classic so it is no longer a theoretical threat!

This has led to the development of more advanced versions of the proof-of-work algorithms. This is taking solutions more towards proof-of-stake algorithms. It is quite typical in all systems, including critical financial systems, to have vulnerabilities and updates for the software are constantly needed. But with blockchain this is more complex, when it is impossible or hard to undo transactions. The solution, for example to return stolen money, is to rewrite history and go back to the point before the attack occurred and recreate all ‘legal’ transactions after that so that all parties agree on those transactions and create new blocks. It is also a complex legal and ethical question, if the community has the rights to rewrite history.

A strength and weakness of blockchain and smart contract solutions is that different parties can see and investigate the software core. It means different parties can also find potential bugs and vulnerabilities. If a party finds a problem, then it is up to this party, to determine how it wants to use the information, i.e. to either hack or improve the system.

The blockchain hype also included theories that blockchain had no security issues and could solve all the issues of traditional financial systems. Of course, this is a naïve belief. All systems have their issues. The positive side is that many developers and parties are already active in the blockchain community constantly developing improved solutions. However, as with any system, it is a continuous race between threats, problems and better solutions. But blockchain is different from traditional centralized systems, and the risks and new solutions are different, and different competence is needed to handle them.

Banks, stock exchanges, telcos, logistics companies and many startups are already working with blockchain solutions. Nothing, it seems, can stop this development path. As yet, we don’t know which blockchain solutions will make a real breakthrough and rule in the future. Maybe those solutions will be unlike existing blockchain solutions, or other distributed ledgers because computing and data are now distributed in many different ways. They must first solve the issues of existing solutions, especially getting from private pilots to multi-party transaction models. 

The article first appeared on Disruptive.Asia.