I had an interesting debate this weekend with a friend of mine who’s also in tech which triggered this thought: will software in itself be worthless in the future?
The paradox is that software is finding its way into everything in our lives – from how we order food, a taxi, to how we run a business and increasingly recreation, gaming, education… virtually nothing is left untouched by software. And as hardware becomes more pervasive it literally will be distributed everywhere, from our attire to our kitchen appliances, the couch we sit on, the toys we play with and so on.
Yet the cost of writing software is also dropping exponentially, or seen conversely, the ease of writing software is increasing exponentially. The cost of building a simple website in 1995 was easily in the tens of thousands, would have taken months and would require deep technical knowledge. Today anyone can build a website for free on sites like Wix.com, or WordPress, or even from scratch with minimal coding. Or if you’re lazy you can hire someone for a few hours on PeoplePerHour.com and be done with it within hours.
Moore’s Law, one can argue, applies to software too. But whereas for hardware what powers that is advancement in microprocessor technology, in software it’s mainly the open source phenomenon.
Coders are willing to work on projects for free, for the greater good of the community. Equally the sheer number of coders is increasing exponentially as writing software becomes easier and easier and penetrates into our earlier education system. Kids now can write programs that were a challenge to software engineers in the 90s. That trend will only grow.
Microsoft which arguably was the first software company has been battling with this for decades as has its founder Bill Gates who’s always been very vocal about coders writing software for free. Yet the business model of selling pure software has been undermined largely by that trend. Operating systems, word processors and spreadsheets are now accessible for free (Linux, Google docs is just one example of each). Microsoft in the end had to adapt its business model – arguably the most powerful business model in the world to date – to survive and launched Microsoft 365.
That’s scary: even the most powerful business model in the world, that created the worlds richest man (or vice versa arguably:) was eventually undermined and had to adapt.
Equally, compilers that translate code to a functioning program have – and will carry on – becoming more sophisticated. Its quite imaginable that soon we will be able to input English language which gets translated into the end product by the magic black box in-between. The need to be able to code in zeros and ones (Assembly Language) has already become virtually extinct. The first coders had to write in zeros and ones; now there’s a myriad of ‘higher level languages’. The cap for that trend is simple written language – the first language we learn as kids. So if one can explain what is needed to be done by the program, the compiler will do the rest.
Seen differently, we’ve gone from a world where companies had to go to specialist to write their software – the Accentures & IBMs of this world – and spent hundreds of millions on bespoke systems that enables their business to function. Now, they can rent Software as a Service (SAAS) at a click and tailor it to their needs with very little coding needed.
Pretty much all core assets in a corporation can be managed by software that can be rented today as opposed to written bespoke. And some of it will have been coded by the open source community. The best example is Linux, and MySQL. You can manage your customers and sales pipeline this way with CRM, your procurement with ERP software, your computing in the cloud with AWS, Docker and so many more, your accounting, your people with Human Resource Capital Management (HMRS) software and so on.
SaaS or rental models have largely replaced the licensing model. But as software becomes more accessible that model will also die. Software itself will have to be offered for free in the future simply because some other open source competitor will do if you don’t.
Therefore new business models will have to be invented to sit on top of that software. It doesn’t mean that software itself will be extinct, but the business of selling software will be.
Businesses and people will still carry on being empowered by tools that operate with software in them. They will however pay for them in different ways to today’s models.
Personally I think this is a good thing. Apart from the fact that open source development a) engages so many more engineers and b) leads to much better innovation, it also aligns pricing with the value to the customer. This happens in every industry as it matures: pricing goes from a ‘cost plus’ model (because monopolies simply can do that) to demand-driven pricing: what is it worth to the customer?
Ultimately if one gets academic about this, valuing software should be no different to how one values any asset. In the end it’s the sum total of future cash flows the asset yields you (either in the form of cost savings via productivity gains or extra revenue) discounted to present value. It’s the DCF of the cash it unlocks to perpetuity.
Which is why I think the SaaS model will also die. SaaS is just a smarter way of charging for the software than licensing, but it’s still supply-side pricing. And as virtually every SaaS product will sooner or later have a free open source competitor, they will have to evolve or die.
Seen differently, in economic theory the two opposite ends of the competitive landscape are Monopolies at one end and Perfect Competition at the other. New industries often start as monopolies (the first airlines, utilities and the first software company- Microsoft), charge supply-driven cost-plus or cost-plus-plus-PLUS, reap profits for decades and then over time as they trend to the other end of the spectrum lose the power to do so. In a perfectly competitive market profits trend to zero. The sum total of all Airline’s profits in the U.S. for the last five years is something like 0.2% on sales. In other words, petty much zero. The current monopolies however like Facebook and Google (they are monopolies because last year alone these two alone accounted for virtually 100% of growth in Ad revenues in the U.S.) have profit margins north of 30%!
Software as a product in itself is trending at – or very close – to the perfect competition end of the spectrum. Which is why its price will trend to zero. What’s unique about software though is that its not just a product itself, it’s also a platform to build on. Amazon, Google, eBay, Facebook, and much smaller players like my company PeoplePerHour are all platforms. The value in a platform is its network and that’s its defensibility. Network effects means that every user that joins the platform makes it more valuable for the next user. And this is without writing a single additional line of code. Yet the business sitting on top of the software becomes more valuable.
This holds true of marketplaces but hasn’t yet been applied to B2B models where a company buys or rents software to use to run its business. But it will have to because the ecosystem that builds that software is the same and is becoming perfectly competitive.
We don’t just need innovation in product and software design, we need innovation in pricing.
My hunch is that the answer will come out of data science. Data is very quickly becoming the most valuable commodity in the world. And the tools to analyze it are becoming exponentially more advanced. We are already making big strides but much like in the C2C and B2C worlds the tracking and measurement on every click’s ROI has become an exact science, in the B2B world we will in the future be able to measure the value of software to an Enterprise to the dot. We will be able to measure to the dollar the exact amount of contribution it’s making to your cash flow, and from then on it’s a simple DCF calculation.
The models that will prevail are the ones that a) help you calculate that to the dot, and b) charge a fair share of that as fees.
While i am writing this we are having the same debates internally at PeoplePerHour over the launch of our new product, TalentDesk.io (watch this space;)… more on that in a future post but for sure we will be innovating on both points above – the product itself but also its pricing. If Marc Benioff coined the term ‘the death of software’ with the launch of SalesForce.com, we are declaring the death of SaaS with TalentDesk.io.