Today, software isn’t just eating the world, it’s killing people. Algorithms mean companies haven’t just moved fast and broken things, they’ve moved recklessly and broken society. We have to accept with new technology comes great responsibility and software has never been more vital.
Good software feels like magic. When a Tesla becomes faster or drives further or lets you keep dogs inside thanks to an over-the-air update, it’s wondrous. When you can move to an earlier flight at the touch of button and for free, it’s magnificent. But the better some things get the more shocked we become that so many other experiences suck. The more we rely on software, the more dangerous things become when we place lives in the hands of bad code. A crashed screen on the BT Tower looks clumsy, but the blue screen of death in a self driving car could actually be deadly.
Our shocking digital foundations
If the IT systems for most companies were physical constructions, we’d never dare to venture even close to these buildings. Most systems and software are an amalgamation and aggregation of patches and quick fixes, hacks built on temporary workarounds augmented by kludges. Things that work on a good day, just about, but nobody quite knows why.
Because management can’t fix the structure or foundations quickly, efficiently, or with certainty, we slowly add middleware to legacy infrastructure. We get servers to talk to other servers through a proxy, we build for common use cases and let only the weird ones fail, and on top of this we build fancy new user interfaces to cover it all up.
It’s always fascinating to me that the most vital IT equipment in the world is the oldest. The US Department of Defense (with an $800bn annual budget) uses IBM Series/1 computers from the 1970s and floppy disks to control its nuclear weaponry. Nuclear power plants and public transport are routinely based on both hardware and software from over 40 years ago.
The cracks appear
You rarely see into these systems’ problems but you feel it.
Online travel firm Kayak can scan more than 600 airlines and millions of permutations in seconds but an agent is still required to type furiously into a keyboard for five mins (and then often make a phone call) to change your ticket. Airline apps offer the chance to make changes, but are typically not able to make them “for exceptional reasons”. Departure boards at airports often show different delay information to airline apps, which have different information to flight tracking sites. It’s all a mess because an entire $750bn industry is built on COBOL, a computer language from 1959 that only works on a good day, just. The banking industry is similarly precarious. In the US, around 80% of in-person transactions and 95% of ATM swipes are based on the same software from 60 years ago.
Time to prioritise
IT in the minds of CFOs, CEOs and most of all CMOs has never been considered vital. Airline execs assume its the champagne served on board that matters, not the call centre slickness. Hotels assume they are in the business of pillow fluffing, not the “we will actually call you when your room is ready” industry. Car companies make faster cars, not better ways to book service appointments. They are all wrong. Today what matters is every touchpoint.
Yet for decades IT used to be a service department, it was like the cleaners or receptionist – an overhead, something peripheral – but now it’s core. It is a profit centre to maximise investment in, not a cost centre to offshore.
Companies’ P&L statements don’t create arguments for projects that take more than 10 years to pay off. They don’t reflect savings that are currently not made. They don’t allow for increased revenue that could happen if a fundamental change is made. Instead, organisations bury their heads until something like the Millennium bug or the Year 2038 problem arise and then they do the absolute minimum they need to.
The way to focus
The most modern technologies get the best teams and the greatest budgets. Trying to argue for desktop website improvements is an impossible task, but if you have no app and the competition does, suddenly the money appears.
In business there is what I call the ‘CEO in airport effect’. CEOs rarely have time to read, but they do notice glossy outdoor ads from Deloitte about the internet of things, or from IBM about Watson and AI. They read in-flight magazines about the next big thing being blockchain. And these spark a thousand emails a day to their teams saying, “Hey, what is this, and what are we doing about it?”
I personally think blockchain is an inelegant solution to an unknown problem. The notion seems confused by marketing fluff, it seems way too complex and often deeply naive in its misunderstood ability to solve problems that are typically people/computer interface issues.
But what I love about blockchain and the trojan horse we must use it as, is a way to get CEOs to care: to find money and to build systems that customers deserve. The same way that the hyperloop can make public transport topics for mayors seem sexy, or that Amazon Go gets retailers thinking about better self-checkouts, or that Bitcoin gets payment companies to wake up to digital wallets, blockchain can be a fairly silly but sensible excuse for further thought.
My absolute favourite and most awkward question in business is, “If you were setting up your company today what would it look like?” Blockchain can get us to rethink what we’d make if we could change things at a deeper level. What can we finally make possible in a world where things just work?