Wednesday, January 24, 2018

You may not be a software company, but that isn't an excuse to lame-out at computering

If pizza delivery, coffee joints and banks can do it, so can you


I don't begrudge organisations who want us to start calling them "software companies". People are free to do whatever they like with such trivial labels, I guess. But the tick of such labelling has always been an annoyance to me.

No, you're a company that uses software effectively
Most companies saying "actually we're a software company" are anything but. They very rarely sell software as their core business. Of course, I'd never shy from bombastic overstatement (or too much redundancy). These companies are trying to make a valuable point: they're now using custom-written software to do more than digitise paper-driven, manual processes and customise their ERP systems into cement. They're now able to program their business.

Everyone's favourite pizza provides a hot, steaming example. While Domino's boasts that you can order pizzas from your wrist and Papa John's makes it lickity-split easy to customise your pizza on an app (for some reason, you can't add anchovies except by phone – file a ticket!), these two companies are still fundamentally, well, pizza companies.


Starbucks has long been an example of a company comfortably creeping up the "digital transformation" curve. Their software-driven orders have been so successful that mobile orders have been known to clog their meat-space. Still, when I go there (hey, get off my back, tapered sweatpant milinums! I just want some coffee!), I'm happy to find coffee in my cup instead of a numbered stack of those mini CDs begging me to click on "startup.exe".

Do you even computer?
In an era where Amazon and its three friends are trundling through everything, it's easy to get wrapped up in the need to transform to a software company. That said, would you even call Amazon a software company? Clearly, in their cloud business they are, but the retail business is more about ruthlessly creating and using software to, well, sell stuff. You can throw a "multi-sided platform" flashbang into the mix to befuddle this point, but at the end of the day, Amazon's nickname, "the everything store", tells you exactly what the company is.

Finance has for a long time used software of all sorts – custom and off-the-shelf – effectively and it's little wonder that they're one of the few industries to have quickly staved off Fear of Silicon Valley Eating Your Lunch. While incumbent banks have been slow to adopt mobile payments, they're now spinning heads at how quickly they're catching up. Banks have a good track record of acquiring pesky finance startups and they've been stuffing themselves to the gills with nerds. Recently, Goldman said that a quarter of their employees are engineers, supporting over 1.5 billion lines of code. JP Morgan Chase has somewhere in the region of 19,000 developers.

Banks have always understood IT well enough to gorge themselves on it; many an IT vendor salesperson has filled their wrist with heavy watches and pegged out their retirement accounts by going up and down Wall Street. Sure, I'll concede that you can get all intellectually crafty and point out that money is, largely, just numbers in a spreadsheet somewhere, but it'd be odd to call these banks "software companies".


Creating software is the art of failure...
The problem with calling yourself a software company – beyond the obvious fact that you're not selling software – is that you must now think and act like a software company. Software companies, especially young ones that are no longer just extracting maintenance fees, are built around one of the core problems of innovation: failure.

Failure in the software startup world is enshrined in the idea of failing fast. Software companies have an unnatural comfort with failure. They continually throw software at the wall to see what sticks, observing how people use the software and tweaking it to get the features just right. There are all sorts of fancy phrases like "product/market fit", but at the end of the day it's plain old common sense: you rarely get it right the first, or even 31st time.

Funding software is driven by the idea of failing – the more the better, even, so long as it's quick. Venture capital's model spreads risk over numerous startups, hoping for that one giant payoff. The creation of new software is an extremely risky business. While figures like 90 per cent failure rates seem at first astounding, after a few decades of anecdotes of startup failure across the industry (and at least two myself), that 10 per cent success rate starts to look amazing.

Companies outside of the technology world are ill prepared for this kind of gut-wrenching ride. Expectations are more incremental in business improvement rather than transformative: if we put more cash into our existing business, perhaps making it cheaper to run in addition to simply selling more of our product, we can increase our return on spend. You throw together a business a case with the audacious notion that you know how much your new venture will make in the future and how much it will cost to get there. Thus, you can figure out the return on investment, or "ROI", that snipe that finance people make unsuspecting nerds hunt out in the Forest of Finance. A seasoned software innovator would sniggle at the notion that you'd trust such figures: hurtling into the unknown can't be put into a spreadsheet.


That all sounds daunting, but it's good to whiplash back to the fact that doing software actually is core to succeeding and surviving in business. While "that's fine for Amazon" may seem to apply to its profitless chewing up of every industry except mining and cement manufacturing (so far), their business and others' prove both the value of creating a programmable business, and that it's actually possible to do so. You just have to know what you're getting into and structure your executive minds correctly.

... so get started failing
Judging by surveys that show a still slow adoption of "digital transformation", there's likely a good five or even 10-year window open in various industries to become the "actually we're a software company" of your industry. Estimates vary, but surveys are showing that something like a third of organisations are actually doing anything about improving their software. The field is wide open for companies to make themselves more programmable by fixing how they do software. Now is the chance to grasp at some new innovation levers and actually do something different with your business. Either that, or look into Big Cement.


EmoticonEmoticon