Close ☰
Menu ☰

Learning from HMV

Posted on: Thursday 24th of January 2013

Many people have a nice, neat explanation for the demise of UK music and video retailer HMV. Senior management should have acted sooner, when they had the time and the resources to respond to the Internet threat (from online retailers like Amazon and new forms of music distribution such as iTunes).

The trouble with this explanation is that it doesn’t explain anything. Yes, of course HMV should have acted sooner but why didn’t they? The stock answer – that HMV management were clearly shortsighted and stupid – doesn’t hold water. Businesses like HMV aren’t run by stupid people. They are run by very clever people. So the real question is ‘why do very clever people make this sort of mistake?’

The trouble with ‘evidence’

Here’s a suggested answer. The mistake comes from trying to do the right thing – make decisions based on the evidence.

The trouble with ‘evidence-based decision-making’ is that it begs difficult questions such as ‘exactly what evidence should we rely on?’ and ‘If there are two conflicting bits of evidence which one should take precedence? And on what grounds should we make this decision, using what evidence?”

Put yourself in the shoes of an HMV manager five or ten years ago. You are not stupid. You have read the newspapers like everyone else. You are perfectly aware of this Internet thing.

Of course, you don’t want to cannibalise your business. The more money you make doing one thing the less likely you are to welcome doing something different that makes less. Quite rightly. Why give up making lots of money in a risk free way in favour of making less money in a new way that’s more risky because it’s not tried and tested? You will only embrace cannibalisation if you really have to. Which begs the question ‘do we really have to’?

To find out, you commission some research. Not political research designed to justify a decision that’s already been made but real research designed to inform an important decision. Chances are, that research presents you with the evidence – that you don’t need to do anything quite yet. Why?

Levels of innovation

There is a qualitative difference between incremental innovation based on an existing business ecosystem and disruptive innovation which creates a new ecosystem. New ecosystems change everything: necessary infrastructure, skills and knowledge, critical resources and assets, price points and business models, key stakeholders, customer offer (utility, user experience etc) customer expectations, buying and usage habits and so on.

In its early days, it’s incredibly hard to get a new ecosystem right. To achieve lift off, you need to get many different ducks lined in a row (infrastructure, skills and knowledge, incentives, partners etc). They all need to be put in place together. There’s no point in getting the infrastructure right if you don’t have the skills and knowledge to use it properly, the resources and assets to feed and drive it, or the business model to make money out of it.

When you’ve got to get many different things right at the same time the probabilities multiply. Let’s say you’ve got a 50% chance of getting each one of seven key ingredients (infrastructure, skills etc) right at the first go. The probability of achieving this is 0.5 x 0.5 x 0.5 x 0.5 x 0.5 x 0.5 x 0.5, which turns out to be less than 1%.

That’s what happens in disruptive ecosystems’ early days. They bump along the bottom, year after year, full of obvious potential but never quite achieving it because, still, one of the key ingredients isn’t yet right. As a result, incumbents, who should supposedly be quaking in their boots, get blasé. “Ah yes, that revolution you were talking about,” they say with relaxed confidence. “It’s not happening.” And they’re right. All the available evidence supports them.

Exponential dynamics

The trouble with this is that they are right … until they are wrong. New ecosystems bump along the bottom remaining effectively irrelevant until all  their key ingredients are finally put in place. Then they achieve lift off, shifting from flat-lining to exponential growth.

Take e-books as an example. Sony’s Reader was a brilliant product, launched over eight years ago. But it didn’t have the ecosystem it needed to achieve lift off. Publishers were worried about copyright and margins. Customers found it hard to find and download books. Many different devices were being launched, each using its own standards thereby creating cost, complexity, risk and confusion for publishers and consumers alike. The devices were also expensive, selling at £300 or more.

Then Amazon addressed all these issues with Kindle – and ignited a market revolution.

Such revolutions can be deceptive however. By definition, revolutionary new concepts start out very small. Say an ecosystem innovation starts out with 0.1% of the market in its first year and doubles its market penetration every year thereafter. Even at 100% growth per annum it will still take five years to break the 1% market share barrier. Looked at from the incumbent’s perspective, it’s still bumping along the bottom for five years … five long years for decision-makers to look at the evidence and say, with apparent complete justification, “It’s just a niche. It’s not impacting our business. Look at the evidence!”.

Missing the tipping point

Now, however, there’s a different dynamic at work. If market share doubling continues over the next five years, the new concept will soar from less than 2% share to over 50%.

Having watched and waited for years because ‘the evidence shows that the new competitor/trend is not a threat yet’ the incumbent is then ‘suddenly’ forced on to the back foot, needing to make massive changes to infrastructure, business models, supplier, customer and investor relationships, internal culture, etc all together, all at the same time.

And here’s the rub. Faced with a competitor enjoying exponential growth, to respond in time, the incumbent has to act before the competitor has gained significant market share. Say the incumbent decides to respond when the upstart has 4% market share. Once it decides to respond, it will take a year or so to get all its ducks lined in a row. By that time, however, the upstart has jumped to 8%, which means the incumbent has even more to do. But while the incumbent is trying to do that, the upstart is rocketing away to 16% share, and so on.

‘Suddenly’ the incumbent is on the strategic back foot, racing to catch up with far too much to do in far too short a time. ‘Suddenly’ after years of dismissing the threat on the grounds of all available evidence, the incumbent can’t adapt far enough, fast enough. Yet, when you’re in the thick of it – looking at what appears to be a continuum – it’s almost impossible to divine exactly where and when that tipping point is.

This has happened time and time again, from the horse-drawn carriage and the motor car through today’s multiple digital revolutions, from film vs digital cameras (remember Kodak?) to music, books and lots more today.

Complements vs substitutes

Of course I have simplified to make a point. In reality, there is not one single, easily defined bucket of ‘incremental innovation’ and another bucket of ‘disruptive ecosystem innovation’. It’s a spectrum involving some elements of complementary innovation that acts as ‘an added extra’ and some elements of direct substitution. HMV was at one extreme. Its entire product line was effectively being replaced. For other retailers like John Lewis or Argos, online retailing is a complete departure in some ways (new infrastructure, new processes) but at another level it is just a continuation – ‘just another channel’ to sell the same set of goods.

That makes judging the tipping point, and the nature, timing and scale of the right response, even harder.

What this means for consumer empowerment

The same dynamic is now being played out – once again – in the arena of consumer empowerment.

Empowering consumers is emerging as a new industry – a new type of business, based on and driven by deploying information as a tool in the hands of the individual. This has many dimensions (helping individuals manage their own data, helping them make better decisions, helping them manage tasks and processes more efficiently etc) but at root it’s an ecosystem innovation involving new infrastructure, new players and stakeholders, new channels and touchpoints (and disintermediation), new sets of relationships, new value propositions, changing consumer habits and expectations, new incentives and business models.

Its becoming manifest along every part of the innovation spectrum:

  • Incremental e.g. a more informed consumer thanks to word of mouth recommendations in social media
  • Intermediate e.g. price comparison sites creating new channels to market
  • Disruptive e.g. personal data stores and related services creating new infrastructure and making completely new value propositions possible.

The challenge for existing brands and companies is to know not only how to respond but when. Here the lesson is clear. If you wait until ‘the evidence is clear’ it’s almost certainly too late.

Alan Mitchell