How tolerating failure sustains competitive advantage

Dan Levinthal is a professor of management at the Wharton School at the University of Pennsylvania. He is also the Michael J Crouch Visiting Professorial Fellow at the Australian School of Business where he will be giving a series of seminars on innovation and the experimenting organisation.

Levinthal says that companies seeking to foster new possibilities need a mix of discipline and some element of tolerance for at least partial failure. It is about balancing exploration and exploitation. Levinthal recently spoke with Knowledge@Australian School of Business.

This is an edited transcript of the interview.

Knowledge@Australian School of Business: Dan Levinthal, what do you mean by “the experimenting organisation”?

Dan Levinthal: I think it might be useful to back up one step before we get to that question. I would suggest that organisations face a dual challenge. One is that you need to meet performance demands in the current time period. At the same time, you have to prepare yourself for some unknown and changing future circumstance in which you’re going to find yourself.

So, in general abstract terms, the firm needs to leverage or exploit its current market positions and refine that. At the same time it needs to explore, discover new basic competitive advantage.

The question is, how to manage that dual challenge? [This brings us back] to your question about the innovative organisation. Most innovative efforts are likely to be a failure. So if we think about a learning adaptive system, most of the pushback you’re going to get is saying, “You know what? Actually leveraging our current position, doing what we already know how to do, is [looking] pretty good.”

So part of the first challenge is to recognise the likely negative reinforcement you’re going to get in innovative activity. The returns are going to be more distant and more problematic. And so we need to set up systems that can allow [and] sustain this sort of balance; systems that might be tolerant for perhaps initial failures, initial feedback that may be negative, but for initiatives that may over time be more promising.

Knowledge@Australian School of Business: Is time the issue? Many organisations have to report their financial results quarterly. The shareholders may expect their cash at the end of the year. For many organisations coming up with innovations, that’s just too short a time frame.

Dan Levinthal: Exactly. If [an organisation] wanted to maximise next quarter’s return, it could turn the faucet off on R&D. That would make next quarter’s return look better. So there’s this kind of shifting between past returns and the future, versus the immediate current returns.

And so, internally, an organisation has to address the issues of how it allocates resources and thinks about that challenge, but also how it relates with external audiences. It has to explain the positioning for forward returns versus current ones. And it needs to create a sense of milestones, [while still maintaining] controls. We can’t just say: “Have fun. Here’s 6% of revenue, go to score, and play.” But at the same time, there needs to be some tolerance for the temporal delays, the possible failures, of those innovative efforts.

Knowledge@Australian School of Business: Google famously gives people one day a week off to go and develop their innovations. Why don’t we do that in every company?

Dan Levinthal: Google actually borrowed it from 3M, which did it earlier – where you allocate engineers [and give them] a little time to do what they want. That’s a wonderful way to begin to nurture exploratory efforts and I think that’s useful.

But you also have to recognise [or] imagine that your clever little hobby may work out. So you’ve been pottering around with something that gets to be called Gmail. At a certain point, you’re going to need to scale up and [obtain] some serious corporate resources to do that. But bear in mind that [individual experimentation] is just one day a week and that is precisely that sense of balance of exploration and exploitation. You are working hard [the other] four days to pursue the existing corporate mission, the mandate given by your operating unit, and some fraction is exploration. And so it gives you that balance. And then again, secondly, it gives you that chance to begin to prove the business case. And at that point, you can take it to the more traditional corporate resource allocation process and say, “Look, I have a prototype. It actually kind of works.” And it may have taken you a couple of months to get to that point. But now you can demonstrate a more credible business case for pursuing this initiative called Gmail, and not, “Gee, I walked in. There was a glimmer in my eye, maybe it will work, maybe not. But you know, give me some time.”

So that’s a quite useful mechanism. The new initiative is often the way you can ride up through an organisation, certainly it’s been true at 3M to some extent, at Google as well. That is, by defining a new business opportunity, a new project. And if that gets approved, that’s the kind of fast track to rise up through the ranks as that initiative grows with your efforts.

Knowledge@Australian School of Business: When rising up through the ranks, managers can have many conflicting demands; they judge against a whole range of criteria. How are they going to have the commitment to see that this is going to be judged in the future against those criteria and will be a success?

Dan Levinthal: Yes, particularly when the new initiative may not have any direct positive financial outcomes, it’s going to take some time. So what is the benchmark by which we are making progress? Do we define it in terms of technological milestones? Do we define it in terms of some thresholds of market acceptance?

One popular business management approach, the balance scorecard, is a way to recognise the polarity of objectives. We care about customer service, we care about being innovative, we care about being efficient. And when level, that’s all true and good. But per your question, what does that mean for an operating executive? [They may think]: “You’ve given me 20 things to worry about. Each one of them is quite demanding of my time and energy. And not only that, they might even run in conflict. If I want to maximise reliability, I wouldn’t want to innovate, because that’s likely to have some glitches and some problems.”

When level, the balance scorecard can be useful as a kind of indicative diagnostic but I think it’s sometimes pushed or suggested as a way to guide managerial action. What I would suggest and observe is that it can actually freeze managerial action. And then what you get is, “OK, here’s our balance scorecard”, but over beer at the end of the day, I and my buddies are going to [be saying]: “They told me these 20 things, but really, what do you think is going to drive our promotion, our bonus this coming period?”

And so you end up speculating about what are the true drivers by which evaluations are occurring versus goals [that] are supposed to align action, align activity of the organisation. So be explicit; recognise the focus is going to change over time, but any moment in time you need some clarity and alignment of activity and energy within the organisation.

Knowledge@Australian School of Business: When we’re talking about the experimenting organisation, we’re often saying that managers should think outside the box. What constitutes that proverbial box?

Dan Levinthal: One of my suggestions in that regard is, often when people think about innovation, the emphasis is on variety. There’s the phrase: “We need to let a thousand flowers bloom.” That’s lovely, but imagine, to keep the gardening imagery up, you only have one type of lawnmower. Imagine that we have this variety, but the organisation says: “You know what, this is our current product market, these are our customers, this is the sort of thing we like.” You’re squashing that.

So I think one thing, per this out of the box – and it often gets unattended to in thinking about organisations as innovative and changing – is being open to some kind of heterogeneity, some eclecticism on what the selection criteria is. This has been our current market; fine, these are our current customers, this is what they like. Maybe this new initiative is unappealing to them; but if we took it to another context – take this technology to a new environment, take this product or service to a new customer segment – they may in fact be receptive to it.

If you’re not open to different criteria about what it means to be successful and effective, you really close in the system and you might actually squash lots of interesting and wonderful innovation that the organisation has generated.

Knowledge@Australian School of Business: How do we get managers to be committed to jumping to a new type of action and being committed to innovation?

Dan Levinthal: If managers feel that this effort is going to be recognised, reinforced, then they are kind of inclined to do that. While life is simpler operating in that proverbial box and with the status quo, I think there’s a natural human desire and energy to be a little bit creative, to be innovative. But there’s [also] a strong presumption that we’re going to get our knuckles rapped for doing so.

So I think there’s some latent intent. Also, it’s important to recognise when a manager’s out in the field talking to customers, or the engineer is exploring the new technology, that those efforts are creating natural experiments. We often think, “Oh, here it is, the central R&D lab. They’re the creative types, and the rest of the organisation is just aligned with our current activities.” But lots of interesting new ideas about possible products, or a service way to approach the market, are occurring in the context of the ongoing business operation. [It’s a case of] then being able to tap into that and allow those creative thoughts to bubble forth, be reinforced, and scaled up.

Knowledge@Australian School of Business: When you first look at new innovations as they come on-stream from R&D, is there some easy way of working out which ones are going to be a success and which ones are never going to make it commercially?

Dan Levinthal: Typically, in the corporate side – whether it’s exploring the possibility of a new geographic market, a new product market – the feedback is going to be highly mixed. There is some element and ways it kind of worked, there’s some element where it didn’t quite work; and maybe if we tweaked the proper business proposition a bit, it might succeed.

Management scholars, by pouring over this notion of real options – often in a somewhat naïve way – are really not confronting the difficult question that these interim outcomes are likely to be partial failures, partial success; [nor confronting] the business judgment about how to sort that out.

On the one hand, we don’t want what in companies is sometimes talked about as “the living death”. You know, it’s the project [that’s] not really going anywhere, but we never pulled the plug; it’s just tying up time and energy and resources. But at the other end, you have to recognise that even initiatives that ultimately prove quite successful are likely to look not terribly promising in their early incarnations.

We need to face full-on that this is going to be a difficult judgment matter. You’re going to need milestones with which to gate these efforts – you don’t want to be in the realm of the living dead, the zombie that doesn’t disappear – but at the same time have some sense of flexibility, [be] open to ways in which [something] kind of worked and didn’t, and that maybe [you could] shift 20 degrees from the initial business trajectory.

There needs to be this mix of discipline and some element of tolerance for at least partial failure. If you’re not tolerant for the failure, why would I experiment?

Knowledge@Australian School of Business: If someone following this discussion was running a company, what one idea would you want them to take away?

Dan Levinthal: I’d go back to the fundamental idea of a balance of exploration and exploitation, and recognise some of the challenges of sustaining that balance, that the pressures will always be pushing to exploit and certainly you cannot neglect that. But what are the systems that will provide for the ongoing exploration of new possibilities inside the firm?

Knowledge@Australian School of Business: Dan Levinthal, thank you very much.