I feel like I've recently been on a crusade against business analysis. As some of you will know I really, really like to prove things so I am most certainly in favour of analysis, my problem is the amount of it that people feel that they need to do in order to make a decision. And worse still, having asked for lots of data and analysis it then goes on a shelf without being read. I've been trying to change that situation within our practice, project by project.
Part one of my crusade is to persistently remind people that analysis must ultimately help someone make a decision and then enact that decision in the world. That may seem banal, but my observation is that it's very easy for analysis to become the task, subsuming the intention of the strategy process. This leads to intellectually pure answers, but not decisions.
Part two is more recent. It's the culmination of something that's been nagging at me since we started to build ventures for clients, initially using Lean Startup, more recently our own proprietary method that extends and improves that idea. In venture building one is always creating new tests to increase certainty about the business in three domains: do customers want it, can we do it and can we make money out of it.
Working with my colleagues with a more design-related background reminds me constantly that each of these tests is a prototype of a component of the overall business (as opposed to the prototypical products or solutions that they typically focus on). This makes me think about a reframing of business analysis: if we considered every analysis as a prototype of a decision would we reduce wasted effort and increase the likelihood of realizing a better decision in practice?
My train of thought starts with the domains of strategy. Oddly, this isn't something that I've given much more than idle consideration over my career: I've just tried to solve problems (note that I didn't say 'help leaders make decisions' - I'm very guilty of everything I've said above). A simple construct for the development of strategic intention would be:
Although this type of thing can show us about reactions to a value proposition, it doesn't show us anything much about our employees' response to our new aspiration, or about competitor response or our capabilities to scale the proposition or our ability to govern the business. Given the ultimate impact of a corporate strategy and the value at stake, it seems to me that larger scale prototypes could be tried.
For example, rather than analyzing a market endlessly and performing scenario analysis to try and understand how it will respond to new products, why not launch a new business specifically to test how it responds in reality, or buy something that does the same? In scaling such a business, you'd also be prototyping the capabilities and management systems that would be needed to run a similar business at scale, hence derisking the build of large scale enterprise productivity systems (which have a tendency to be iterative rather than transformative, despite the regular presence of "Transformation Directors" in organizations). One could even use incubators and accelerators to do the same through X-Prize competitions of funds in very targeted domains.
In this way, strategy becomes a continuous process based on connecting hypotheses about the world to metrics that describe the behavior of that environment and prototypes that impact it. It probably looks something like this:
Part one of my crusade is to persistently remind people that analysis must ultimately help someone make a decision and then enact that decision in the world. That may seem banal, but my observation is that it's very easy for analysis to become the task, subsuming the intention of the strategy process. This leads to intellectually pure answers, but not decisions.
Part two is more recent. It's the culmination of something that's been nagging at me since we started to build ventures for clients, initially using Lean Startup, more recently our own proprietary method that extends and improves that idea. In venture building one is always creating new tests to increase certainty about the business in three domains: do customers want it, can we do it and can we make money out of it.
Working with my colleagues with a more design-related background reminds me constantly that each of these tests is a prototype of a component of the overall business (as opposed to the prototypical products or solutions that they typically focus on). This makes me think about a reframing of business analysis: if we considered every analysis as a prototype of a decision would we reduce wasted effort and increase the likelihood of realizing a better decision in practice?
My train of thought starts with the domains of strategy. Oddly, this isn't something that I've given much more than idle consideration over my career: I've just tried to solve problems (note that I didn't say 'help leaders make decisions' - I'm very guilty of everything I've said above). A simple construct for the development of strategic intention would be:
- Describing the environment in which a decision will be made (a.k.a. scaring everyone/ creating a burning platform)
- Making strategic choices that define how the organization will react to, and attempt to shape the environment to its advantage
- Translating corporate strategy into commercial or business-unit level strategies
In my experience (1) and (2) are typically bundled together. I'm not really happy with this anymore as I very rarely arrive in an organization in which the leadership have a similar view of the nature of the market in which they play, the likely future of that market and the internal issues that impact the business. Spending some proper time setting the scene probably makes sense, before moving on to corporate-level strategic choices.
I typically see two types of analysis used to underpin these three domains: anecdotal and research-based. As stated in the previous paragraph, anecdotal tends to be the dominant method in describing the environment, I suspect because its easiest to abstract the subject of the nature and future of the environment into "it'll change and it could be BAD". Since there will almost certainly be a business case produced in the formal strategy processes, research-based analysis dominates (2) and (3). I personally don't like business cases very much as they are justifications of the overall decision, rather than honest analysis used to make the decision. It's therefore both easy and desirable to make the business case look good, hence destroying any real chance of the strategy being realized. Anyhow, the research-based analysis involves lots of people sitting at desks looking at Google, making models and occasionally trying to find 'experts' to interview. It's all a bit ridiculous when you consider how much it costs.
Getting back to the point, could there be a third way? Could we use prototypes to make better decisions by more accurately expressing the nature of the market and judging the likely validity of ideas? I think yes... I mean, seriously, would I have written all of the above if I thought 'no'? :)
The main question is how far we're willing to go with prototypes. Great strategies make deliberate, stretching choices that create the conditions for success and that we ultimately hope will enable us to win. That means being ambitious - not 'playing to play'. To really improve our understanding of the behavior of a market, one has to be quite active and deliberately set out to make a change that one can then observe propagating.
Because of this a good strategic prototype probably needs to go beyond the most basic formats: a mock-up of a value proposition that is introduced to customers to gauge their response or an adcept launched on Facebook or Linkin to test sign up rates. These are far superior ways of derisking a strategy than analysing what other companies are doing and thus inferring that we can do the same, particularly if the prototype's branding is carefully thought through. Responses to a startup are very different to responses to a mature enterprise. Not better or worse, just different. Likewise, we can experiment on the capabilities of an enterprise or the reaction of the channel by changing incentives for sales staff, tweaking real operational processes to see what happens to productivity. All of this is prior to concluding on a strategic option, remember. It makes strategy a more activist process.
Because of this a good strategic prototype probably needs to go beyond the most basic formats: a mock-up of a value proposition that is introduced to customers to gauge their response or an adcept launched on Facebook or Linkin to test sign up rates. These are far superior ways of derisking a strategy than analysing what other companies are doing and thus inferring that we can do the same, particularly if the prototype's branding is carefully thought through. Responses to a startup are very different to responses to a mature enterprise. Not better or worse, just different. Likewise, we can experiment on the capabilities of an enterprise or the reaction of the channel by changing incentives for sales staff, tweaking real operational processes to see what happens to productivity. All of this is prior to concluding on a strategic option, remember. It makes strategy a more activist process.
Although this type of thing can show us about reactions to a value proposition, it doesn't show us anything much about our employees' response to our new aspiration, or about competitor response or our capabilities to scale the proposition or our ability to govern the business. Given the ultimate impact of a corporate strategy and the value at stake, it seems to me that larger scale prototypes could be tried.
For example, rather than analyzing a market endlessly and performing scenario analysis to try and understand how it will respond to new products, why not launch a new business specifically to test how it responds in reality, or buy something that does the same? In scaling such a business, you'd also be prototyping the capabilities and management systems that would be needed to run a similar business at scale, hence derisking the build of large scale enterprise productivity systems (which have a tendency to be iterative rather than transformative, despite the regular presence of "Transformation Directors" in organizations). One could even use incubators and accelerators to do the same through X-Prize competitions of funds in very targeted domains.
In this way, strategy becomes a continuous process based on connecting hypotheses about the world to metrics that describe the behavior of that environment and prototypes that impact it. It probably looks something like this:
If this were feasible (and I believe that Digital Economy sensing and data processing technology makes it so) then it would revolutionize the practice of strategy, which is for the most part a periodic project-based exercise, realized with waterfall delivery that generally fails to have significant impact. Productivity increases. Everyone wins! A victory for theory...
...but, there's a problem, and it's a big one. Corporate governance means that anything that involves an acquisition or significant capital outlay needs to be reported to the market in some way or another. This makes it quite difficult to launch prototypes of sufficient scale without making market-facing announcements that may bias the validity of the test. It's a massive problem, and one that I'm going to try and tackle in the next couple of posts.
Comments
Post a Comment