Last year I snapped a photo of a curious bumper sticker, and posted it on Facebook. It read, If you’re prepared for flying irradiated zombies that can swim, then you’re prepared for anything. I figured the car’s owner to be either a risk manager or an insurance agent. Who else would be moved to share this wisdom?
When dealing with uncertainty and risk, we humans follow a pattern. We collect an array of facts about things that matter. We relate these facts to other facts. Then, we assess that mass of information to glean understanding for how future outcomes might unfold. Ultimately, we must untangle this messy conglomeration of fact and feeling to answer a vague question: now what?
“Any decision relating to risk involves two distinct and yet inseparable elements: the objective facts and a subjective view about the desirability of what is to be gained, or lost, by the decision,” wrote Peter Bernstein in his book, Against the Gods. Here’s where things get interesting, because at this point, the pattern begins to fray. The actions that we plan are based on our dynamic, individual mix of optimism, confidence, and loss aversion. A constant mental tug-of-war that has shaped our personalities since we were tiny infants. These emotions combine within us as uniquely as water crystals in snowflakes.
Maybe, just maybe, that irradiated zombie visage pushes a bright-red risk button for someone – especially someone who has learned about drone technology, and recognizes its potential sinister uses. But one person’s sincere concern over possible zombie infestation can be another person’s perfunctory dismissal of an irrational fear. In business development, I still marvel that people unabashedly proselytize rules for things.
Software algorithms, by contrast, are coldly indifferent when it comes to assessing risk. Give a computer clean data along with a set of logical rules for analyzing it, and you’ll get consistent interpretations. Don’t like the results? Simply refine the algorithm! Alas, for now, we humans are stuck with pesky biases that interfere with the uniformity we often crave.
This yin-yang of risk seeking and risk aversion between and within individuals creates immense organizational challenges because people – not algorithms – still make most of the high-level, strategic decisions in an enterprise. And executives suffer a love-hate relationship with uncertainty by sometimes confronting it, sometimes sweeping it under the rug, and sometimes, doing both. So here’s the problem: how do you bubble up the most relevant, consequential uncertainties, and put them into a collaborative space for people to consider, analyze, and use for strategic planning and decision making?
Not surprisingly, there’s a process for that! Here’s how to put uncertainty to work for your company:
1. Start with a deterministic statement. In most organizations, they’re easy to find. For example, “In the next five years, we will increase our annual revenue by seven times our current level,” or “our target operating margin for next fiscal year will be 20%.”
2. Identify areas of concern that might inhibit achievement of that goal or target. This requires people – preferably, many people – to raise a hand and say, “well, what about, what about, what about, and what about . . ?” Write those what about’s on the white board, and you’ll develop a picture of specific uncertainties that exist in what was an opaque swirl of unknowns. Some thought starters: “Customer demand,” “parts availability,” “meeting hiring targets,” “economic conditions,” “currency valuations,” “pending regulations” “competitive product introductions.”
3. Prioritize those areas of concern by ranking them from most likely to apply pressure on revenue results, to least likely.
4. For each high-priority area of concern, take a view on a related process, and, over a specific planning time frame, forecast the minimum, most likely, and maximum values that could occur. Example 1: “in the next 12 months, revenue from service agreements will not be less than $10 million. The best result we could achieve will be $30 million, but we’ll probably be somewhere around $22 million.” Example 2: “next year, our worst case for customer churn will be 18,000, our best case will be 7,000, but we should anticipate churning about 14,000.”
Note: the most likely value is not necessarily the average between minimum and maximum – and most often, it isn’t. For example, the most likely revenue produced by a new sales rep will skew toward the minimum value, and the opposite is typical for a more experienced rep.
5. For every minimum value, explain why it’s not possible to achieve a result that is lower, and for every maximum value, explain why it’s not possible to achieve a result that’s higher. For example, there might be a ceiling on units sold because factory production might be unable to exceed a specific capacity, and outsourcing manufacturing isn’t feasible. Or, for planning quota achievement by sales rep, the minimum value could be derived if every territory generates run-rate revenue of $X million.
As daunting as uncertainties might be, they serve a vital function to open conversations, and to enable people to develop an understanding of probabilities for business outcomes. By now, you have recognized that for every target or goal, there are identifiable circumstances that are consequential for achieving them. The variables have many possible values, and they can combine in thousands – or millions – of different ways. If unemployment increases by 6%, AND average sales price is $145.00, AND salesperson productivity remains static, AND the development team delays the new software release by six months . . . will the company meet its financial goal? You get the picture. When forecasting an outcome of interest – revenue, net profit, new customer acquisition, average revenue per transaction – the sheer magnitude of number crunching requires software for simulating the results.
Through analytic tools, insight for very complex uncertainty problems can be revealed. Managers can ask which outcomes are most likely given a particular condition, or set of conditions. How might price increases affect demand? Which projects will likely achieve the biggest increases in revenue? Probability modeling makes the answers accessible.
Next month’s column, How to Model Revenue Risk, adds five steps to the five provided in this article, and using an example, I will illustrate how to solve common problems in revenue uncertainty through Monte Carlo simulation.
Put uncertainty to work for your company. What begins as a whirlwind of uncertainty can be used to gain clarity on how to achieve your most important, mission-critical goals. An infestation of flying irradiated zombies won’t be on everyone’s list of worries, but without first having a conversation about what is, we can never know for sure.
To read the first article in this series, please click here: Revenue Uncertainty – Part I: Known Unknowns, Unknown Unknowns, and Everything in Between.