It’s summer, and while most folks are packing novels and guilty-pleasure magazines for lazy days sprawled on the beach, my beach bag is a bit different this year – I’m going to catch up on some business classics like “The Effective Executive” by Peter Drucker while also getting up to speed on the latest behavioral science trends. Before you dismiss me as a complete and hopeless nerd, let me tell you that an article in the most recent edition of Behavioral Scientist has already paid dividends in the form of a concept that has crystallized my thinking on two critical aspects of OKRs. Bet I have your attention now.
Production Functions (Sound boring, but are very cool)
The article, by University of Chicago Law Professor Lee Anne Fennell, is titled “What shape does progress take? Don’t assume it’s a straight line.” Progress…assumptions…straight lines…this article had me at hello. Fennell explores the common misconceptions around the predictability and linearity of progress in various endeavors, from personal goals to large-scale projects. She argues that our default assumption tends to be that inputs (like effort, time, or resources) will lead to outputs in a predictable, linear fashion. In other words you put in the work and, bada-bing bada-boom, you get the results. But that’s not always the case, as this “production function” she’s describing can take different forms. For example:
- Step function: Sometimes, considerable inputs result in no visible output until a certain threshold or “critical mass” is reached. Think of building a bridge. Unless you’re filming the next Ryan Gosling action flick you don’t want to be driving on a half-completed bridge. In the case of a bridge, no useful output is realized until all the segments are constructed to complete it.
- S-Curve: Initially, progress is slow and returns are minimal, but once a certain level of effort is invested returns increase dramatically, before eventually hitting a plateau. I can personally relate to this one – several years ago I became interested in performance driving and my initial attempts on the track were clumsy and frustrating, causing the destruction of many orange cones. But with practice and effort I was able to competently navigate the track at speed – with no menace to innocent bystanders or cones I should add. Eventually, however, my progress stalled which I understand is typical. A classic S-curve.
- Other examples of non-linearity: The article also notes other forms, like inverted curves where initial large gains are followed by diminishing returns, and scenarios where returns may fluctuate over different phases of input.
These insights have significant ramifications for how organizations approach the use of OKRs. In most organizations, the default cadence for OKRs is quarterly – every ninety days you set, hopefully execute, and then reset your OKRs. That simple process assumes a linear production function – you set OKRs, you apply effort to achieve the OKR, and hope/assume your inputs in the form of time and other resources will lead to the achievement of the OKR. The process is then repeated quarter-in, quarter-out. But is the linear production function suggested here a realistic assumption? I would argue that if you’re pursuing challenging OKRs, a non-linear production function is very much in your future, and that reality will impact at least two critical decisions you’ll need to make when implementing OKRs: The right cadence, and whether you’ll allow the use of milestone key results.
What is the right cadence for OKRs?
While some organizations have pivoted to trimesters, for most the default OKR cadence remains quarterly. Given what we now know about production functions, particularly step and S-curves, are either of those timelines the right choice for your organization?
Let’s imagine an OKR related to entering a new market, which for most companies would represent a significant challenge, and hence be a good candidate for an OKR. An outcome-based key result could be the percentage of share captured in the new market. Even if we assume that whoever developed this OKR did the necessary “pre-work,” for example: carefully defining the target market, calculating the total addressable market value, estimating sales volumes, etc., etc. it’s still going to be difficult for them to achieve their market share key result target in ninety days. Promotions will undoubtedly have to be staged, sales materials generated and distributed, and myriad other activities engaged in to drive the market share number. In other words, ninety days may not be enough.
The velocity of change in modern organizations is undeniable, and the agile approach to goal setting represented by OKRs is a much-needed antidote to ensure you stay ahead (or at least keep up with) the change curve. But a question you must ask, assuming once again that you’re selecting challenging, stretch OKRs, is how long you’re willing to invest in their achievement? In the example above, if you don’t hit your market share number in a quarter, do you abandon the OKR? I would hope not. My advice, as it is with all things OKRs, is to remain flexible. While a quarterly cadence will work for many OKRs, others, especially those that are vital to innovative and experimental endeavors, may require significant initial effort to produce results, and must be given the requisite time to produce results.
Yes, I’m going to talk about milestone key results…again
The first installment in my “Courting OKRs Controversy” series focused on what I consider, under certain circumstances, to be the prudent use of milestone-based key results. As I mentioned in that post, many experts in the OKRs field write and speak in a near religious fervor about the necessity of using outcome-based key results. “OKRs are about outcomes!” they cry, and nothing else will do. Of course I’m in favor of using outcome-based key results to gauge the achievement of an objective, but I’m also open to the reality that many organizations face when implementing OKRs, and that is the fact (yes fact) that you’ll often require milestone-based key results before you can get to an outcome.
Let’s return to the example above of a company that wants to enter a new market. As I pointed out, before they’re able to rake in revenue from their arrival in the new market several things must be accomplished. When sharing the example I was generous in suggesting we might assume they completed the necessary pre-work, but based on my experience of seeing innumerable examples of comparable OKRs, many times that pre-work hasn’t been done and is exactly what will initially be captured by milestone key results. Is that a bad thing? If you’re a proponent of the “Key results must represent outcomes only!” then yes, it is. But I would argue that milestone key results are both acceptable and appropriate in this situation. They allow you to gauge progress, make necessary course corrections, and provide early warning signals as to the likelihood of ultimately hitting your outcome-based key result.
A winding road to success
At the end of the day, the key takeaway here is that progress rarely follows a straight line. Whether you’re working towards ambitious OKRs or just trying to become a better driver (without terrorizing innocent cones), you’re going to hit some twists and turns along the way. Rather than getting frustrated when things don’t move in a linear fashion, it’s crucial to embrace the messy reality of non-linear progress. Recognize that some OKRs will start slow and then accelerate rapidly. Others might deliver big wins early before tapering off. And sometimes you’ll be stuck spinning your wheels until you hit a critical mass.
The best advice is to stay flexible and adaptable. Don’t get too hung up on arbitrary timelines or cadences just because you read about them in “the book.” If an important OKR needs more runway, give it more runway. Use milestones to chart your progress, make course corrections, and keep momentum up during the challenging phases.
Ultimately, by internalizing the wonderfully complex nature of non-linear progress, your OKRs can become living, breathing guides that flex with reality. It keeps things fun, engaging, and prevents total meltdown when the road gets twisty. So buckle up and enjoy the ride – straight lines are overrated anyway!