You're the newly hired CMO of a SaaS company called ZetaTech.
As the new CMO, your first task on the job is to improve customer retention for your company's flagship project management tool.
All eyes are on you. The rest of the leadership has gotten you onboard, expecting you to hit the ground running, and deserve your large equity stake + compensation package in the company. The team expects you to move fast and deliver results.
You're feeling the pressure. To tackle the problem, you quickly apply a straightforward segmentation of the customer base into three categories based on company size: small businesses, mid-sized companies, and large enterprises. You believe that understanding and catering to the unique needs of each size category will improve retention.
Based on this segmentation, your team develops separate marketing strategies, creates tailored content, and even enhances product features to meet what they perceive to be the unique needs of each segment. You and the team invest significant resources and time into this approach.
However, a few months later, you see that there is no significant improvement in retention rates. You and your team are puzzled and demotivated, having put considerable effort into the segmented strategy with little to show for it.
On investigating further and understanding the situation and your customers better over the next 3 months, you discover that the initial segmentation, based solely on company size, was too simplistic. The real segmentation is more nuanced:It's not the size of the companies that matters, but their stage in the digital transformation journey. Some small companies are tech-savvy and have advanced needs, while some large enterprises are just beginning their digital journeys and need more hand-holding and basic features.
You realize that your previous efforts, based on the simplistic segmentation, have completely missed a crucial bit of nuance around the stages different companies are at when it comes to proficiency with integrating tech tools into their workflows. As a result, you have developed advanced features that are not useful to companies just starting their digital transformation, and you have not provided enough support and basic functionality to meet the needs of tech-savvy smaller companies.
So, what happened here?
You prematurely applied a structure in the form of simplistic customer segmentation, which acted as a costly headwind. It created a false sense of understanding and masked the true complexity of the problem — understanding the customers' digital transformation journey. Despite all the effort invested based on this structure, you had to work against the inertia of you initial segmentation to uncover the real issue.
It was like trying to navigate through a dense fog with an inaccurate map.
Had you taken the time to thoroughly understand your customers before applying a structured segmentation, you would have identified the correct, nuanced segmentation sooner and saved your team a lot of time and resources.
The Danger of Premature Structure
Structure, in the form of processes, frameworks, and plans, can be a powerful tool in any business's problem-solving arsenal. It offers an efficient way of reducing complexity to mere complication, providing a sense of clarity and direction.
However, one must tread carefully. Our innate desire for clarity and detail can lead us to apply structure too early, leading to misapplication and obscuring the true dynamics of the problem space.
Misapplied structure actively obfuscates and creates blind spots.
Over time, a simplistic structure that doesn't map well to the territory and vaguely approximates things tends to lock in assumptions that are not true, but made to serve the framework. This can come to obscure the right answers.
For example, MECE stands for Mutually Exclusive, Collectively Exhaustive. It is a widely known framework introduced in business schools and used in problem-solving and analysis. The framework was originally introduced by McKinsey & Company.
To apply this framework, you start by dividing the problem or situation into separate categories or segments, ensuring they are Mutually Exclusive, meaning there is no overlap between them. This helps avoid redundancy and confusion. Next, you verify that these categories are Collectively Exhaustive, meaning they cover all possible options or scenarios related to the problem or situation. This ensures that no important aspect is overlooked.
But funnily, what students in business schools don't understand is that the MECE framework does not prescribe how to segment the problem, nor does it instruct us on what are the individual parts that collectively make them up.
This is where your domain knowledge, research, and judgment come into play.
MECE is a tool to aid in structured thinking and problem-solving. But it's not a magic wand that will instantly reveal all the variables that influence a given problem, let alone reveal the solution.
Yet, I often find students using MECE as a substitute for thinking deeply about the scope and breadth of the problem and identifying its boundaries. To apply MECE effectively, you need to have several pages worth of raw research and analysis, post which you can think about how to compress the problem elegantly using MECE.
Structure should guide thinking, not replace it. And process should only retroactively pave pathways.
Bad structure is actively costly, and in complex systems, it can often be unclear what the right answer is. So the safest thing is to formalize how things already informally worked. If you are still regularly being surprised, that's a sign it's too early to structure that part of the problem.
In complex environments, imposing process prematurely is bad because it breaks flexibility as you learn more about the territory. Bad processes need constant energy to overcome the mismatch between structure and reality.
This usually happens when amateur managers try to apply a rigid process to a bottoms-up culture — just for the sake of having established a “process.”
Startups trying to scale prematurely succumb to the same problem.
In the late 1990s, Webvan was a startup that aimed to provide a home-delivery service for groceries. However, Webvan expanded too quickly, announcing plans to expand to 26 major cities in 1999. The following two years were a logistical nightmare, with Webvan ultimately losing a total of $830 million before filing for bankruptcy.
Mike Moritz, former Webvan board member, and partner at Sequoia Capital, said that Webvan committed the cardinal sin of retail, which is to expand into new territory before demonstrating success in the first market. In fact, they were still demonstrating failure in their initial market, the Bay Area, while they expanded into other regions.
The collapse of the wave function here occurred when Webvan decided to scale up before their business model was proven successful, resulting in inevitable failure.
History is testament to many such examples, where mis-applied structure has obfuscated the true nature and dynamics of the territory. When the model is wrong or ill-fitting, you'll have blindspots. Things that fit within the efficient path for the problem will be easy to see and easy to deal with. But things that are orthogonal to the efficient path will be extremely hard to see; perhaps impossible.
“That doesn't look like anything to me.”
Those blindspots can then allow problems to fester that can cause the entire system to shatter. You might lose your best employees, your best customers... sometimes, entire markets.
It's a form of ossification.
As a business leader, resist this temptation.
Understand that you are operating in a complex problem space, where the opportunity, the solution, and even the market are not fully known. Recognize that the urge to scale is driven by a desire for clarity and detail and increased revenues — mirroring the way humans tend to crave certainty and optimize for simplistic low-hanging fruits in complex situations.
But instead of rushing to structure and formalize the business, remain in the superposition of possibilities for a while longer.
By not collapsing the wave function prematurely, you maintain your flexibility and ability to adapt to new insights and market shifts. You allow the complexity to remain, understanding that you're still being regularly surprised by new developments. As your understanding deepens, your rate of surprise goes down, and you start seeing generalizable patterns, you can gradually start making more concrete decisions.
This is great advice for navigating the first 5–10 years of your career as well.
Don't prematurely optimize for salient features like titles, labels, and paychecks while ignoring the territory: what you're uniquely suited to do and excel at.
Structure is tempting, and in the absence of complete information, it serves as a refuge to shield you from uncertainty. But to solve any problem well, you need to have a large appetite for dealing with ambiguity and unknown unknowns.
Don't collapse that wave function just yet.
