When is it wrong to focus on the 20% of the use cases that meet the needs of 80% of your customers?
When your goal is to develop a Minimum Viable Replacement, i.e. retire and migrate all of your customers off an existing system.
Unlike in the scenario of an MVP, where your goal is to capture a slice of the market, in the case of a Minimum Viable Replacement, your goal is to migrate your entire existing user base from one technology platform to another.
Instead of just focusing on migrating early adopters or early majority, your goal is to drive adoption by everyone, including “the laggards” as well.
In these cases you need to focus on the 1% scenarios for your UX & architecture first, not last.
Because, if you can’t support the extremes, you’ll design a system that won’t be flexible or scalable enough to serve the needs of all your existing customers, and either fail to migrate them, or have a system that fails key segments and use cases.
Complexity and impact tend to be concentrated in certain segments that require radically different approaches to serve than everyone else in order to migrate them off of an existing product/technology platform.
Complexity in this case refers to the difficulty required to meet customers expectations, e.g. the customer expects the new system to integrate seamlessly into ERP, just like the old one.
Complexity can be driven by multiple factors, depth of relationship, distance, size, variety of use cases, etc., e.g. rural customers in regions with low customer/population density far from central distribution centers are going to cost more to serve than customers with higher density in close proximity to a distribution center
Impact refers to the results associated with serving that segment or solving a specific problem/capturing an opportunity, e.g. While heavy-duty trucks make up a small percentage of the vehicle market, they represent a huge percentage of the pollution emitted.
In B2B companies, 40–90% of your revenue is driven by a relative handful of customers, whose scale and complexity dwarf everyone else’s, driving 90% of the resulting product complexity.
Since your large customers generate so much revenue and have so much leverage over you, you cannot afford to fire them, nor easily force them to adopt new technologies, especially if they view the new solution as either inferior or costly to adopt.
In addition, you’ll discover annual peaks and unusual events that, while rare, are critical to support, e.g. Black Friday traffic, or ability to close books at the end of year, respond to audits, etc.
So as part of your UX & architectural research prior to development, you need to be able to have a clear understanding of your customer base and systems, such as:
- Who are your most impactful customers, either in terms of cost to serve, revenue, profitability, and how do they differ from everyone else in terms of size and complexity, e.g. revenue, volume, # locations, geography, usage patterns, customers?
- How important are your most impactful customers, and how much power do they have in the relationship?
- What are the peaks and valleys that you need to be prepared for, and just how big are they?
- Are there rare but critical events that need to be supported?
Remember complexity grows exponentially, so don’t underestimate just how difficult it can be to meet the needs of your most complex and impactful customers. In B2B scenarios, 1% of your customer base typically drives:
- 50–90% of your revenue
- 90% of your complexity
- Has the most clout vis a vis your organization, and therefore you have the least leverage to force them to upgrade
Electric cars: An Example
The transition to electric vehicles illustrates how extremes drive impact and complexity.
Unfortunately, I couldn’t find the detailed data required to create detailed charts, but since my goal was to illustrate a concept versus provide an accurate detailed analysis of the electric car transition, I went ahead and created some data to illustrate key points.
The data, while not necessarily accurate, is directionally correct enough to highlight the pitfalls and analysis required to help you think about transition strategies.
Loss Aversion & Edge Use Cases: Why customers often do not care about averages
Average Daily Miles Driven By Consumers
As you can see in the above chart, 90% of all customers drive less than 60 miles each way to work, therefore if you can develop a car with a 100 mile range, that should meet 90% of your consumer needs. And today, the typical electric car can easily be driven nearly 100 miles without recharging, so why haven’t customers adopted them?
Could it be because customers don’t just care about having a car that meets their needs 90% of the time, but want a car that meets their needs 100% of the time?
In the chart below, I’ve incorporated not just the average daily miles driven/commute, but the max driven for vacations as well. Even though these longer trips are semi-rare events for most people, they loom large in our mind.
Average Daily Commute vs. Maximum Miles Driven for Vacation
As individuals, despite the fact that our average commute is only 16 miles, many of us refuse to buy electric cars because of the rare days that we actually drive more than the current 200 or 300 mile range of electric cars.
Most of us are accustomed to be able to just jump in our cars, and know that if we wanted to, we could drive across the country in our existing cars without worrying about running out of gas, and the thought of switching to a car that might prevent us from doing that is unappealing to say the least. Even if we rarely if ever use this capability, and we could potentially just rent a car for those few days, the idea of giving it up is very disturbing to say the least.
Loss Aversion: A bird in the hand is worth two in the bush
The challenge is, we’re hardwired to focus much more on the potential loss of something and maintaining the status quo, than we are on the potential gain of something else, i.e. a bird in the hand is worth more than two in the bush, even when we rarely ever use the bird we have.
The concept is known as loss aversion. According to psychologists Daniel Kahneman and Amos Tversky, users generally fear a loss twice as much as they are likely to welcome an equivalent gain. In the world of technology, loss aversion makes people more likely to stick with what they have despite compelling reasons to change, i.e. the devil you know versus the devil you don’t, which is just one reason why it’s so difficult to get rid of rarely used features.
So not only do we have a psychological aversion to losing rarely used capabilities, but our fears have an actual basis in reality. After all, even if you’re doing a great job 99% of the time, you’ll still get fired by your boss or your customers, when you fail at critical points, e.g. even if your ecommerce system works great 363 days a year, but it can’t handle the loads on Black Friday and Cyber Monday, you’re still going to lose customers, revenue and your job.
If the concept of loss aversion is too abstract, just remember that it takes 10 or more, “Great job!”s to make up for every, “Oh crap!”, whether at work or at home.
So whether it’s range for an electric car, ability to handle spikes in demand or address rare but critical audits, we’re unlikely to trade our existing technology for a new one if it can’t handle these extreme use cases.
Recommendation: Focus not just on the repetitive events, look for spikes and unusual occurrences. Rather than just look at data on an aggregate basis, which tends to hide anomalies, structure your queries in such a way that accentuate the probability of finding more extreme events, e.g. look for spikes or aberrations that happen at the end or beginning of the day, week, month or year. Ask about the worst times of the year, have customers tell stories about their worst or hardest days, etc. No guarantees you’ll uncover all the important extremes, but at least you will have tried.
Understanding which segments drive the biggest impact
Total Miles Driven by Segment vs. # of Drivers by Segment
If you’re developing an MVP, it 100% makes sense to target your efforts on the 90% of the market that drives less than 60 miles per day. However, 40% of the miles, and therefore greenhouse gases are driven by 10% of the drivers, so even if you migrate 90% of your customers, you are still nowhere close to being able to declare victory.
In fact, even if you migrate 99% of all drivers over to electric, you still face the fact that the 1% are responsible for 10% of the miles driven.
In many companies, 1% or less of the customers actually drive more like 30 to 90% of their business, e.g. despite UPS literally having millions of customers, just one, Amazon, represents 12% of UPS’s revenue.
Unfortunately, getting the data required to really understand which portion of your customer base really drives your business can be difficult, but it’s critical to understand, otherwise your expected victories may just be mirages, with the actual finish line, much, much further away.
If your goal is to either migrate an entire customer base to a new solution, or to make the biggest impact possible, it’s critical to both understand the “extreme” use cases that are the hardest to solve for, and which customers drive the biggest impacts.
If you don’t understand those two aspects, you’ll fail to:
- Design and plan for solutions that meet all of your customers’ needs, thereby settling yourself up for both overestimating adoption rates and underestimating the costs.
- Achieve your desired impacts, since you won’t have correctly allocated customer/segment value, and therefore may be fooled as to your actual progress by focusing on the wrong metric.
So whether you’re working on reducing pollution or migrating your customer base off an existing product, you need to:
- Understand the extreme use cases that impact the complexity/size of the challenge ahead of you.
- Identify the right success metrics so you focus on the most important challenges, truly understand your impact and know where you are on your journey.