As belt-tightening has become a discussion theme in the Valley, teams are debating how to cut burn to survive through a perceived winter.
Whether or not employees need to be cut, I believe that startups who want to build truly game-changing products would be best served by figuring out which customers they should be firing.
There are lots of bad customers in the world. Some require too much expensive support. Others have expectations mismatched with what a product provides or can provide. Others are just vocal curmudgeons who are likely to cause more brand damage than good. Others want a service, but just can't afford it.
Big Company Limits
Google, Amazon, Facebook, Airbnb and their brethren need to reach the maximum number of people to drive overall engagement and monetization, so it is nearly impossible for them to fire customers. And, in some of the most extreme cases, users even end up with some sort of right to a product either socially or legally (telephone, cable, big-box-retail etc.), which makes them truly impossible to fire.
This need to serve everyone, even people that don't individually make business sense to serve, is an enormous cost for big companies.
There is the obvious support burden of difficult customers; there is the brand cost of having a lot of negative people use your product. There is also a major regulatory cost. If everyone needs to be a customer, then those customers—and the governments that represent them—end up wanting and having a lot of say in what big companies do.
There are also large, less obvious costs to needing to serve everyone. If you have to build a product for everyone, you are building to the mean—the mean intelligence, the mean income, etc.
That means big companies basically have two options. They can respond by building average products at average prices, an approach that works for all their customers but can't be truly great for any of them. Or, they can respond to this—average—demand by managing the extreme complexity of customizing their offering for each user constituent. That approach can slow down development and makes managing a brand and service extremely difficult.
Neither is particularly palatable, but that is the cost of operating at real scale.
Startup Firings
Unlike big companies, if startups build average or overly complex products, they are dead. That means that they have to be willing to fire users.
The good news is that small companies, unlike big ones, can do exactly that. So, setting aside the pure marketplaces (a space I don't think startups should be building in anyway), the best way for startups to compete is to unburden themselves of customers who don't work.
This means startups need to do things like setting a strategy to select the best customers out of the pie, and serving them with better products. It also means startups need to be willing to discontinue service to people whose needs, demands, or demeanor, don't work for what they are trying to achieve or build.
If, and when, startups do a good job of selecting the best customers for what they are trying to build, and a good job in quickly firing their bad users, then they can move fast and build a differentiated product.
Over the last few years, I think that the funding environment has allowed startups to be sloppy with managing their customer base. When capital values growth above all else, it is easy to spend money to acquire and maintain bad users. It is time to rationalize.
In practice
If possible, it is always better to set up a company or service so that the wrong customers never come in the front door in the first place.
This might be the Tesla model, where you make the product expensive enough that you skip a bunch of bad customers while you grow, allowing you to compete against companies like Ford that must serve everyone. It could be the original Facebook model, where you limit your service to a specific population—Harvard—where you have much less worrying to do about scary Internet people until you are ready.
If you don't have a key strategy like that, then it becomes a matter of how you acquire customers or what channels you prioritize.
If you end up being a tech-media darling and build up a big waiting list of technologists, you will probably need to be ready to serve a bunch of demanding people. If you grow through social advertisements, be ready to serve the types of people that click on ads.
When a bad customer makes it into your system, eat the cost and rip off the bandaid.
That means letting an opportunity to save an unhappy customer float by and letting them leave, or it might mean charging more when people ask for too much. While I have never heard of anyone doing this, I have always thought it would be intellectually interesting to decide to slow down service for the worst customers until they naturally leave.
Conclusion
It might be easy to mis-read what I am suggesting and fire the customers that are most engaged and the most expensive to serve. That would be an enormous mistake. Those are the best customers for a startup and if your model doesn't work for them, that is your problem to solve. Rationalize the heavy users, don’t fire them.
But, that point aside, if you are a startup looking to do some belt tightening, the key point is that you shouldn't forget that your ability to fire customers is a hugely valuable tool, and one which big companies can't replicate.
For startups, discipline on the customer side is going to be as valuable as discipline on the team side in the coming years.