Justifying difficult customers in support

In my past life I was working in health insurance. Insurance works in a method called pooling. 10 people pay $10 each, and when one has an accident that costs $80, the insurance company can pay the claim, and still make some money. There’s the reverse of course, where health insurance companies collect $100, but the claim is $150. Then they’re screwed, but they make up for it in investments and interest.

We handle a lot of support between Beanstalk and Postmark. We have, and I say this all the time, the best customers in the world. They are designers and developers, just like us. So most support is really great and easy to answer. On occasion a client or two come along who starts building an entire reputation amongst our team. They are difficult, and not in an angry way, more like a lazy or inexperienced way. Our team will help to a fault sometimes. We will provide articles and screenshots and links to books, but the customer would rather have us do it for them. Then we summarize what’s in those screenshots, and books. Still nothing. We’ll spend hours trying to teach someone how to use version control, or even write them some code to access our API. In one of those moments, I had a thought: How much support is too much for $15 a month?

I think it’s like pooling. In insurance, we take an approach that many pay for one. The same goes with this. If we had 10,000 accounts that pay $15, and 5 of them cost me $500 in hours of support, it’s totally ok. When it get’s bad is when one customer starts taking away from others, using that $500 to themselves. I don’t think we need to hire more support people, I think we need to fire some clients. And, we do. We don’t want to owe more than we make, we don’t have investments to make up for it.

So, going through the extra effort for some customers, even when difficult, is fine. Just make sure there is a balance.


Our products