Some Assembly Required

May 6th, 2017

The perfect product requires no humans. It just works. Opening up, turning on an iPhone, one-click buying on Amazon. Removing humans gives our industry not just incredible margins, but consistency in services provided as the business grows and changes.

Sentry is the same. Our documentation and FAQs guide first time setup while design builds a product that is intuitive to use. This allows us to offer a cheap service while focusing our resources on improving the product.

Coming from this, why hire customer success and sales?

A few definitions

Before I go on, let me define these roles. My definitions are specific for B2B SaaS. To avoid flowery job descriptions (the worst kind of bullshit), I define roles by how they are measured. Yes, jobs are more than their metrics (particularly the creative fields), but in business roles, the metrics make the man.

Sales is measured by revenue from deals closed. This means their day to day is always finding customers and getting them to buy the product or speaking with existing customers and getting them to buy more. The question “Why hire sales?” is the easiest to answer. Because they make your business money better than your self-serve business does.

Customer success is interesting because it is often confused with account management.1. Account management was fundamentally measured by churn. Once sales had closed the deal, account management ensures the customer remains a customer. Generally, the best way to reduce churn is to get the customer to use the product they just bought, but it’s a means to an end.

Customer success, on the other hand, is focused on the success of the customer. Feels obvious, but obvious is often lost in the business doublespeak. The “success of the customer” is how much the customer actually uses and gets value from the product. There’s a nice side effect of churn reduction, but it’s not the core goal.

Sounds familiar, right? The difference arises from the odd incentives of business purchasing, where the buyer is usually separate from the end user. The reason why products like Concur are so unfriendly to use is because the buyer almost never uses the product and the users don’t get to weigh in on the purchasing decision.

An account manager’s time is better spent teaching the buyer how valuable the product is (which usually involves driving internal adoption). Usually this involves showing an admin how to make dashboards. Customer success will spend its time driving usage amongst the end users, usually through product education or training.

Back to Product

If the perfect product has perfect adoption, why rely on humans to drive it? Is it to “monkey patch” the product while on-boarding is perfected? A rapidly iterating product should never have perfect on-boarding because the feature landscape will always be shifting. Humans can smooth over bumps until the product can be fixed and minimize churn in the process.

Yet, that seems unsatisfying. Sentry has minimal churn. Per head, customer success costs less than engineer, but not by much when you full account of the cost of “a head” (office space, insurance, etc). Why not hire 1 engineer instead of 2 customer success managers?

A lot of joining and growing an early stage startup has been wearing a lot of hats, including customer success. I’ve easily spoken to hundreds of customers, and while I found it hard to quantify the value of those interactions, yet I always knew that without a doubt they were valuable.

Trust. Belief.


People don’t come to our site because of our code. They come because someone they trust has told them it’s a valuable service.

  1. It’s a common mistake and sometimes a deliberate one. When customer success is the new sexy, employers and employees alike love to rebrand account management as customer success.