I recently started using Superhuman, the popular $30 per month email application, that's getting lots of buzz. It's a wonderful product. It solved my email overload problem.

I would've started using it sooner, but before they would grant me access, I had to complete a thirty-minute consultation with one of their staff members to configure my email and learn how to use the product most effectively. That seemed unnecessary to me, so I passed.

I eventually got desperate and agreed to the consultation. I now see why they force this — they go deep on how you use email, do some real-time customizations, and make sure you know how to use the product. All of this makes users much less likely to churn.

That said, it's surprising that Superhuman, an application with thousands and thousands of users, would make this kind of investment in onboarding new users. For a $30 per month consumer email application, this seems like the definition of something that won't scale.

I recently came across an interview with Superhuman's co-founder, Rahul Vohra, where he talked about the importance of these consultations and was asked if he thinks they can scale. He responded by saying that organizations need to identify the things they do that won't scale and decide which of them they should keep on doing. These are things that, from the outside, may seem small and wasteful but are actually core differentiators consistent with the heart of the organization's strategy and competitive advantage.

I've been thinking about this a lot lately. As an organization scales, the things that aren't scaling start to become really obvious. And smart companies find ways to outsource, automate or completely stop doing them.

The hard part of all of this is identifying those things that, on the surface, seem like they obviously won't scale but actually drive big value.

At the Ritz-Carlton, every single employee (even the maintenance folks) has a budget of $2,000 per guest to make things right. On the spot, without asking.

Zocdoc, the medical appointment booking service, sends a $10 Amazon gift card to users every time a doctor reschedules an appointment.

Zappos maintains a 24/7 call center, posts their phone number on every page of their website, and doesn’t have a phone tree.

In the early days, most startups will tend to overinvest in high-touch and high-cost activities. They have to do this because they're forcing their way into a market. They can't cut corners and scale isn't an issue.

One-on-one product training. High-touch recruiting and employee onboarding. Ultra-fast customer service response times. Even small things like sending hand-written holiday cards to every customer. These are obvious and easy to do in the early days. But many of them won't scale and there’ll be pressure to stop doing them over time.

The easy part is dropping things that don’t scale. The hard part is continuing to do them.