Everyone says that the next million cloud customers are coming from enterprise IT. It’s the narrative that all of the major cloud platforms have been saying for a while — to the point that I started accepting it uncritically. I even recently wrote about those next million enterprise IT customers.

Oof. No! That’s not how any of this works! My job is to think critically about the things “everyone” says and to rebut them. My apologies. I will do better!

Let’s be clear: There’s a huge market of enterprises still to migrate to the cloud. The trouble is that it’s an inherently limited market in many ways.

A company founded today that grows to become a giant enterprise is likely going to have its technology stack built inside of a cloud provider’s environment, not in a data center it owns. This is a trend that will only continue. Throughout most of the industry, there’s already a shift away from needing justifications to use public cloud to instead requiring justifications not to.

Where does this enterprise customer of the future leave AWS?

As AWS continues doing its thing of launching purpose-built managed database offerings, it only makes this harder. Today, AWS’ product page lists a dozen different databases from which customers can choose. As those offerings continue to expand, what happens when there are 30 or 40 of them? Someone setting out to build a new application on top of AWS is going to have any number of skills that they bring to the problem, but “disambiguating between a bunch of AWS services that all kinda look the same” is extremely unlikely to be one of them.

AWS has to keep an eye toward future and growing trends. It should be stimulating more applications to be built on AWS; pure infrastructure plays will no longer be enough.

Let’s play out a completely random example

Let’s assume that AWS has internal data that demonstrate its offering is super effective at helping customers build out email newsletter creation and delivery systems.

Today, AWS talks about how Amazon SES is used to send email and Amazon Pinpoint is an “outbound and inbound marketing communications service.” Those are necessary parts of the bits that send email, but customers looking to build the next ConvertKit or Substack or Revue aren’t going to care about only those things.

They’ll need services that help them build out the website that lets customers manage their subscriber lists and information. They’ll need help building something from the ground up that complies with GDPR and other similar regulations. Most notably, they’ll need insight from a company that specializes in this space to figure out what challenges they’re likely to have as they find success and scale out. Those insights are easy to implement early on in the planning stages; they’re way harder to apply once you’ve got a working product in customer hands.

How AWS could actually empower future enterprise customers

Today, AWS’ “investments” in startups with ideas like these look a lot like a pile of free credits and an account manager who has the shockingly broad remit of “startups.” What would an actual AWS investment in a customer profile like this represent? As I see the industry, it would look more like connections to AWS customers in orthogonal spaces who can speak to some of the market’s complexities, an articulation of which services solve which problems for a company like the one the customer is building, and introductions to their first 10 likely customers.

Systems would have to be built up to support this kind of insight, and AWS would have to face some hard truths about the markets it’s in versus the ones it only thinks it’s in.

The entire AWS sales and technical field teams would need to learn to have very different conversations with customers. But consider how unworkable the alternative is. If today’s sales and marketing motions continue doing what they’re doing, customers are going to continue to be barraged with feature and service enhancements, a periodic partner recommendation that looks an awful lot like it might compete with what the customer is building, and a growing sense that AWS is and remains an “infrastructure company” instead of a company that builds services that empower its customers but can’t articulate how it might do so.

But is it possible?

This isn’t merely a story about theoretical email newsletter creation companies. This model would apply equally well to companies setting out to build CRMs, phone systems for customer sites, next-generation manufacturing floors, smart home appliances, and oh-so-much more.

The tools to build these kinds of products and businesses already exist within AWS. The trouble is that the only people who truly understand how those tools all fit together to do that are the customers who’ve already built those things with them. AWS needs to understand its customers in new ways and get away from the current pattern of “spinning up a Linux VM in EC2 requires at least basic awareness of a dozen distinct AWS services.”

AWS risks irrelevance if it ignores the actual next million cloud customers

The baseline reality is that customers are going to demand something that encapsulates all of the infrastructure building blocks and lower level “plumbing” into something much higher level. If AWS doesn’t give it to them, others absolutely will.

The long-term threat here is that AWS becomes similar to one of today’s large bandwidth providers, such as CenturyLink or NTT. Every company depends upon them, but few people are aware of them in any realistic sense. If this happens, then it turns out that customers’ margin will, in fact, become some other company’s opportunity.