No, You Don’t Own The Customer Anymore — The New Age of SaaS Competition

Much has been written by VC’s about building “full stack startups,” “owning the customer with a strategic product” and finding the “killer app to build a platform on.”

One interesting phenomenon we’re seeing is that an increasing number of users don’t need or want you to be “full stack” at all.

Actually, at the extreme, that’s pretty much the opposite of what developers want — which is something a lot more modular (preferably open source) that solves their specific problem well yet offers the flexibility of a diverse, transparent and rapidly innovating ecosystem of components: consider Docker’s batteries included, but removable philosophy.

This is not limited to developers.

If I’m any knowledge worker, I can find and begin using Quip and Dropbox on my own. If I’m handling customer success, I can use FullStory to replay a user problem and Intercom to communicate. As a marketer, I can A/B test my website with Optimizely, slice and dice my mobile events in Amplitude, and build richer cross-channel profiles with Lytics. As a salesperson, I can use ClearSlide to present to potential customers, and RelateIQ to keep track of my meetings and customers.

And those just scratch the surface — the long tail is a lot more interesting (and often a lot more risky!) than that. The average company (per SkyHigh’s quarterly cloud adoption report, based on data about 17M users) uses 923 different services delivered via the web. This is > 10x more than IT expects. The average individual employee now uses 28 cloud apps for work. They don’t want to wait for their technology team to tell them what limited set of technology they can use to do their job.

Consumers / employees want to assemble their own application stack, just like developers do.

Other companies report similar app ecosystem diversity. Okta has over 4,000 application partners, and their average 2014 deployment included 14 identity-managed apps per user (up from 8 in 2013).

SkyHigh and Okta are already helping to solve the visibility, security and accessibility problem for SaaS. However, companies also need supporting data infrastructure. Without that, this Cambrian explosion of individually and departmentally-adopted apps and services creates information silos and an ongoing integration burden for the technology team.

In other words, a proliferation of web services without a rational integration plan results in just another untenable compatibility matrix from hell that results in lost productivity, useless and duplicative engineering effort, and vendor lock-in.

Significant companies offering “enterprise integration platforms” have been built, from Mulesoft to SnapLogic. But one of the things we learned from Docker is that not all customers want the entire proprietary Platform-as-a-Service, given the choice. They’d rather have the right (open source) spec and a single standard interface — and they are quite inventive about what to use and build on top.

We’re seeing a lot more companies pursuing this vision, in different areas — companies like Segment, which is focused on building a lightweight customer data service to support productivity in a paradigm of application choice. The success of services like this will enable companies to use best of breed applications for every need, while maintaining user ownership of consistent data, from across their different platforms, available across all their separate systems.

No, software vendors don’t own the customer any more.

They can’t rely on their deeply integrated suite of products to lock the customer in. Developers were the first to break out, but we think the rest will follow, and this is a good thing.

Let a million flowers bloom. Let us find rational architectural points to see them, secure them, enable them, and bring them together in open but integrated ecosystems. And let us be freed from the shackles of bad software.