Built on AWS, Zendesk Sunshine is a new take on enterprise SaaS and CRM

One of the most interesting — and so far least discussed — aspects of Zendesk's recent launch of its Sunshine CRM platform is the way that it is built on Amazon Web Services (AWS). This is an example of a new class of enterprise application platforms I see emerging that are built on a serverless architecture. This is a big move by Zendesk, and one that points to larger trends in the evolution of the enterprise CRM space.

First let me fill in some details. Zendesk is a fast-growing SaaS vendor specializing in customer service and support ticketing. Founded in 2007, it's now established as an up-and-coming competitor against Salesforce Service Cloud in the midmarket, with $600m in annual revenues and more than 125,000 customers, from Airbnb and Rovio to Stanley Black & Decker and Tesco. Its core Zendesk Support offering is a popular choice among digital businesses and others with sizeable customer service teams who need a modern, flexible, customer support and ticketing solution.

Expanding Zendesk's CRM reach

At its annual conference held last November in San Francisco, the company made three announcements to expand its reach into new territory within the CRM space:

• Zendesk Sell — marking Zendesk's entry into the sales force automation (SFA) market, Sell is the product formerly known as Base, which Zendesk acquired in September. An easy-to-use, mobile-first SFA tool with an existing global customer base of 5,000 small-to-medium-size businesses, ZenDesk Sell already has close integration into the core Zendesk Support product, and is to be integrated into the Sunshine platform.
• Zendesk Explore — adding data analytics into the Zendesk family, Explore brings data together from within and outside of Zendesk to help teams monitor and improve the customer experience. It provides pre-built and custom reports that are easy to share with colleagues across the business.
• Sunshine — the new CRM platform that ultimately brings all this together, Sunshine is built on Amazon Web Services infrastructure and opens up Zendesk's underlying data model and event model for connection using the AWS environment. It is able to use AWS services to connect to other data sources, and businesses can build their own applications on AWS that work with Sunshine data and events.

Built on AWS services

As I found out in conversation with Zendesk's VP of Product Sam Boonin a few weeks after the launch, Sunshine was born out of Zendesk's own journey to migrate its SaaS infrastructure to AWS. This wasn't a straight 'lift-and-shift' of the existing code onto Amazon servers, he says. Instead of building its own data store, event bus and analytics tooling, Zendesk made full use of AWS services such as the Aurora database, Lambda serverless functions and Kinesis data streaming and analytics. That led to the decision to launch Sunshine, he explains:

One thing we realized is that [building on AWS] actually unlocked a lot of developer productivity for us ... We had this epiphany moment — why don't we offer that same flexibility to our own customers?

Sunshine connects natively with a raft of AWS services, including Kinesis, S3 data storage, the Redshift database, SageMaker machine learning, and Amazon Connect, AWS’s cloud-based contact center service. The data flows across a direct, secure, private AWS network connection.

Flowing in and out of Sunshine

This means that all the customer data in Sunshine, along with other data from the Zendesk applications such as knowledgebase content, usage metrics and agent productivity, sits on the same cloud platform where many businesses already run other digital products and systems. Data and events can flow freely between Zendesk and other AWS endpoints — with no rate limits, adds Boonin, in a sly dig at Salesforce, which restricts data flows into and out of its cloud platform. It's important to customers to enable this integration in an open way, he explains:

Customers want to remove the friction of having to integrate into us.

The Sunshine data platform has three elements. Profiles hold a single, unified view of each customer. Events capture customer activity in a historical timeline — anything from customer service interactions and website visits to purchase transactions and shipping history. Custom objects allow companies to bring in useful information from other systems, such as products owned, equipment rented, or data streamed from connected products and sensors.

A fresh approach to integration

The announcements are part of a move upmarket to compete with incumbents such as Salesforce, Microsoft, Oracle and SAP, while taking a fresh approach to integration, says Boonin:

They're experts in extracting money from their customers for things like [integration]. We've always been easier to use, set up and understand. We've always had a relatively large cost advantage and we're leveraging what our strengths are.

Those strengths stem from Zendesk's pedigree as a newer generation of cloud vendor, he believes:

Philosophically, companies like us — born of the public cloud, with restful interfaces, well after open source had become mainstream — we don't think of vendor lock-in, we think of openness.

We have thousands of customers that use our prebuilt integrations ... If people use a different SFA tool or support tool, that's fine. We don't make that assumption. We're not SAP — we're not thinking in terms of being a monolithic stack, that you have to use all of it. That's not how modern software companies operate.

Convergence of sales and support

Expanding to include SFA — and also engagement and marketing through Zendesk Connect, which launched in May — chimes with the growing convergence of sales and support that's emerging in the digital enterprise. Businesses are building continuous relationships with customers in which customer service will often be involved in renewal, cross-selling or churn prevention, while closing a sale often means resolving support issues. As Boonin explains:

Customer support doesn't always happen at the end of the relationship — we need a platform to help companies understand where customers are in their lifecycle. It needs to be much more open, [where you] put all your data in and understand your customer.

Concepts such as the digital customer journey will accelerate convergence within CRM — Zendesk certainly doesn't expect its product to mirror the separate 'clouds' for sales and service that Salesforce offers, says Boonin:

We think they're going to merge together in very interesting ways ...
We don't think CRM is going to look like it does now in five years.

My take

I'd been meaning to catch up with Zendesk for quite some time — I've known the company since inviting founding CEO Mikkel Asger Svane to speak at a European SaaS conference in 2009. But the timing of doing so now couldn't be more fortuitous.

With these announcements, Zendesk is putting itself at the forefront of two very significant trends. The first is this convergence of sales and support, which at diginomica we see as characteristic of The XaaS Effect in which businesses use digital connections to build a continuous relationship with customers rather than the old-school pattern of sporadic one-off sales. As I wrote last year in the context of Salesforce's evolution:

Rather than being an isolated encounter with a salesperson, the customer experience becomes a longer-term relationship that’s nurtured by marketing and sustained through customer service.

The second trend is at an even earlier stage. This is a refactoring of enterprise applications into collections of serverless functions and resources that allow for much more agile adaptation of new information flows and functions. While larger vendors such as Salesforce, with its acquisition of MuleSoft, are taking steps in this direction, Zendesk is one of a new wave of vendors that are taking a more radical path.

The decision to build its new CRM platform on services within the AWS infrastructure is especially interesting. Instead of attempting to build its own platform ecosystem, Zendesk is piggybacking on AWS's. This makes sense not only because its customers are already making use of that universe, but also because it can then leave AWS to build many of the capabilities such as data streaming and machine learning that Zendesk's larger competitors are investing extra cost and resource in building for themselves. This is a new take on enterprise SaaS that has intriguing possibilities, which I'll explore further in a follow-up post.