I keep going back to NoSQL vendors because they keep delivering customer stories.
Which brings us to my recent chat with DataStax CEO Billy Bosworth. Considering the leadership turnover at several NoSQL upstarts, Bosworth has bucked that trend. Next month counts seven years and counting for Bosworth at DataStax. Bosworth was amongst the first twenty Datastax employees - now the count is 500.
Last year, my colleague Derek du Preez reported Bosworth's intentions to transform DataStax into a business-focused company, via a customer experience agenda (DataStax CEO launches new CX strategy - focus shifting from tech to business). So how's that going?
We haven't stopped selling the technology, and we certainly haven't stopped talking about the technology. But if you look at the mainstream enterprise, which is our target, our customers are faced with a problem.
CX dilemmas - building cloud apps isn't a cakewalk
Building real-time cloud apps is harder than it seems:
[They tell us], "Okay, I get it fundamentally. I need to start building a new classification of cloud apps that has a 'right now' characteristic, but where do I begin?"
That means advising on more than speeds and feeds:
It's a very big transition for a lot of these companies. For us to come in and talk distributed databases is really good if you're a very technical person, but what customers need help with is understanding: "How do I apply that? What use cases make the most sense?"
In our world with enterprises, we think hybrid cloud is the model for the next 10-15 years. The faster we can help our customers to understand how to leverage that hybrid model for these right now applications, that's a big mission for us.
Bosworth talked about a bank that's taking on retail:
Macquarie Bank's headquarters is in Sydney. They're one of the largest investment banks in Australia... Now they're moving into retail. They gave a comment I thought was one of the most enlightened comments any enterprise can make. They said, "We're not competing so much against other banks. We're competing against the other applications on your phone."
But it's not always clear where a CX overhaul should begin. So what does DataStax advise?
Well, first of all, I need to know all about you - we call that a customer 360 project. Then I need to give you a very individualized experience. That's called personalization and recommendation. Then I need to make sure you're okay and safe, so I've got to have anti-fraud. All these fall under that bucket.
DataStax provides this type of collaboration and through its customer success group. But Bosworth wants to avoid the services clichés:
"Customer success" is not just a nice way to give a new name to services. It is very much a mindset and a model that says you have to really understand what your customer is trying to achieve.
That advice also means architecture planning, tying DataStax into an array of tools and vendors, from the storage layer to the security layer to the middleware layer: "How we interact and engage with our partners is all very important."
So is this a revenue play for DataStax, or is it about solidifying the customer relationship and making sure the projects deliver? Bosworth says it's very much the latter. Without opening the entire financial kimono, he offered this:
We don't share a lot of financial information. One thing I can tell you is our gross margins run north of 75 percent - that's our blended gross margin as a company. That's really how you can figure out if a company is a services company or a software company. Certainly anything upwards of 70 percent puts you in the software category. But for us, we really stay focused on how do we make the customer successful as quickly as possible... kind of time-to-impact if you will.
Rip and replace isn't the right NoSQL benchmark
One thing I find bothersome is the emphasis on "rip and replace" as the benchmark for NoSQL's legitimacy. Yes, NoSQL vendors have been known to fan these flames, bragging about relational database replacements.
But as a general rule, I've found NoSQL's value to be in pushing ahead with next-gen projects - without disrupting core transactional systems. Perhaps we wind up with a subtle form of rip-and-replace in the end, but if so, it's a very slow burn as cloud apps achieve critical mass. Turns out Bosworth has a similar view, informed by his own humble pie over the demise of the mainframe:
When I was 22 years old, and fresh out of school with my shiny new Computer-Sci degree, I was running around telling everybody, "The mainframe is dead. We're going to kill the mainframe. We're all going to re-write these applications." And you know what? The reality is we didn't re-write almost anything. We did a bunch of new stuff on a client-server, and then there was a couple of things we had to re-write because we had this little deal called Y2K, where the world was going to end.
He sees that dynamic now with NoSQL. Still, NoSQL has advantages, especially for cloud apps:
You also have to be true to your technology... We kind of deal with the same general workloads as relational databases, but we do it better, and in a more unique way. Schema-less, for example, is a big deal. We manifestly take on workloads that are just not possible on a relational architecture.
DataStax calls real-time cloud apps "right now applications." Bosworth believes a core component of a"right now" app is its architecture is very highly distributed. He wants to cut through the false hype of pseudo-distribution:
We've been saying that we distribute things for years, but the reality is we haven't. Not really at scale for full read-write capabilities.
But that doesn't mean you rip out the relational:
I tell customers all the time, "Hey listen. You've got enough to worry about in terms of writing your new applications, making them successful, figuring out how to get these new engagements with customer experience... Don't try and replace anything. If it's not broke, don't fix it."
So leave your relational environments the way they are, run them probably for another 10-15 years, do whatever you want. You might move them to the cloud, you might move Oracle to Aurora, you might move SQL Server to Oracle. I don't know. But whatever, it's still relational.
My take - enough with disruption, let's help somebody
Bosworth and I talked about the limits of disruption as a sales pitch. Yes, customers want to make digital moves - without having to throw out their entire landscape.
He brought up another DataStax customer - Babylon Health. With over a million members, Babylon Health brings personalized health advice to consumers via mobile devices. That means tackling serious security and UX issues. Babylon Health has now expanded from the UK to Rwanda. They use an AI bot to relieve health bottlenecks:
Jon, if you don't feel well, and you suspect something might require a doctor visit, you can go through the very laborious painful healthcare system to get to a doctor live, be very frustrated all day, or you can take out this application and start chatting with a bot. You know it's a bot, but it's going to help navigate you via the AI to the right consultant. Then you can interface with a doctor or healthcare professional via text or video messaging.
This is how we live our lives everywhere else now; this is how we need to start living our lives for healthcare. And so for them, being able to come in and give that health advice in an instant, with an underlying data technology that they know is going to be able to scale whatever needs they have, that's the kind of impact and change that just, again, fundamentally not possible with the old technologies.
That's the motivation that drives Bosworth forward:
We use this tech for our video games, and taking pictures of our food. Let's get to doing it for stuff that matters to society, right? That's where I get super excited.