Talend CEO - IT must get unstuck from a legacy cycle to turn data into a business asset
- Summary:
- Talend Connect brought heady conversations about avoiding cloud lock-in and realizing a multi-cloud, serverless potential. But as my conversation with CEO Mike Tuchen shows, it's really about a practical way to fund innovation within IT budgets - and deliver data assets to the business.
In the flurry of spring events, Talend wins my award for the most buzzword-compliant announcement, with their unveiling of a demo for "multi-cloud serverless containers." They also nabbed second place with the "Industry’s First Cloud-Native, Apache Beam-Powered Streaming Data Application."
Fortunately, Talend's CEO Mike Tuchen took my buzzword award in good humor during our recent sit-down at the Talend Connect user conference. That's good, because behind the catchphrases, Talend's event was as provocative and important as any I saw this spring.
Why? Two reasons: Talend puts the nitty-gritty of data transformation and portability into focus. Digital transformation isn't happening without better/cleaner/faster data. And without data portability, cloud is just a new flavor of lock-in.
Talend is also part of a loose alliance of companies - a multi-cloudy/DevOps/containers movement if you will - that make a compelling case for next-gen IT within the context of existing IT budgets.
Customer results from modernizing IT are coming in
As I wrote in my Talend use case on how TD Bank avoided data swamps:
The Talend Connect 18 keynotes pushed a heady new phase in next-gen applications, where portable/cleansed data (via Talend) and portable apps (via containers and orchestration a la Docker and Kubernetes) allows IT shops to avoid futile fundraising by moving from (expensive) legacy IT to cost-effective, modern infrastructures.
The customer results are coming in. My Talend Connect interview with Docker CEO Steve Singh was a telling indicator of this shift, because Singh hails from the business applications side (Concur, then SAP). See: Can IT finally deliver innovation without busting its own budget? Docker's CEO says yes. During my chat with Singh, he cited some hefty customer numbers to illustrate the "innovate from within" IT push:
MetLife is doing exactly that. They’re taking their legacy apps, putting it on Docker, running it on modern infrastructure. They’re saving tens of millions a year, and they’re using that to actually drive next-gen innovation. So the next-gen innovation is entirely self-funded. They haven’t gone back to their CEO and said, “Look, I need more budget.”
During his Talend Connect keynote, Tuchen shared similar customer stories, including Lenovo:
Tuchen said that in 2012, Lenovo had a "modest analytical setup." They realized for the next stage of their growth, that wasn't going to cut it. They started by moving to the cloud - with Amazon Redshift. They analyzed a variety of data streams in Tableau. In the process of working with Talend, Lenovo got their run time up 50 percent at 1/3 the cost, as per the slide.
It wasn't just about IT efficiencies. As they progressed into predictive, Tuchen says Lenovo was able to deliver to the business a 12 percent bump in per unit sales through personalized offers to customers - promotions that leverage data on what other similar customers have bought in the past.
But new data challenges require scale and dexterity
But Lenovo's next stages point to the challenges. Lenovo is now dealing with a petabyte of data. They have moved into Google TensorFlow for machine learning, and AWS Fargate for serverless containers. As Tuchen says:
Lenovo changed their tech approach multiple times in the last half dozen years, and they are just getting started.
To make those moves, companies need a lot of data dexterity - not to mention a better ability to move between clouds than most companies have now. Tuchen believes these challenges add up to "completely disruptive data economics."
All of this is driving a set of requirements is dramatically outstripping what IT is able to do; they are stuck in a legacy cycle.
The new data economy undermines the IT status quo
A skeptic might question what's different about today's data economy. After all, the data warehousing movement traces back to the 90s, and analytics goes back much further. Tuchen sees a "confluence of events" - the biggest inflection point since the origins of computer data economics 48 years ago:
Cheap sensors, ubiquitous networks, cloud-based data processing, real-time data streaming, data-driven employees taking advantage of data in a self-service capacity, and, last but not least, machine learning.
But Tuchen doesn't sugarcoat the obstacles ahead. Companies must get out from under this data disruption if they want to use it to their advantage. As he told the Talend Connect audience:
- 47 percent of data has integrity issues (and half of most corporate data is inaccessible).
- 80 percent of an analyst's time is wasted preparing data.
- Data is doubling every two years, adding in cloud data, social data, and a vast array of data types.
- A 3x growth in self-service users keeps data administrators on their toes, including huge security, compliance, and privacy concerns.
Tuchen says we are driven by these fundamental questions: "Where is my data? What is my data? What's happening to my data? Is my data correct and secure?" And finally, the most important:
In this world of accelerating change, how do I make sure I don't get locked into any one technology, because [technical innovations] are happening all across the world? How do I harness all of that innovation industry-wide going forward?
Practical issues of hybrid clouds must be addressed
That points directly to the need for data portability and so-called "multi-cloud" flexibility, which Talend believes it has good answers for. But that provokes another crucial question: can you do this on-premise, or do you have to move to the cloud to achieve these results?
Tuchen sees the future in hybrid cloud terms, but when it comes to Talend's own customers, the cloudy nature of the roadmap is clear. So, I asked Tuchen, how is Talend managing that with their customers? Tuchen:
Almost every customer we talk to is either already in the cloud, or they know they're going to the cloud. To every single one of them, we say, "Our functionality comes to the cloud first. Then in the next on-premise release, you get to pick it up." They'll often say, "Okay, I understand that. I'm on the way."
What advice does Tuchen have for large enterprise customers, most of which are managing a hybrid reality?
What we're talking to all of our customers about right now is whether they're on-premise, in the cloud, or some mix of the two. Most large customers, to your point, are starting on-premise, and they're just now starting to move to the cloud, which means they're going to be hybrids for many years to come. That's just the world that we live in.
And how do you solve these data problems in a hybrid setting?
To me the choices they need to work through are: where is their center of data gravity today, and what's their intention? Where do they want it to be over the next couple of years? Where is their center of gravity in the cloud? Is it in one of the clouds? Is it several of the clouds, and why and why not?
As we have those strategy conversations, we can say, "All right, here's how you solve your data problems with that in mind."
There isn't one way to solve this:
It's not one-size-fits-all. We've worked with some of the largest banks in the world where their center of gravity is very clearly on-premise and they're experimenting with the cloud, but they're going to be mostly on-premise for a long period of time.
Tuchen contrasted this with a couple of healthcare customer that are moving aggressively to the cloud:
Healthcare is a very compliance-oriented industry. They tend to be relatively slower moving. Now we have two of them that are multi-million dollar customers for us, which means they're spending far, far more with the big cloud players as they've moved enormous parts of their infrastructure out there. For them, the architecture has to start with a cloud-centric approach, where premise is an adjunct to it as opposed to the reverse.
The wrap
Talend just announced their serverless offerings at this year's show:
So, in a sense, Talend is still disrupting themselves to see how far they can take this. But what matters is not some type of serverless idealism, but whether customers are deriving value they can use now. During an AstraZeneca customer presentation, this line stood out:
We now have high-quality data that is in constant demand.
Sounds like a good target to shoot for. Tuchen told me Talend is moving at a 40 percent growth rate for 2017 and 2018 to date, and "well over 100 percent" growth in their cloud solution. Given that Talend is heading towards 2,000 customers or so (their last public customer statement was 1,600 customers a year ago), there should be plenty more stories to tell. I plan to write up more Talend use cases myself, and see what can be learned.
End note: here's the Talend multi-cloud/serverless demo announcement slide: