Cloudera CMO Mick Hollison joined Computer Business Review to talk company focus after a $5.2 billion merger with rival Hortonworks.
All Cloudera projects will be 100% open source, you’ve announced. Several other traditionally open source service providers are doing the opposite and making their licences more restrictive. Aren’t you worried that public cloud providers will eat your lunch?
It was core to the merger talks from the beginning that this was something that both executive teams wanted to do. However, being open source, and which licensing mechanism you choose for your open source software are two independent thoughts.
The fact that we’re going 100 percent open source, does not mean we may not in the future change our licensing structure. The reality is that the large cloud infrastructure providers, as it stands today, can very easily just take the open source bits – whether they do or do not contribute them back to the community – and compete vigorously.
So we are evaluating licensing alternatives independent of the fact that we’re going to open source the entirety of the core platform, if that makes sense…
Is there enough urgency in your move to bring a unified cloud data platform to market later this year? Competitors are moving fast…
As a marketing guy I’m never satisfied with how fast a product comes out.
This is not a far distant project however; you’re talking about something that’s going to hit in the next two quarters.
Considering that these were two independent companies, bringing software together inside of six to nine months to me is extraordinary.
Post-merger your companies are providing over 30 open source-based products and keeping both Hortonworks and Cloudera iterations of tools. Are you focussed enough?
It’s certainly a very valid question. I think it’s more focused than it’s ever been. This whole galvanising concept of an enterprise data cloud is really pulling it all together.
Once all of the open source components are delivered as cloud-based services, you quit worrying so much about which open source project you picked; you let the customers make those decisions. And honestly even for them it’s delivered as a service.
They just know that it’s a service that provides them with a data warehousing capability a service that provides them with a data engineering or data pipeline capability.
Does not that spread your development team a little thin?
One of the great benefits of the open source community is that you are benefiting from the great work of others on other projects.
So it’s not as if we’re the only or even the leading contributor to many of those projects. We just had IBM on stage. They’re massive contributors to a number of the core open source projects.
We benefit from the innovations that are delivered by their committees to those projects.
To your point, it is about thoughtful selection which ones we place our own engineering emphasis on, versus which ones we benefit from the community on.
The markets have treated Cloudera brutally over the past 12 months. What will it take to regain confidence?
It’s really simple. It’s just numbers and execution. The markets are waiting for us to show long term top line growth with great operating cash flow margins.
It is our goal in FY21 to be at a 20 percent top line growth number. And by the end of FY 21 beginning in the FY22 to be producing 15 percent operating cash flow margins.
Should we do that, that would make us one of only nine enterprise software companies on the planet that have those particular financial characteristics.
And those companies have valuations that are two or three times the current multiple that Cloudera commands in the marketplace.
The enterprise data cloud strategy is central to realising that vision and getting to those growth and profitability metrics.
What are your priorities to make that happen?
Some efficiencies are already baked in. Others are in play.
For example this year in fiscal 2019 we’re going to we’re going to keep having Data Work summits; we’re also going to keep having Strata summits.
In FY20 we won’t do that any anymore; we won’t have the same level of event marketing spend.
Similarly on the engineering side there’s places where, to your point, you need to double down on a given project and there’s other places where we may back off a bit.
I wouldn’t be telling the world a secret to say, look, it’s getting more competitive out there.
The house offerings of the big public cloud vendors are more competitive than they’ve ever been in the past. And there is certainly a rise of single function vendors out there as well that offer just one analytic capability; so they just offer data warehousing or they just offer machine learning and data engineering.
We are a platform company that’s about delivering multifunction analytics.
We think getting to this enterprise data cloud vision will put us in a very good competitive position where we’ll be able to command better pricing than we have historically.
Who do you see as your main competitors?
AWS’s EMR and Redshift are very much competitive offerings to what we have in the field.
Microsoft has some competitive offerings. Google has some competitive offerings and then in a single function world… you mentioned Databricks; I would put Snowflake in that same kind of category.
That said, we’re the standard bearer for specifically what we do: we’re really the only major multi-function analytic vendor out there that runs across multiple public clouds.
And while the offerings from Amazon and Google are good in their own right, Amazon is not in a hurry to start running their workloads on Azure and vice versa. That’s where we have an advantage.
What’s your sales focus?
We’ve identified around two thousand accounts that we want to go chase. We’re going to go seek additional opportunities in the telco and automotive and life sciences spaces, because we’ve already got a very strong footprint in those at the top end of those industries. So we will continue to target companies in those industries that have a great demand for the kind of offerings and solutions that we bring to bear.
The other thing that’s probably worth noting is that it’s a little bit unusual in the marketplace. The bulk of our growth. From a bookings perspective is in expanding our existing accounts, as opposed to necessarily generating brand new ones. The great news about the merger is a cross-sell opportunity for our salespeople
Your head of AI said very clearly that you are “not a services company.” Data science is hard though, and a lot of customers don’t have capacity. Is offering more of a managed service on the cards at all?
There are a couple of different couple different ways of answering that question. First of all at the strategic and in terms of what problems are you’re going to go try to solve with machine learning and A.I. that is where Hilary [Mason]’s previous company that Cloudera acquired, Fast Forward Labs, steps in.
So Fast Forward Labs is great at strategy and and also with building out functional. requirements and even delivering snippets of code for people to get started with for solving well-known challenges. The next click beneath that is offered through our pro services organisation; that’s more of the data pipeline cleanliness and sort of the plumbing part of data science which is super important.
People forget that there is no AI if you don’t have a clean data lake and you don’t have well organised data; and in the first place or our professional services team can help with that. There are managed service providers out there that are doing work with us today (also). There’s one for example Cazena.
So there are a number of third party vendors that are offering managed services not just for CDW but for the totality of the Cloudera platform. How much of what we decide to do ourselves versus what we let our partners deliver, is something that we’re debating in-house right now.
Is it likely that you will do a lot more of that.
Would you say “I’m not at liberty to say”?
It’s where the money is…
There you go!