Formerly known as Wikibon
Search
Close this search box.

How Hortonworks is weathering the big-data market’s shift away from Hadoop

More often than they like to admit, tech vendors must put on a brave face when they’re groping for a robust growth path.

In the big-data arena, that go-to-market messaging approach is very much in evidence among the vendors who pioneered the commercialization of the Hadoop programming framework for storing and processing very large data sets across many computers. Hortonworks Inc. Chief Executive Rob Bearden, for instance, attempted in a recent interview to put a positive spin on a fairly cloudy outlook for his company’s growth prospects.

The bottom line is that Hortonworks, like longtime rival Cloudera Inc., is still struggling to achieve consistent profitability and momentum. On the positive side, both companies continue to grow revenues, add new customers and deepen their footprint in existing customers every quarter. Most recently, Hortonworks reported total revenue of $75 million in fourth quarter 2017, up 44 percent from a year earlier, with year-to-year annual revenue growth rising 42 percent from 2016, and it forecast strong revenue growth to continue. In addition, Hortonworks closed 20 deals of more than $1 million in the latest quarter, more than twice as many as in the same quarter of the previous fiscal year, with deal sizes continuing to grow and renewing customers even expanding their spending on Hortonworks products.

But nobody’s taking a victory lap. Profitability remains a touch-and-go proposition for Hortonworks. For the Santa Clara, California-based company, the fourth quarter of 2017 was its first cash-flow-positive quarter ever. This is in sharp contrast to Cloudera, which is still racking up operating losses and burning investor cash at a feverish pace even as it forecast revenue growth to slow this year. However, Hortonworks can’t take continued cash-flow positivity for granted, and it has been reluctant to forecast that it will at least break even in coming quarters.

Both vendors find themselves challenged to prove that their business models can succeed long-term. Both have grounded their businesses in open-source software, with specific emphasis on Hadoop and on-premises deployments. What they both face is a market where momentum has shifted to the public cloud and to providers that offer a full suite of big-data analytics offerings plus the infrastructure and platform offerings to address every conceivable enterprise deployment scenario.

Like Cloudera, Hortonworks has seen its stock price knocked around recently by an investment community that harbors doubts in this regard. No big data vendor — not even IBM with its huge enterprise customer base — imagines that it can grow indefinitely from premises-based deployments alone.

Acutely aware of these misgivings, Hortonworks continues to tweak its messaging to suggest that it isn’t being straightjacketed by its Hadoop legacy. Bearden recently told investors that Hortonworks has “evolved from the leading Hadoop provider focused on the big data market to the leading global data management platform company with over 1,300 customers throughout the world.”

Though Bearden boasts that two-thirds of its revenue stream is from on-premises deployments, the company continues to place a major emphasis on public cloud partnerships. Its Cloudbreak offering enables simplified deployment, provisioning and scaling of its Hadoop platform in the Amazon Web Services Inc., Microsoft Corp., IBM Corp., and Google Inc. clouds. IBM resells Hortonworks Data Platform in lieu of its own deprecated Hadoop distribution and has placed HDP at the heart of its Hadoop platform-as-a-service offering. Microsoft has OEM’d HDP and integrated it into the heart of its Azure HDInsight public cloud service. If you’re an AWS customer, you can acquire HDP in that cloud provider’s online marketplace and deploy it with full S3 storage connectivity and a shared Hive metastore.

Even as it struggles to avoid encroachment from innovative public-cloud solutions, Hortonworks has had trouble standing apart from traditional rivals. Its portfolio — anchored by HDP, Hortonworks DataFlow and Hortonworks DataPlane Service — provides a comprehensive platform for unstructured, streaming and edge analytics in hybrid clouds. But taken as a whole, it’s not head-and-shoulders differentiated from the equally sophisticated Cloudera Enterprise Data Hub or MapR Converged Data Platform. The range of enterprise-grade functionality — cluster management, data governance, query optimization, policy management, security and the like — that they all offer in these portfolios are simply table stakes in the enterprise market, not competitive differentiators.

Likewise, these vendors’ growing emphasis on industry-specific and horizontal solutions for data warehousing, cybersecurity and other business applications are essential for growing their share of pocket among established customers, but it won’t dislodge the public cloud providers from these and other growth opportunities going forward.

Even their growing emphasis on providing data science DevOps tools is not necessarily going to buy them much share in the burgeoning artificial intelligence market. Considering the depth and breadth of AWS, Microsoft and Google investments in AI, deep learning and machine learning, legacy Hadoop vendors such as Hortonworks are in danger of becoming marginalized in this arena as well.

Hortonworks remains hamstrung in the AI market by the fact that it’s merely a reseller of one IBM tool, Data Science Experience. Cloudera has a stronger competitive position with its Data Science Workbench, but even that faces formidable competition from AWS, Microsoft and hot startups that are providing level of machine learning automation conspicuously missing from Cloudera’s tool.

For perspective on how the IBM partnership fits into Hortonworks’ strategy, here are Bearden and IBM’s Rob Thomas interviewed on theCUBE at Big Data NYC 2017:

 

Book A Briefing

Fill out the form , and our team will be in touch shortly.
Skip to content