After filing for an initial public offering last month, Hadoop provider HortonWorks is ending 2014 with a bang: The company’s stock had an enthusiastic reception on Wall Street and by the end of the first day had a valuation of more than $1 billion.
After making its debut with an IPO price per share of $16, HortonWorks’ stock soared to more than $25 in the first day of trading and has stayed above $20 per share in the intervening days. And it was in good company: Real-time analytics and mobile app management company New Relic made its IPO on the same day on the New York Stock Exchange and also saw more than 40% gains on its first day. Both companies are now in a coveted club of $1 billion startups, showing the high expectations for big data and analytics companies.
The HortonWorks IPO may be the latest and most visible advance of Hadoop, Apache’s in-memory database system, but the technology has been advancing, both in adoption and venture investment, throughout 2014 and is likely to gain even more steam in 2015. HortonWorks’ CEO Rob Bearden told Gigaom this month that he believes Hadoop is “ready to explode.”
A few recent developments in Hadoop-related investment include:
- Altiscale, which touts itself as a purpose-built Hadoop-as-a-Service offering, closed a $30 million series B funding round led by Northgate, bringing its total funding to $42 million.
- Qubole, another HaaS solution provider started by the creators of Apache Hive, recently raised $13 million to add to its coffers.
- Tel Aviv-based XPlenty, a Hadoop provider, which also has offices in San Francisco and uses public cloud services such as Google and Amazon Web Services to run its Hadoop offering for as little as $100 per month, raised $3 million.
- One of the largest big-data funding rounds of 2014 was the $900 million raised by Hadoop provider Cloudera in June, which followed a $160 million funding round just a few months earlier.
Bearden isn’t the only one predicting enormous growth for Hadoop and the big-data market. ABI Research has projected that global big-data spending will reach $114 billion by 2018, reflecting a compound annual growth rate of nearly 30%. Notably, the company pointed out that about half of yearly spending will be taken up by salaries as businesses seek to hire big-data specialists, as well as professional services for businesses with a lot of data, but no in-depth understanding of how to manipulate that data and apply intelligence to it. In addition, the HaaS market alone is predicted to see substantially more growth than on-premise implementation as big data and cloud services cross paths: Allied Market Research estimated earlier this year that HaaS could reach sales of $16.1 billion by 2020 after doubling its market size last year.
One of the central challenges to widespread Hadoop adoption will be businesses’ ability to monetize the analytics technology, according to Ashish Gupta, CMO and senior VP of business development at Actian, which has an analytics platform that runs natively in Hadoop.
“You can adopt (Hadoop), but still not be able to monetize it,” Gupta said. “That’s where the ecosystem becomes really important.”
Evidence points to the ecosystem developing rapidly as well. Intel has been a strategic investor in Cloudera, and that $900 million funding round was accompanied by the addition of Intel CIO Kim Stevenson to Cloudera’s board of directors. Cloudera has been steadily adding to its lengthy list of partners as well as extending some of those partnerships throughout the year – along with gaining expanded global presence with new offices in France and China during the fourth quarter. HortonWorks’ major partners include Microsoft, Teradata, Hewlett-Packard, SAP and Red Hat, and the company touts relationships with more than 500 consulting companies and systems integrators to help companies with their Hadoop implementations.
“Hadoop is here to stay,” Gupta said. He said that Actian sees significant growth in Hadoop implementations moving from proof-of-concept to production. He noted that companies are increasingly able to create reservoirs of data accessible at a low cost, due to Hadoop – making it cheaper to store data than delete it, because those large swathes of data can be analyzed for more accurate business intelligence. Gupta said that for a retail customer, traditional data storage infrastructure might have stored several years’ worth of data from point-of-sale transactions, but only a few weeks could be analyzed.
He said there are particularly valuable use cases for telecom operators. Being able to add more data sources than CRM records reveals a deeper view of the customer experience – such as correlating dropped calls with location data in order to pinpoint which customers are likely to be unhappy with their service and therefore more likely to churn. Gupta said that Actian worked with a European operator that previously offered across-the-board discounts to customers who were approaching the end of their contracts, but that unhappy customers often churned anyway and customers who remained paid less. By adding analysis of log data, device data and call detail records, as well as GPS data where available, operators can offer more targeted offers as well as get better insight into their network operations, he said.
The IPO of HortonWorks, he added, is a “huge validation” of the mainstreaming of Hadoop and a recognition that both enterprises and investors see the value in the technology.