The Data Economy is an occasional analysis column by Wikibon Senior Analyst Jeff Kelly covering the business of Big Data.

Any start-up trying to commercialize Hadoop as their main line of business has two important challenges to tackle:

 .

1. Deciding on and executing a business model (open core, services only, largely proprietary, etc.)
2. Building a world-class company, from sales and support to marketing and HR.

 .

At Cloudera’s first-ever analyst meeting last week, the company spent an entire day educating the analyst community as to their progress in tackling both challenges. Here’s what I learned.

Doubling Down on Open Core

.

From a business model perspective, Cloudera is doubling down on its open core approach to Hadoop. Last fall, the company simplified its pricing model, introducing a three-tiered product portfolio, in order to better execute this model. Cloudera’s flagship product is its Cloudera Enterprise Data Hub Edition (CEDHE)*, which packages its open source Hadoop distribution with proprietary management software and an all-you-can-eat offering of support services for various data processing “modules” on top of the platform, including Impala for SQL-on-Hadoop and Navigator for data management.

CEO Tom Reilly reiterated to analysts that CDH, Cloudera’s Distribution for Apache Hadoop, is and will always be open source and open architected. This is important as it enables choice for customers, ensuring they can move applications and workloads from one Hadoop distribution to another, thus avoiding vendor lock-in. It’s also enables other software providers to integrate their capabilities inside the Cloudera platform, extending its value to enterprise practitioners.

With the business model clear, how is Cloudera executing? During the analyst event, Cloudera executives revealed they currently have over 300 paying subscription customers, adding between 50 and 60 new customers per quarter. Current customers include marquee names such as Monsanto, Allstate and the U.S. Army. What Cloudera didn’t reveal is how many of these paying customers have successfully deployed Hadoop applications in production.

On the partner front, Cloudera has clearly taken significant strides over the last year, having established several important relationships with software partners such as Splunk, Informatica, and SAS Institute. In all three of these cases, as in others, the partner organizations have worked closely with Cloudera at the engineering level to enable their respective software packages to run natively on Cloudera Hadoop clusters. For example, in a hypothetical 1,000 node Cloudera cluster, 20 nodes can now run SAS for statistical modeling while another 50 run Informatica for data transformation workloads, with both accessing and processing data stored across the entire cluster. Such capabilities are critical for Cloudera to credibly deliver a comprehensive data management platform.

Cloudera said they now have over 240 software partners, over 50 channel partners (mostly hardware providers including HP and Dell), over 50 SI partners (including Capgemini, with which it has co-branded a Big Data reference architecture) and ten cloud partners (AWS, Verizon Enterprise, and IBM Softlayer among them.)

Building A World Class Company

 .

Building a world-class company is obviously important for any start-up, but particularly so for a Hadoop start-up such as Cloudera that intends to go public in the not-too-distant future (as Cloudera executives have stated as their goal on numerous occasions.) This requires Reilly, CFO Jim Frankola and the rest of the executive team to get their financials in order, build a strong revenue base ($100 million plus annually, in my opinion) and sure-up customer support services, marketing and internal operations.

News broke during the analyst event that the company had raised an additional $160 million from lead investor T. Rowe Price, with participation from Google Ventures and MSD Capital. This only adds to the pressure. Cloudera has now raised over $300 million, likely making it too expensive for an acquisition. An IPO is likely the only way for investors to make a return. Such a large raise at this point in its life suggests to me the company is not as close to being ready for an IPO as many (including myself) previously thought. This makes 2014 a critical year for Cloudera. It must ramp up its sales, cross the $100 million revenue mark and get within shouting distance of cash-flow neutrality. The new funding will certainly help, but Cloudera still needs to execute.

Having said that, it was clear from the analyst day that Cloudera is firing on all cylinders in many parts of its business, in particular customer support services. Peter Cooper-Ellis, Cloudera’s Vice President of Engineering, said his team is using Hadoop internally to provide proactive support to customers, analyzing data on a 40TB cluster to discover potential problems before they manifest themselves to customers. His customer support team is up to 60 members with offices around the world.

Cloudera has also made significant improvements to its marketing messaging. It’s new pricing model, with the Enterprise Data Hub at its heart, is a much easier story to tell to potential customers and enables the company to shift the conversation from one of technical details to one of business value. This should shorten sales cycles and result in more CEDHE deals, which are significantly more lucrative over the life of the subscription to Cloudera than the company’s Basic and Flex subscriptions.

Time to Execute

My major takeaway from Cloudera’s analyst day is that the company has taken some important steps over the last year to up its game in anticipation of an IPO, but there is plenty of work to do.

The company is innovating on the technology front, adding new capabilities such as search and SQL and partnering where it can to fill holes and further extend the functionality of its platform. It is also sharpening its go-to-market operations and messaging, and making customer support a top priority.

But at this point in its lifecycle (the company was founded in 2009), I would have expected Cloudera’s customer count to be higher than 300 or so. The company needs to ramp up its customer wins and significantly scale its revenue in 2014 if it wants to go public in 2015. Its reseller arrangements will help somewhat, but, unlike competitor Hortonworks, Cloudera is largely betting on its internal sales team to do the job.

The good news is the company has thousands of potential customers using its free distribution, many of whom will be looking to take the next steps towards production deployments in 2014, including paying for better management software and customer support. Cloudera needs to win its fair share of these deals, including a good number of CEDHE deals.

Cloudera’s business model is sound and the company is maturing rapidly. The Hadoop market is hot and largely wide open. Cloudera has a major opportunity this year to solidify its market leading position. Now it all comes down to execution. If it falters, there is no shortage of competitors waiting to pick up the slack.

*This post was updated at 12:37pm ET to clarify the name of Cloudera’s flagship product. The correct name is Cloudera Enterprise Data Hub Edition.

feature image: Marius B via photopin cc

photo credit: opensourceway via photopin cc