As you know by now, Cloudera and Hortonworks have announced their plans to merge. The companies reportedly have a combined fully-diluted equity value of $5.2 billion and expect to generate approximately $720 million in revenue.
As after any notable merger, articles and commentaries surface, asking things like whether this represents the death of Hadoop and if the consolidation was a move for the two companies to improve their margins. Those will get played out eventually but they’re fundamentally dependent on whether the companies can execute the merger successfully and minimize the disruption to customers.
Having gone through several mergers of this size and larger, I have experience with the challenges involved. In this blog, I’m going to focus on the challenges with the merger and what it represents for the competition, the companies themselves, and most importantly, their customers.
Although it appears that Cloudera is slightly more dominant with the 60/40 split given the way the companies are structuring the deal, it’s likely to be more of the merger of near equals. In my experience, this operating paradigm is harder to pull off than a merger that involves a dominant company, where that dominant culture enforces itself and execution risk is minimized with the dominant management team remaining in place.
With a merger of near equals, everything needs to be carefully negotiated, including product line, marketing strategy, pricing and positioning, which can lead to disagreements and politics that take time to work out. The first thing the companies need to determine is what the executive team will look like and ensure that the leaders work well together de-facto – not just de-jure. That executive team composition will help identify the culture of the new company and how it will go about making decisions — everything from sales and marketing and positioning to product and customer support decisions.
In addition to management and culture challenges, with this merger there’s a huge overlap of product lines, which creates issues to be resolved. Important decisions regarding which product to continue and which to phase out are fraught with politics. Again, these discussions and decisions take time, which will undoubtedly cause some confusion and uncertainty in the marketplace. Customers will need to be told sooner than later what products survive and how they will transition from one to another. This will be a great opportunity for MapR and others to take advantage of the uncertainty.
Pricing and Vendor lock-in
Pricing is going to be an interesting issue for customers using products from both Cloudera and Hortonworks. Consolidation between the two largest vendors creates lock-in risk for some enterprises in that they become too dependent on a single vendor and the vendor could exert pricing leverage. This will force the largest customers to evaluate alternatives simply to mitigate risk. Additionally, if the spend is large enough with the combined entity, tech savvy enterprises will need to consider going to Apache and supporting the infrastructure themselves as it is likely cheaper in the near term and puts the control in their own hands.
The Cloud and AI
There are two dominant use cases for Big Data: ETL and Analytics. In the case of Analytics, a direction on which the new company will undoubtedly focus, the cloud is a big competitive concern during the consolidation process. Without a well-defined and articulated cloud strategy, Cloudera and Hortonworks face the risk that vendors like Microsoft and Amazon or Databricks and Snowflake could come in and displace them.
AI is also a long term strategic goal for the new company but it is hard to see how they can even get to par with Google, Microsoft or Amazon. The big guys are far more knowledgeable and if the data gets moved to cloud object stores negating one of the key principles of Hadoop – cheap storage, it will be increasingly more difficult to get and retain customers.
As the merger moves along, workforce cuts are inevitable. There will be pressure for retention because the first people to leave are usually those you don’t want to lose because it’s easy for them to find other jobs. With companies and talent firms — including those at the cloud and technology vendors I’ve already mentioned — surely using this opportunity to recruit, we’ll start to see key people leaving, particularly in engineering.
From a customer perspective, these losses will raise significant questions about the stability of the new company. It may also cause some customers to question whether Hadoop is in fact dead and should they consider accelerating their move to the cloud or another solution.
What It Means for Pepperdata
With a merger like this, discussions inevitably arise about the future of the companies as well as the technology. The management team at the newly formed company will figure things out. However, that does not mean it is without uncertainty for the next 12-18 months and they’re going to have to plan and manage that transition very carefully.
At Pepperdata, we’re rooting for the company’s success. Regardless of the outcome, we have customers relying on us to ensure their big data platforms and analytics solutions are running well. We play an important role and it will get even more crucial going forward.
Pepperdata is not married to Hadoop or YARN or any one commercial solution. With proven products, experience and expertise and our demonstrated ability to work with Kubernetes, we are focusing on workloads that go beyond Hadoop and HDFS. We will continue on that path irrespective of this merger. As a matter of fact, it will likely accelerate our progress.