In a move that may indicate the beginning of a new wave of activity in the business intelligence (BI) market, Oracle has announced its intention to acquire Endeca. Founded in 1999, Endeca originally focused on search capabilities for online commerce. Users selected a product attribute, and the software automatically revised the remaining selection criteria based on products matching the previous selection. We have been covering Endeca as part of the BI and information applications marketing. For instance, if the products only come in one color, the color attribute would be removed from the selection criteria and possibly replaced by other relevant criteria. Most of us take this behavior for granted as it has been adopted or imitated by many e-commerce sites and other Web properties.

More recently Endeca began applying its search and navigation capabilities to the business intelligence market with the introduction of Latitude which it positioned as a separate product line from its InFront e-commerce products to gain flexibility in rolling out new features. Both product lines were mentioned as being valuable in the acquisition process, but in my opinion without Latitude the deal would not have happened.

As I’ve written before, end users understand and prefer search-based navigation, popularized by Google. BI vendors long since incorporated search capabilities into their products. However, BI search capabilities have been limited largely to structured data and metadata. Endeca cracked the code for searching both structured and unstructured data at an auspicious time. The proliferation of unstructured data has occurred in conjunction with explosive Internet use and the rise of Hadoop to manage large volumes. Our recent benchmark research shows that Hadoop users often work with unstructured data. In August Endeca announced that it had expanded Latitude to support Hadoop.

Congratulations to the Endeca team: They spotted an opportunity and created an innovative solution leading to its acquisition. Beyond that success, there are implications for the broader market. We identified the importance of search capabilities and the need for business to assemble information into what our firm has defined as Information Applications for which our Information Applications benchmark research validated. I expect to see more activity and investment in making it easier for organizations to leverage unstructured data. Several other companies are drawing attention in this context. My colleague Mark Smith has written about Datawatch which has technology for extracting information from Adobe Acrobat, HTML and XML documents. Actuate acquired Xenos to incorporate similar capabilities as part of its BI product line and to be part of the larger demand for information and not just analytics. With its $10 billion valuation, the pending HP-Autonomy deal suggests that the importance of handling unstructured data has not yet peaked at least in the eyes of one investor which also is in the business of providing information based applications to its customers. Another company, MarkLogic, that my colleague assessed recently has also seen significant growth.

Venture capitalists also are likely to invest more in this market. They tend to flock toward segments where some success has been proven but has not yet gone mainstream. As part of their portfolios of investments, many VCs want to have one or more companies that compete in the same market. The presence of more investment will also draw more startups into this segment, and thus the cycle of investment and acquisition will continue as suggested in my post on the software ecosystem. The result for enterprises will be positive as software companies deliver more information applications capable of helping them navigate more easily through increasing volumes of structured and unstructured information. Oracle is not new to information and applications, so the Endeca acquisition will compliment its Fusion middleware efforts that has been continuing to expand its footprint.


David Menninger – VP & Research Director