The past few months have been a time of "big vision" in the search business. Google and Microsoft shared announcements about their respective visions for enterprise search, and Elastic made an IPO offering that received a strong commitment from the investment community.
In all three cases it is indeed about vision rather than current technology, but a critical barrier stands in the way of their achieving these visions that no one seems to have considered.
Google: Convincing Companies to Make the Switch
Google announced a cloud enterprise search solution that will offer a range of connectors to facilitate a federated search across multiple applications. Technically, this is a very elegant one-stop approach, but the big unanswered question is what the pricing model is going to be. The Google enterprise search appliance offered a notionally transparent per-document pricing model, but many users found the definition of "document" somewhat idiosyncratic. At present, pricing appears to be something agreed upon on an individual basis, probably linked to the publicity value of the customer. I’ve lost count of the number of Workplace by Facebook presentations I have seen recently where no money has yet changed hands!
Google’s major problem will involve displacing existing applications, mainly in SharePoint. Proving that one search solution is better than another on a "proof of concept" test is no guide as to how it will perform in a production environment. The inevitable change in the user experience alone will pose a significant challenge.
Related Article: Google Search Appliance Fades Away
Microsoft: Unraveling a Complicated Search Legacy
Microsoft acquired FAST Search and Transfer 10 years ago and then buried it after stripping out some elements that ended up in FS4SP. In particular, Microsoft discontinued the content processing pipeline. At present, Microsoft has a complex range of search applications, including "classic" and "modern" search, Azure (which is Elastic under the covers) and Bing for Business. Complicating matters, Microsoft's announcements often lack clarity on whether it is a product or a promise. It took Agnes Molnar 40 minutes to explain the current situation at the IntraTeam meeting in Stockholm while the audience was frantically making notes.
Recent announcements about a more coordinated Microsoft strategy mention enterprise-wide solutions being available in 2019, but this refers to the Microsoft enterprise. How third-party applications are going to be implemented remains to be seen. I am curious to see how or if the underlying graph search technology will cope with indexing non-Microsoft applications. Unlike Google, Microsoft has to support a substantial legacy market and that limits its ability to throw bits of its technology stack away.
Elastic: Strong IPO, But Is it Sustainable?
Elastic recently completed a very successful IPO, with the stock currently trading at almost double the offer price, valuing the company at around $5 billion. The NASDAQ description of Elastic (which is well worth reading) positions Elastic as a search company. However, the company seems to focus primarily on the log analysis market. There is a difference. Euan Jones shared a good analysis on Seeking Alpha of the market potential for Elastic that suggested the $70 pricing is not sustainable for a company with revenues of $200 million.
Related Article: Enterprise Search Has an Open Source Secret
The Major Barrier to Growth
All of this technology is only as good as the people who can use it to deliver a satisfactory user experience. The problem, however, is there are not enough of these people. Only a small number of search integration companies and a few Microsoft channel partners have the skills to manage complex search environments. With the dramatic increase in complexity and the positioning of "personalized search" as the ultimate vision, there is an urgent need for search specialists inside the organization. Managing relevance ranking needs to become a profession, ideally with some form of competence certification.
It is remarkable that the only person providing a range of training courses is Agnes Molnar through her Search Explained business. The highly successful conferences such as Haystack in the US and the UK, developed by OpenSource Connections and Flax, clearly prove the demand for training. Unfortunately, there's a lack of people with the expertise and experience to lead training courses. Information Schools also ignore enterprise search, but again who would lead the courses and the tutorials?
A Search Academy?
The only solution I can see is for vendors to appreciate how the lack of search skills will impact their bottom line and come together to support the development of search academies in the US, Europe and the Far East, which teach search management on a vendor-neutral basis. Between us, Molnar and I have developed all the course material — we just need someone to come up with some funding and an ability to promote them.