I’m asked regularly whether Microsoft has abandoned the Enterprise Search market. This was a frequent question in 2015, and less frequent in 2016, but there’s been a recent uptick, and I got this question 10 times last month. As a long-standing search nerd that lives close to Microsoft, I know the answer is NO. But I was baffled about why this question keeps coming up.
So I decided to investigate. This blog takes you through what I’ve found and how you can answer the question when it comes up. Search Explained is the perfect place to publish it.
Why do people presume Microsoft is out of the Enterprise Search Market?
“Reports of my death were greatly exaggerated” – a quip attributed to Mark Twain (actually a misquotation) – has become a popular culture trope found in movies, song lyrics, and tweets. When this happens in real life, it’s usually a matter of perception and miscommunication. That’s definitely the case with SharePoint Search.
Overall market perceptions change slowly, so the roots of this question go back over a few years. When I went through it, there are 6 reasons people presume Microsoft is out of the market:
1) Changing Marketing and Terminology – the market as a whole, and Microsoft in particular, has a tendancy to rebrand things or change they way they talk about them, which can be confusing.
Marketers and analysts in the industry have been uneasy about the word “search” for a long time, and there have been periodic attempts to redefine it or replace it. “Search” has several weaknesses as a term: it’s primarily used as a verb (which is weak), it denotes an incomplete activity that’s not useful in and of itself (seeking something rather than locating it), and it has accreted some baggage over the years, including the expectations that enterprise search is just like web search. People instead have used ‘find’ and ‘findability’; ‘discovery’ and/or ‘exploration’; ‘question answering’ and different variants of ‘analytics’. The enterprise search field has been called ‘information retrieval’; ‘unified information access’; and recently, ‘cognitive search’ and ‘insight engines’.
There’s a place and role for all these terms, but it’s confusing. Meanwhile, ‘search’ continues to be widely used and none of the alternative names have really displaced it. The term search just refuses to go away.
When Microsoft stopped talking much about Enterprise Search and started talking more about Delve and Discovery, it contributed to a sense that Microsoft wasn’t doing search any more.
2) Long, confusing product evolution – within Microsoft, enterprise search has been through a number of transitions, not the least of which was the acquisition of FAST back in 2008. I drew out a picture (see below) about all the transitions – the dotted blue lines denote a major re-architecture, which you can see has happened a lot.
When I co-wrote “Professional Microsoft Search” about the 2010 generation, there were 7 variants in the product lineup. When the 2013 generation came out, fully integrating what had been a next-gen technology project from FAST, it changed search dramatically. SharePoint Search is Dead, Long Live SharePoint Search was the title of a TechEd session by my friend Neil Hodgkinson at that time – touting how completely the new search had been revamped with SharePoint 2013.
There’s been plenty of name changes here too, contributing to the confusion/perception. The FAST brand was dropped in 2013 (though it was partially resurrected in a variety of Microsoft material). I have heard people say that Microsoft acquired FAST then dropped it after a couple of years! This is the complete opposite of the truth, but contributes to the rumors of SharePoint Search’s demise.
3) Bundling – along with the 2013 generation, Microsoft dropped their Search Server products. This was to simplify the product lineup. After all, if you wanted a great standalone search product, you could just use a few SharePoint servers, configure them with search services, and not use anything else. SharePoint still had a very strong search capability. In fact, the free version (SharePoint Foundation) became a very capable search product, up to 1M items.
However, the ‘Search Server’ name was gone, another easy-to-misinterpret signal. Plus, there was now a genuine constraint: to use SharePoint as a standalone search product, you still needed to be licensed for SharePoint. If you already own SharePoint, the price of using a few more servers to create a search farm is very small. But if you don’t already own SharePoint, it usually does not make financial sense to bring it in for the sole purpose of search.
4) Gartner Coverage – among analysts covering the search market, Gartner is the best known. In 2013, when Gartner published their Magic Quadrant on Enterprise Search, Microsoft was an upper-right market leader. But in 2014 Microsoft was gone. In 2015 Microsoft was also not included. Why? The criteria for inclusion was to have a standalone search product, and Microsoft did not qualify because of their bundling (neither did Oracle or OpenText). Gartner has a big influence (probably way too much), and this was a signal to many that Microsoft was out of the market.
5) Partner Strategy – in search, as in many other areas, Microsoft uses partners to fill out solutions, more so than any other major software vendor. This partner ecosystem is a big advantage in many ways. But there are many areas left to partners, sometimes referred to as “white space”. Microsoft Enterprise Search has a lot of white space – things like connectors, autoclassification, and analytics are provided by partners. A company looking for a complete enterprise-scale search solution cannot really get it straight out-of-the-box with SharePoint search; they have to either do extensions themselves or get add-ons from Microsoft partners. There’s a perception that this means Microsoft is not fully engaged in the search market, since they don’t address all the requirements.
6) Development focus – after the release of SharePoint 2013, the search development team within Microsoft turned to a new, innovative area: the creation of the Office Graph (initially called “project Oslo” because that’s where most of the development team is based). Bug fixes for search were slower, and new features for search were less frequent. SharePoint 2016 search is essentially the same as SharePoint 2013 search if you use it on-prem. Even people quite close to the development team expressed frustration with the lack of focus on traditional search issues. That also added to the sense that Microsoft had moved on.
When you look at these six areas, it seems natural that people might assume Microsoft had exited the market. Four of these are purely perception, but two have some reality as well. Bundling search into SharePoint truly did make Microsoft search a non-starter for some companies, and Microsoft’s development focus actually did shift away from traditional on-prem search towards the cloud.
What’s now clear to me is that most of these factors came to the fore in 2014 and 2015. Certainly, 2015 was the year I got the most questions on this topic. Ironically, this was shortly after the rollout of SharePoint 2013, which was from Microsoft’s point of view a great success cementing their leadership of the search market. I find that Microsoft has a “mission accomplished” mentality – when think they’ve won the momentum on the current battlefront, they shift resources and focus almost entirely to a new battlefront. This is what happened in search – focus shifted fully to the Office Graph initiative, leaving the market with the impression that Microsoft had left the room.
See the second part of this blog post here: “SharePoint Search is still alive and well”