I just read an excellent article by Michael Healey at InformationWeek titled, “Go Rogue with Enterprise Search.” In this article, Michael makes the point that a relatively small percentage of companies have implemented enterprise search, and that most of the companies who have implemented it have typically indexed only one or some of their sources. He refers to enterprise search as the “the most powerful but underused technology available to IT” and suggests that the limited use of full enterprise search is not the fault of the software, but rather a problem with how companies view and deploy search. Anyone considering search should read his article.
In the article, Michael reports on companies who experienced low search adoption rates. There are many possibly causes for these low rates, including controllable factors (such as those I am discussing in my blog post series, “An Intentional Approach to Adoption”); and factors that have to do with the type of enterprise search solution implemented.
Many of the existing enterprise search implementations might be referred to as “Enterprise Search 1.0” solutions. That is, these are solutions which are generally embedded in specific systems or content sources, rarely access both structured and unstructured data, and in some cases, due to the proliferation of search within different systems, can only “federate” search results. Consequently, these legacy enterprise search implementations might require people to know where information is stored, might require the user to search in more than one location, and may struggle with producing quick relevant results. Given the prevalence of enterprise search 1.0 implementations, it is likely that this older approach to search might have been a good part of the lower adoption rates mentioned in Michael’s article.
I implemented a comprehensive enterprise search solution 4 years ago at my former company and have talked with people about enterprise search at a couple of hundred companies since then. My company’s implementation achieved 100% adoption and absolutely great (and measurable) benefits for our business. It became clear to me that companies should be flocking to enterprise search and I couldn’t agree more with Michael’s statement that the “lack of search is quietly sapping productivity”. I have left my former life as Director of IT to spend time with companies who are considering and planning enterprise search implementations in order to do something about this. We have too many difficult problems to solve in our businesses to ignore the benefits we can get from something like enterprise search.
My Experience With Full Enterprise Search
What should a “full enterprise search” solution look like? In my former company’s case, we went beyond what most analysts would describe as a full enterprise search solution, to include something referred to as an “enterprise search 2.0” approach. The basic elements include:
- Indexing of all our unstructured data sources: including all file shares, all files on local hard drives, all SharePoint sites, the Intranet, our company web site, all confidential repositories, and all email (on the Exchange server; and email saved in archive files on local hard drives, saved on the network, and in our email archiving)
- All of the company: all business units, all corporate departments, all offices, virtual workers, departed workers old content, etc.
- Data from our databases (ERP, CRM, etc.)
- Search features and simple search-based applications that combine information from the structured sources (ERP, CRM, etc.) with all of the unstructured content to provide very high level knowledge. This stuff is really amazing.
- Interfaces that return more than just lists of documents, but consolidate information from various sources, sometimes using dashboard elements
Enterprise search paid off for my company in a few months and then started delivering a very real financial benefit that grows each year. Experiencing enterprise search 2.0 is like walking through a hidden gate and seeing a whole new world on the other side. Once one has seen that world, it is impossible to imagine not having instant finger tip access to all of the information, knowledge and expertise in an organization.
Problems Companies Have With Seeing the Value of, and Embracing Enterprise Search
There are a number of fundamental human and organizational problems that get in the way of companies seeing the great value of enterprise search.
An IT perspective: As a former IT Director who has had this conversation with many other IT Directors, enterprise search is not something that seems worth putting aside other conventional IT infrastructure projects for. The conscious or subconscious thinking goes something like this: “We have a backlog of important IT infrastructure projects we need to get to, in order to deal with growing storage, security, disaster recovery, etc. Why should I take on a project like enterprise search when I have many other IT infrastructure projects I want to do and where I know and understand the need, when I don’t understand the information access needs in my company? If this is a business problem; they should come to me if they need this.”
A business perspective: For people on the business side of a firm, perhaps the problem is even more basic. First, having never seen a full enterprise search implementation, they don’t realize the great benefits they are missing and thus will never go to IT asking for enterprise search. Second, people have been told that they could deploy an information system to give quick access to a company’s information and knowledge many times over the last few decades. Most companies have implemented database and document repository systems multiple times, hoping that they would eventually get the “right technology” and deliver on the promise. These projects have consistently under-delivered or failed and thus have bred a very healthy skepticism for promises about information and knowledge access. An enterprise search 2.0 solution is truly a radically different approach, one that if implemented well is finally the answer to all of those former broken promises. However, anyone who has been a part of those past IS projects should be asking tough questions.
Those of us who have implemented full enterprise search solutions that have reached 100% adoption and who have worked on dozens of database and data repository systems over the years are prepared to answer those tough questions. We are, and must be, able to explain why and how enterprise search will actually deliver on those past promises, and how we will be able to help them drive toward full adoption.