Pervasive BI: Survey Finds Ease of Disuse
How often do people actually use BI technologies? BI adoption can't seem to crack 25 percent. What if the users actually working with BI are using it to the max?
- By Stephen Swoyer
- May 27, 2014
Why does it matter if pervasive business intelligence (BI) remains a dream deferred?
For that matter, why does it matter if technologists give BI poor marks for business impact? A great potential metric of BI success is frequency of use: how often do business people (i.e., potential BI users) actually use BI technologies? Are they using BI regularly, interactively, throughout the day? Do they use BI at least once daily, or only weekly, or -- worst of all -- less than weekly?
Although BI adoption can't seem to crack 25 percent of all employees, what if the people using it are using it to the max? What if BI provides critical, ongoing support to decision making? What if it's an integral component of the business process?
Luckily, the latest edition of BIScorecard.com's The Successful BI Survey provides answers to these and other questions. At first blush, however, the answers it provides aren't terribly reassuring.
Each year, BIScorecard.com founder Cindi Howson, a principal with ASK Consulting and a long-time TDWI faculty member, promotes The Successful BI Survey. This year's edition yielded a treasure trove of information about BI use and disuse, BI impact, BI drivers and depressors, and other issues.
How frequently do business people use BI technologies in any given organization?
The answer is: probably not as often as most in BI would like. According to this year's edition of the survey, almost two in five (38 percent) BI users actually use BI daily -- but a majority still uses BI on a weekly or on a less-than-weekly basis. (These are qualified users: i.e., business people to whom BI apps, tools, and services have already been deployed or provisioned.)
As Howson stresses, frequency of use "depends greatly on the frequency of the decision making as well as the timeliness of the underlying data that would support a decision." This is doubtless true. Why, then, do so many in the industry are concerned about flat metrics such as "adoption" or "usage?" Frequency of use is a function of two important variables -- how often decisions are made and the timeliness or freshness of the data that business people use to make them; this makes it irreducibly contextual -- not static or flat. Not all people require BI. Not all people will benefit from BI. For some people, in some contexts, decision-making is an aperiodic event. It doesn't happen that often or the data to support a decision just isn't there. This is true for a finite number of people.
The survey attempts to control for these people by asking respondents how many of their employees could realistically use or consume BI. In 2013, that number stood at 50 percent. In any given organization, then, half of all employees could use or benefit from BI.
If people who could benefit from BI aren't frequent users of BI, it's likely that enterprise BI teams and BI software vendors aren't doing enough to address their needs. Timeliness or freshness of data is one problem -- a business person isn't even going to bother going to a BI tool if she knows the data isn't there -- but so, too, are inflexible or unresponsive BI teams; unwieldy (hard to change, hard to manage) data warehouse systems; hard-to-use tools; and overly rigid or authoritarian corporate cultures.
To combat this, Howson says that BI boosters need to focus like a laser on what works -- on what's efficacious. This might mean seeking flexible alternatives to (and, ultimately, replacements for) restrictive or inflexible architectural components, IT initiatives, tools, and the like. (As Howson notes, "[s]ome BI architectures and products may also be more suitable for certain usage patterns." She cites MicroStrategy, which had the highest proportion of daily use, at 47 percent.) It likewise might require a degree of political involvement: BI boosters from both IT and the line of business must agitate for corporate and/or cultural change, solicit (or woo) backing from important executives or stakeholders, and do what it takes to create the conditions for BI success. In spite of the challenges, she remains both optimistic and unequivocal.
"For business intelligence to have a more profound business impact, BI needs to be deployed to all personnel who make decisions, whether strategic or operational. In expanding BI beyond information workers, BI teams must recognize that field workers rarely ask for BI," Howson writes.
She makes a few prescriptive recommendations.
"Power users such as business analysts often act as a front line for information requests by other decision makers such as managers and field personnel. Once their information needs are fulfilled, BI teams should proactively understand what drives the business to consider where BI can be applied to new groups of users, including field sellers, customer service representatives, and external customers and suppliers.
"Integrating BI with an operational application [i.e., by embedding BI] or providing BI at the place of work via a mobile app are two ways to improve BI adoption beyond power users," Howson concludes.
The nearly 60-page report of survey results deals with much more than just BI's disuse. Other topics include a discussion of (and recommendations about) factors for technical success; executive support and the success of BI projects; and organizational factors.