Q&A: Mobile, Search, and Cloud Top "Cool BI" Innovations
In this two-part interview, well-known BI analyst and TDWI faculty member Cindi Howson discusses some of the top innovations affecting business intelligence, in keeping with her annual "Cool BI" list of business intelligent innovations.
- By Linda L. Briggs
- January 17, 2012
In this two-part interview, well-known BI analyst Cindi Howson looks beyond the hype to discuss some of the top innovations affecting business intelligence, per her "Cool BI" list of business intelligent innovations. Howson is founder of the BIScorecard.com Web site (BIScorecard.com), which offers independent evaluations of leading BI products. In Part 1 of our interview, she discusses trends in mobile BI, search, and software-as-a-service (SaaS).
BI This Week: Let's start with mobile BI. With smartphones and now Apple's iPad driving things, mobile computing and BI seem like hot areas. Are BI companies taking advantage of mobile and BI?
Cindi Howson: There's a lot of hype about mobile at conferences, so I think two things are happening. There's messaging about it because of the increase in adoption of tablet computers. You mentioned the iPad, which I think has reinvigorated the interest in BI for many executives and unleashed some of the potential for frontline workers.
That's one phenomenon, but I also think some companies are still wondering, "What's the big deal about mobile? Most of our people are desktop workers, so do we really care about this?" I think it's true that there are some job types, some functional areas in some industries, where mobile is less relevant, but I think there's excitement about it because the iPad is cool. Business intelligence has not been cool, it's been boring, and I think the iPad helps reinvigorate it.
Who are some of the vendors in the space that are doing interesting things?
All the BI platform vendors have their solutions, of course, so it's then a question of which devices they support and how. "How" is where there is almost a holy war going on in terms of whether users have to buy specific apps or should there be an optimized Web app -- a browser approach? There is absolutely no agreement in the industry or among vendors on which approach to take.
Do you have advice on which direction to choose, especially if a company is already committed to a specific BI platform?
Each vendor has a particular preference. MicroStrategy, for example, believes in a device-specific app. Roambi believes in a device-specific app, saying it gives users a better experience. At present, I agree -- I prefer device-specific apps -- but there is a trade-off in terms of which devices are supported. There's also a higher burden on the vendors for development costs, but I'd always rather be harder on the vendors and easier on the users.
QlikView initially had a device-specific app; now they have a browser-based approach. The good thing about their approach, however, is that you can run it on any device, the iPad as well as a Droid tablet. You don't have to re-author existing content, and users don't have to download any particular apps. On the downside, I think there is some loss in usability and an inability to exploit some of the device-specific features.
Along with QlikView, who else is espousing the browser-based approach?
Information Builders takes somewhat of a blended approach. They have one app called Mobile Faves that provides some of the offline viewing capabilities and richer experience, but they also have a predominantly browser-based approach.
Moving on to search -- which is another of your Cool BI choices this year -- who are the BI vendors in the search market right now?
Here, too, a number of BI vendors have search options built into their platforms. Some of them seem to view it as a checklist item; some view it as wanting to have a simpler interface based on search concepts. SAP BusinessObjects, with their Explorer interface, uses a search paradigm that combines search and visualization. Other vendors, such as MicroStrategy, don't believe in it. QlikView uses search to a good extent, and there's Endeca, just acquired by Oracle.
Was Endeca's search capabilities one of the reasons for the acquisition?
It was probably partly for search, and partly for Endeca'sMDEX engine, which combines structured and unstructured content into a columnar database.
What are important search capabilities to look for in a BI platform?
What's important about search is ease of use -- the simplicity. Business intelligence should not be viewed just as a power user tool, and yet how do we make BI appealing and usable for casual users -- somebody who just needs to look up something periodically? They shouldn't need any training; they should be able to find what they're looking for quickly and easily. That's the appeal of search. I think right now, it's still more vision than reality for most products and companies.
Have most vendors done a good job making search simple?
No. The market just hasn't adopted search yet. I don't know why. There are certainly cost reasons, and maybe some complexity reasons. There's a whose-needs-do-we-serve-first [approach], and the answer still seems to be the power users and executives. Search isn't necessarily as relevant for that group of users.
The interest is there, though. In my TDWI classes, we use polling software to gauge what people are enthusiastic about. In previous classes, there seemed to be a lukewarm reaction to search, but in my latest class, 57 percent of the class rated search as very cool and effective in reaching more users.
What's happening in the software-as-a-service area? Do you include cloud computing in defining that term?
There are variations. Software-as-a-service is one category of cloud; in that case, you're buying everything. You're buying use of the hardware as well as the software and a subscription license -- it could be monthly, quarterly, or yearly.
SaaS has been around for awhile; is it continuing to heat up now? Is there more interest, and perhaps more trust in letting a vendor share responsibility for the company's data?
There is definitely more interest. I think the trust issue is still one of the biggest concerns holding back both cloud and software-as-a-service. The appealing aspects with software-as-a-service are the rapid implementation time, rapid time to value, and little to no upfront hardware or software cost.
So it's very attractive in a down economy?
I don't know that it's so much because of the economy, but it's especially attractive for small to midsize businesses that don't have the expertise and people to go out, evaluate, and buy everything that's needed for BI.
How well are vendors addressing the SaaS market?
It's very mixed. The vendors that continue to lead in this market tend to be the specialty software-as-a-service vendors such as Birst and PivotLink. There are also some traditional vendors that have expanded their capabilities into the cloud, the software-as-a-service market, or both. SAP BusinessObjects has had CrystalReports.com for years now. Because they've been growing that and learning from it, I think they have some good cloud offerings. This summer, MicroStrategy went full tilt into the cloud, offering the full MicroStrategy suite in the cloud and running its own data center -- which is unique. Most other vendors prefer to use a shared service.
Microsoft has been doing a lot in the cloud as well with Microsoft Azure, and Jaspersoft came out with a release of Jaspersoft Cloud BI late last year.
How do you advise clients who are interested in software-as-a-service or cloud computing in general?
First, they need to look at what their goals are. Are they trying to get rid of owning the hardware? How would they prefer the data to be handled? Is it on-premise and they want the software-as-a-service vendor to reach into the on-premise infrastructure, or will they upload it to the cloud? If they're going to be uploading it, that is probably less suitable for an operational or daily update.
They need to look at what they're trying to achieve, then evaluate the cloud and software-as-a-service products in the same way that you would with an on-premise solution. What are the capabilities you need out of the product? What do they do to address data security and elasticity requirements?
Speaking of security requirements, is that still a valid concern with cloud computing?
It's a valid concern in that you're letting go of data that you previously controlled. I conducted an interview with a banking institution that had adopted Birst as their solution. They had in interesting perspective suggesting that there's a false sense of security that a corporation can do security better than some of these cloud infrastructure companies. You have to look at what that SaaS or cloud vendor has done, and compare it to how secure your own data center is.
There are categories and industries where I think the security has been figured out, and sometimes it's better than what's on-premise, I don't want to say security's an invalid concern. It's valid, but I also think it's something that vendors have addressed.