In-Depth

Q&A: Prebuilt Apps Address BI Complexity

It's no longer sufficient to simply provide people with the ability to get answers. You must give them the right questions to ask.

• By Linda Briggs
• 02/04/2009

"We're in the midst of a huge shift --- from BI as a tool designed for IT to BI providing configurable, prebuilt analytic applications designed for business users." So says Ken Rudin, vice president of market development at the on-demand business intelligence company LucidEra, which Rudin co-founded in 2005. The shift Rudin points out is occurring in response to the complexity of BI, which he says traditional software-as-a-service (SaaS) vendors only partially address.

Rudin has focused on SaaS for years as an executive or advisor with several vendors including Salesforce.com, Netsuite, and Siebel.

BI This Week: The complexity of BI continues to be a real challenge for businesses. Software-as-a-service BI vendors offer a solution in that they host the software themselves. What's the problem you see with that approach?

Ken Rudin: The problem is that it doesn't go far enough. The complexity associated with getting traditional on-premise BI tools up and running is only one part of the problem with traditional BI. Most SaaS BI vendors focus on removing that complexity by providing a hosted BI tool, but they're only solving one part of the problem.

Once you take away the barrier created by complexity and give users the ability to answer whatever questions they have, you uncover the real culprits preventing companies from being successful with BI: First, most people haven't been trained on how to analyze their part of the business, so they have to make guesses about which questions are meaningful. Second, most people aren't sure how to interpret the reports they get and what to do about the information.

Using a medical analogy, you could give me access to a state-of-the-art MRI machine, but since I don't have medical training I wouldn't know what buttons to push. Even if you showed me that, I wouldn't know how to read the output.

For companies to be successful with BI, you have to address all the issues, not just one of them.

What do you see as the solution?

It's no longer sufficient to simply provide people with the ability to get answers -- you also need to give them the right questions to ask. We're in the midst of a huge shift -- from BI as a tool designed for IT to BI providing configurable, prebuilt analytic applications designed for business users. These applications are designed for specific roles in a company, provide built-in best practice metrics and analyses, and in many cases include industry-specific benchmark information. The new generation of business intelligence applications help users focus on metrics that matter, and help them identify what actions they should consider based on the results.

The difference between a tool and an analytic application is huge, especially in terms of value delivered to the customer. It's like being in the market for a new house and someone provides you a table saw and lumber and tells you to go build whatever house you'd like, versus someone selling you a home.

You said that most business people have never been taught how to analyze their area of the business. How does that affect business in general?

The impact is massive. How do people know what metrics they should be tracking? In most cases, it's just passed down as tribal knowledge -- here are some metrics that my manager before me used to track, so I'll track them too, but not tracking the right metrics sabotages your ability to hit your goals.

For example, many companies look at the size of their pipeline to determine how well they're tracking toward their goals. A large pipeline often provides a false sense of security. What if the deals are getting stuck? The pipeline is still large, but it's stalled, so you can't just look at your pipeline size. As a best practice, you also need to look at your "pipeline velocity" -- that is, how quickly deals are moving through various stages of your pipeline.

What are some other best practices for sales in terms of analytics?

The most effective way to increase sales productivity (meaning revenue per sales rep) is to know which deals to focus on -- and it's usually not the largest deals. The best practice for identifying your sales "sweet spot" is to look at the characteristics of those deals that have been your best deals historically. That is, what are the characteristics of deals that have had the best win rates, the largest size, and the shortest sales cycles? These characteristics can include industry, lead source, deal size, product line being sold, and so forth. Once you know these characteristics, focus on those deals in your current pipeline that share those same characteristics.

Another best practice is to look at not just how long it takes you on average to win a deal, but also how long it takes you to lose a deal. On average, reps spend at two and a half times as long on deals they lose compared to deals they win. That's a huge waste of productivity, but since few people track average time to lose a deal, it goes unchecked.

How does LucidEra address some of what we've talked about here?

LucidEra focuses on delivering not just a hosted BI tool but rather a series of configurable, prebuilt analytic applications that include the types of best practice analytics I've been talking about. For example, we help you increase your revenue by finding your sales sweet spot so you can focus on the deals you're most likely to win. Similarly, we can reduce sales cycles by identifying those deals that are moving most quickly through your pipeline, so you can focus on those.

We can help you increase forecast accuracy by first giving you an assessment of the real size of your pipeline -- which of your deals are real, and which aren't (for example, if they've been stuck in the pipeline a long time). We can provide better visibility into your real close rates (by industry, deal size, etc.) instead of using a single average close rate. Third, we can show you what's changed in your pipeline so you can detect issues and correct them.