Speaking of that democratization, we've seen strides in delivering analytics to the business user. Where do you ultimately see this trend heading?
Every knowledge worker in an organization is going to need to have timely, actionable, targeted information available to them. We're not talking about burying them in reports, which is what some of my BI brethren are trying to do. The information is highly individualized or personalized, given your role and what you do. It leads you to insight and action.
The other thing that's a big trend is where knowledge-worker analytics and insight disappears. It's running, it's out there, it's looking at data based on instructions and advanced algorithms. But those insights are delivered directly in context to a call center agent. They should say, 'This is Jon and, because of that, here's how I dynamically alter the script.' And while I'm interacting with him, I'm taking feedback and that's how I'm going to guide how I interact next. I have yet to see a CEO who says, 'I don't want my work force having insight.'
FOR MORE INFORMATION:
Article: Siebel kicks off 7.7 with analytics
Article: Analyzing the CRM analytics race
When did analytics become such an important part of the customer service space, especially the call center?
You used to see the words 'marketing' and 'analytics' used together the most. Siebel's first analysis products were add-ons to the marketing product. We're seeing a push toward using insight to measure effectiveness. On the sales side, it's the ability to enable enhanced sales effectiveness. We're seeing the democratizationof sales information. We're early in that part of the game. We're also seeing it in the call center and with field service, where there's a huge opportunity to improve service levels, increase customer satisfaction and drive real results. We're also contending with things like the 'do not call' [registry] and clampdowns on privacy that are changing the role of traditional marketing from outbound blasting into discovering relevance and receptivity when the customer is calling in on a service experience. Or, even before you call in, contacting that customer and saying, 'Because you're such a valuable customer, we could save you money by migrating you to Plan B.' That segues perfectly into the contention that analytics aren't very useful if they aren't predictive. Wouldn't it be better to know beforehand that the customer was likely to discontinue the service?
Analytics that aren't predictive are still very, very useful. What I could do is look at [the customer's] traditional usage patterns and, if I see a significant decline and know that it's a valuable customer, fire off an alert to a sales or service agent -- along with the ability to give that customer a discount. Is that predictive or not? If you're going to be exact, that is not real data mining from a predictive standpoint. But it is highly useful. Having strong descriptive analytical capabilities to describe the current state of things is important. Putting that together with business rules and treatment is very important. These days, we hear so much about real-time analytics. Does that mean that batch-mode analytics is dead?
Batch mode is dead, meaning if you can't have interactive access to information you're not in the modern era. What you may have is interactive access, but it only hits a data warehouse that's updated every week or at the end of the day. For certain businesses, that's sufficient.
Too many companies are complacent with information that isn't live or very current. The traditional -- or, dare I say, 'legacy' -- BI-tool vendors were not architected to deal with a data warehouse that has historical data that's optimized for [up-to-the-minute] access. For instance, when you're using BI analytics to drive more intelligent actions or interactions, wouldn't it be more important to know that [a customer] has just gone to your Web site to discontinue your service than to wait and find that out much later on?