Thunderdome

« Back to the top page
IDG News Service

CRM Tips: How to Find the Right Analytics Tool

David Taber, CIO11.18.2009
Tags
Comments 0
Like the story? Get Alerts of big news events. Enter your email address

CRM systems typically have a built-in reporting engine, and some are pretty amazing considering they're driven entirely by browser-based wizards. Most systems have some sort of dashboard system that give executives a little eye-candy.

Nevertheless, it won't be long before you need to go beyond the internal reports.

Here are some classic CRM reporting issues:

1. Let's face it: Few users can create an ad-hoc report that delivers what they want. Instead, they'll need lots of easy-to-use templates and examples. Yet a surprising number of CRM system vendors seem to think XML, SQL, PHP, or Reporting Services programming is just the thing for end users.

2. In the other extreme, CRM reporting can be too easy. Users often click through many wizards, not realizing that they've created reports with no meaning. Then they'll usually get into long arguments about why their reports deliver different numbers from the official ones.

3. The internal reporting systems don't scale well. Typically, the result set must be throttled at a few thousand rows. Big reports typically have to be exported to a spreadsheet or database file.

4. The internal reporting systems can't consolidate data that executives want into a single report. Joins are typically limited to two tables. Even with Salesforce.com's innovative custom reports, marketing, engineering, and finance users will be wanting more.

5. Many reporting systems do not have fine-grain access controls (or the controls aren't properly configured). Thus, users can report on things they really shouldn't see in the first place. There's also a lack of audit trails in some CRM reporting systems.

The Excel Alternative

One way to get beyond these issues is to turn to a proven classic, Excel. It's fairly powerful, very flexible, quite intuitive, and already runs on most PCs. People already know Excel's basic functions and can get up to speed with the quirks of CRM data quickly.

At first, Excel will most likely be used to make the format of the reports prettier. But it won't be long before analysts start hacking VB macros to filter and organize the data better.

Of course, herein lies the downside to Excel: Somebody will need to create and maintain those VB macros. Excel's spreadsheet is also a poor way to handle multi-dimensional analysis (which you'll need surprisingly soon with CRM data), time windows, or really big joins. Sure, it can hold a million rows--but just try to move a column when there are eight VLOOKUPs in place.

Most CRM shops will rapidly move analytics to a desktop database for speed, power, and scalability. We've seen amazing systems created out of linked Excel and Access files. Really nice graphics for the execs, too.

The analytical power is there, but analytics will have to be hand-coded. And too often the code is inscrutable to all but the author. Therefore, this approach can only take you so far.

Beyond Excel

Higher up the food-chain, some reporting engines and business intelligence tools plug directly into advanced CRM systems. These pre-integrated tools have less of a learning curve thanks to templates and examples that map directly to the CRM system's object model.

The best of these tools aggregate and analyze data from outside the CRM database equally as well as data from inside the system. Their user interface is usually easier to grasp than general-purpose BI tools, and the initial setup can be measured in hours. The most powerful tools, though, require real training, particularly for their most advanced features.

The largest customers typically have a data warehouse fed from the CRM system using stand-alone BI tools. These systems run circles around everything else, particularly for multi-dimensional and multivariate analysis.

But they're expensive to procure, setup, and maintain. Getting results from these systems invariably means heavy training, maybe some coding, and almost certainly a full-time data analyst.

Which Tool is Right for You?

Consider these three questions when choosing which tool is right for you:

1. How much data do you have?

2. How sophisticated are the analytics you want to run?

3. How clever and well-behaved are


Post new comment

The content of this field is kept private and will not be shown publicly.
Respectful debate is welcome, but comments that are defamatory, indecent, abusive, or in violation of any law will be removed.