CRM systems are where the data about customer relationships are supposed to live, and they typically provide a report writing system as well as dozens of canned reports. But reports are incredibly vulnerable to GIGO, and they immediately expose data quality problems. Let's look at issues that limit the validity and credibility of any reports in your CRM system.
Danger Zone #1
Guess what: CRM systems can have serious data quality issues with duplicate records, phantoms, incomplete records, and inaccurate fields. While there are plenty of tools, services, and methods to clean up the data, the default assumption needs to be that the records are wrong... until proven otherwise. This issue only gets hairier when the CRM system receives data from outside feeds or is integrated with your other systems.
The solutions here are policies and real-time processes that preclude (or at least flag) bad records from being imported or hand-entered into the CRM system. This may seem costly, but it's way cheaper than the inevitable wild-goose chases due to bad data.
Danger Zone #2
So you have clean records, great. But what about the semantics of the entries: does "qualified lead" mean the same thing in the US as it does in your European operations? Are some divisions using a named account model and others a lead generation model for pipeline development? Unless someone has analysed your sales and marketing processes, the CRM system will probably have important semantic discrepancies that make cross-divisional comparisons meaningless.
The solutions here are policies, semantic standards and a data steward function that control the creation of data fields and manage the evolution of their meaning and usage over time. While both low-level people and impatient executives will balk at this data governance, it's essential for meaningful reports (and SOX compliance!).
Danger Zone #3
It's all too easy to construct reports that run and look right if you spot check using only records that you entered, but produce incomplete and silly results if used across the organisation. The most obvious symptoms in this Zone are reports with summaries that don't "foot" due to double counting, omitted records, or both. These issues typically come from a misunderstanding of the system object model, not anticipating consequences of the security system, and inappropriate use of filters and time windows. These problems become highly visible when VPs in different departments present their team's reports, with results that contradict each other.
The solution? Have a trained data analyst produce accurate, meaningful reports that are used by all organisations. Despite the ease of use of CRM report wizards, in large organisations most users and VPs should not be allowed to create formal reports themselves.
The two reports most likely to mislead
Building on the issues above, the common thread of the big misleaders is they try to do too much, drawing conclusions about several linked business processes. Because while the processes are linked, the information about them is disjointed (or even unavailable).
The first classic misleader is the sales activity report that measures how many calls, meetings, emails and other activities each rep does in a sales cycle. While there's nothing wrong with measuring these things for a telesales group, this kind of detailed metric only works for the highly regularised communications and interactions of a call centre. The moment you apply this level of measurement to a field sales rep, two difficulties arise:
- The reps try to understand what the boss is looking for, and then "game the system" so they look good according to those measurements only. They do less of the things that actually make them effective sales reps. All too often, the things that really matter (such as the quality of their interactions or the style of their negotiation) are hard to capture in any kind of automated system.
- Management tries to model "what's the difference in the activities of the successful rep vs the average one?" While understanding the physics of your sales cycle is critical, the Heisenberg Uncertainty Principle takes its toll on the accuracy and relevance of measurements in the "soft skills." So the model you develop is unlikely to be realistic basis for management direction.
The grandaddy of misleaders is the "marketing investment/ROI" report. CEOs, marketing VPs and sales VPs all want to know "how much revenue do we get from this kind of marketing?" So they ask for a report showing recent marketing events, the leads they create, the pipeline that results and the closed business over time. A reasonable request that CRM systems can't answer (at least for B2B companies).
The reason? In an Enterprise sale, you frequently don't know the length or phases of the sales cycle from the customer's perspective. You don't know which people actually made the deal go over the goal line. And you don't know which particular marketing "touch" motivated the prospects to take action. Of course, you can build models and scoring systems to extrapolate the customer's mental state from their behavior, but these models are usually nothing more than a pile of untested assumptions.
Consequently, "pipeline impact" reports using these models yield highly variable results that dont hold up to scrutiny. With stringent assumptions, marketing may be seen to impact only 5 percent of revenue, and with looser ones, marketing may be seen to impact more than 100 percent of revenue (yes, it's possible!)
The answer to these big misleaders? About all you can do is produce smaller, more confined reports that are valid — and remind the requesters that no CRM system can read minds.
David Taber is the author of the new Prentice Hall book, "Salesforce.com Secrets of Success" and is the CEO of SalesLogistix, a certified Salesforce.com consultancy focused on business process improvement through use of CRM systems. SalesLogistix clients are in North America, Europe, Israel, and India, and David has over 25 years experience in high tech, including 10 years at the VP level or above.