As you might imagine, we get quite a few questions through the subscriber Help Desk about the underlying data people are using to perform analyses and inform sales decisions. One of the more common refrains goes something like this:
“Isn’t using dirty sales data worse than having no data at all?”
Of course, there are no absolutes and the safest consultant-speak answer would be, “well…it depends.” In the past, we’ve repeatedly illustrated how the “dirty data” excuse doesn’t actually hold water in most cases and how you can in fact glean a lot of valuable insight from relatively dirty data.
That being said, however, there are situations where dirty data can actually be worse than having no data at all. Not because the data is incomplete or it’s a little messy. But because it actively pushes people toward the wrong conclusions, the wrong priorities, or the wrong behaviors…all with the “air of precision” and the false sense of confidence that numbers tend to engender.
And false confidence is far more dangerous than uncertainty.
A good example can be found in long-term forecasting. If your data is incomplete, inconsistent, or selectively maintained, your forecasts aren’t just “a little off.” They are structurally misleading. They systematically overstate the degree of certainty, understate the amount of risk, and create the illusion that the business is far more predictable than it really is.
That illusion can then work its way into hiring plans, inventory commitments, marketing investments, and executive expectations. Decisions get made as though the numbers are solid when…in reality…they’re molded out of sand. And as a result, the organization becomes less cautious, less adaptive, and far less prepared for the potential downsides.
In this situation, having no long-term forecast at all would at least force people to acknowledge the uncertainty and manage it thoughtfully. Dirty projections, on the other hand, can hide massive uncertainties behind an impressive-but-flimsy façade of charts, graphs, and numbers.
Another example comes from territory and segment analysis. If account data is riddled with duplicates, outdated firmographics, missing hierarchies, and inconsistent definitions, then your segmentation models won’t just be fuzzy…they’ll be inherently biased.
You’ll start to “see” patterns that aren’t actually there. Certain segments will look unprofitable simply because some cost data is misallocated. Certain reps will appear to be underperforming because their accounts aren’t properly attributed or their product mix isn’t being taken into account. And certain industries will look attractive just because they’re over-represented in the “untapped universe.”
The big problem here is that these biases and distortions don’t tend to stay contained in the analytics. Instead, they get operationalized. Resources get reallocated. Coverage models get changed. Incentives get adjusted. And now the business is actively optimizing against a totally warped pictured of reality.
At that point, dirty data isn’t just unhelpful…it’s actively steering the organization in the wrong direction.
A third example appears when data is used to evaluate people rather than processes. Sales Ops teams are often asked to use data to assess rep performance, compliance, pipeline hygiene, or conversion effectiveness. But when the underlying data is inconsistent, selectively entered, or “politically managed,” the analysis quietly shifts from measuring behavior to measuring gamesmanship.
Reps that are diligent and disciplined about data entry actually get penalized because their pipeline looks more realistic. Reps who know how to game the system look better than they actually are. Managers learn which numbers are watched and adjust their reports accordingly. Over time, the data stops being a reflection of reality and becomes a reflection of various incentives.
At that point, the data is no longer a diagnostic tool…it’s a behavioral distortion field.
Again, in that environment, having no data might be less harmful than having data that quietly trains the organization to lie to itself, day in and day out.
None of this means that dirty data is inherently bad or unusable. Again, in most cases imperfect data can still be extremely valuable for identifying directional trends, relative differences, and broad patterns…as long as everyone involved understands its limitations and treats it accordingly.
When dirty data is treated as precise, complete, and authoritative when it really isn’t, that’s when it stops being a decision aid and starts being a decision liability.
So…the goal for Sales Ops is not to demand pristine and perfect data before doing anything useful. That’s unrealistic and unnecessary. No, the goal is to understand where the data is flimsy, where it’s biased, where it’s incomplete, and where it’s being operationalized in ways that could cause harm.
If you know that your forecast is noisy, you can use it cautiously. If you know that your segmentation model is approximate, you can avoid over-optimizing against it. And if you know that your performance metrics are gameable, you can design them more thoughtfully.
Dirty data becomes dangerous not because it’s dirty, but because people forget that it’s dirty!
Used with humility, imperfect data can still help you ask better questions, test smarter hypotheses, and see your operation more clearly than you otherwise could. But used uncritically, it can quietly convince you that you understand your business far better than you actually do.
And in Sales Ops, that particular mistake tends to be far more damaging than simply acknowledging that the data isn’t perfect and that a grain of salt…or two… might be warranted.
Working With "Bad" Sales Data
In most businesses, there are so many moving parts that working with perfect data just isn't very realistic. In this session, learn a strategic approach to making meaningful progress with inaccurate and incomplete datasets.
Delivering Data to Decision-Makers
Providing sales data to executives and other decision-makers is a core responsibility for most Sales Ops teams. In this session, learn what leading teams are doing to make their efforts in this area far more effective.
Building a Data-Driven Sales Operation
Building a data-driven sales operation where anecdotes and opinions have ruled for years is no easy task. In this on-demand webinar, learn about transforming your sales processes and team culture to make much better use of data.
