The Clearly Podcast

Andy: Hello, guys. How are we doing today?

Tom: Good, thanks, Andy. How are you?

Andy: Doing just fine, thanks. Today's podcast topic is data quality. We discussed earlier that the main takeaway is that quality is essential. If it's important, do it right. I've been Andy Clark, thanks for listening, goodbye.

Tom: We should have started with a spoiler alert: data quality is vital.

Shailan: Yeah.

Andy: We could have wrapped this podcast in under 30 seconds.

Tom: Though in reality, we often see poor data quality.

Andy: Good point. Spoiler alert: data quality is often bad. We think it should be good. Thank you, and goodbye.

Tom: We can be lazy at times.

Andy: True, but with important tasks, like timesheets, I'm diligent because it affects billing.

Tom: You put effort into timesheets to save effort later, which highlights the importance of data quality.

Andy: Yeah. When engaging with customers, what do you usually expect regarding data quality?

Tom: Issues often arise from user input. If there's no consequence for not filling out a field, users won't bother.

Andy: I remember reconciling invoices for inventory purposes pre-COVID. Missing narratives made it nearly impossible to understand transactions. Over-engineered applications also create issues by mandating too many fields.

Tom: Mandatory fields, especially drop-downs, often lead to users selecting the first option just to move on.

Andy: Or "Other" with a text field. It's a balancing act.

Tom: True, and "Other" in drop-downs is problematic.

Andy: "Other" in pie charts is hilarious.

Shailan: And null values.

Andy: What's your advice when talking to customers about data quality?

Shailan: We need to monitor data quality continuously. Dashboards can help track and address issues.

Andy: We often say fixes should be at the source, but it gets ignored.

Shailan: A client recently preferred asking us for data over their internal finance, indicating a data quality issue.

Tom: Dashboards often reveal bad data, driving the conversation on data quality.

Andy: Seeing data quality issues affect reports motivates people to improve.

Shailan: Dashboards showing completeness of fields can help change behavior.

Andy: How do you get organizations to value data entry as much as the final reports?

Shailan: Training on the impact of poor data and allowing time to correct errors can drive behavior change.

Andy: Professional pride should prevent people from having to fix their work later.

Shailan: In some places, allowing time to correct errors motivated better initial data entry.

Andy: Shaming people into good behavior can work.

Tom: If people know their data will be used in high-level reports, they're more likely to enter it correctly.

Andy: Assigning data ownership can help with quality, but it depends on the individual.

Tom: Yes, and hiding data quality issues in analytical platforms prevents improvement.

Andy: Have you seen organizations with consistently high data quality?

Tom: Some take it seriously and don't hide issues, but perfection is rare.

Andy: Data quality requires daily diligence, like maintaining a diet.

Tom: Exactly. Treating data quality as an event rather than a continuous practice leads to yoyo dieting.

Shailan: I've seen sales teams maintain high-quality CRM data because they own it and understand its importance.

Tom: Sales are usually the worst, but exceptions exist.

Andy: What are some data horror stories from poor data quality?

Tom: One example is mismatched goods received notes and invoices leading to payment issues. Another is covering up data issues in analytics, which gets exposed in new platforms.

Andy: Poor data quality costs money and hinders accurate reporting.

Shailan: Data migration often involves fixing quality issues, which is time-consuming.

Andy: Making exceptions in business processes can lead to long-term costs.

Tom: Quality should be maintained at the source, not just during migrations.

Andy: Absolutely. Fixing issues only during migrations doesn't solve ongoing problems.

Shailan: Assigning data ownership helps, but only if individuals take it seriously.

Andy: Data quality is a daily practice. We need to focus on it every day to maintain standards.

Tom: Agreed. It's a continuous effort, not a one-time fix.

Andy: Alright, guys, thanks for the great conversation. Next week, we'll discuss long-term data maintenance and cost of ownership.

Tom: Sounds good.

Andy: Thanks, everyone. Have a great week.