The Translation Problem
"We have a 23% null rate in the customer_segment field and significant referential integrity violations in the order_to_customer join" is technically accurate and completely useless for getting a business leader to prioritize fixing it.
The same problem, translated: "One in four customers in our database isn't categorized, so when you ask 'how much revenue comes from enterprise customers,' we're missing data on 25% of your customer base. The report you saw last week was based on 75% of customers, not all of them."
Same problem. One drives action. One doesn't.
The Framework for Business-Language Translation
For every data quality issue, prepare three things:
1. The business impact statement What decision or process is affected? What is the consequence if the quality problem persists?
Sohovi profiles your datasets for quality issues in minutes — see what's broken before it breaks your pipeline — try Sohovi free.
"Our email campaign to lapsed customers is sending twice to 15% of our list, costing us double the send fees and generating unsubscribes from customers who find it annoying."
2. The quantified scale How many records are affected? What % of the relevant population? What is the frequency and trend?
"15,000 contacts out of 100,000 (15%). This has been growing — it was 9% six months ago."
3. The fix and its cost What would remediation involve? What is the estimated effort? What would quality look like after the fix?
Sohovi gives you the data quality picture you need to make the case for fixing it — and to track improvement over time.
"A one-time deduplication project would take approximately 20 hours and reduce the duplicate rate below 2%. Estimated cost: $2,000 in analyst time."
Structuring the Conversation by Stakeholder Type
For finance/CFO: Lead with cost. "This quality problem costs approximately $X per month in [wasted spend / labor / lost revenue]."
For sales/revenue leaders: Lead with pipeline impact. "This quality problem means our CRM shows $Y in pipeline that isn't real, affecting forecasting accuracy."
Sohovi scores your dataset against your own accuracy standards and highlights the columns and rows where values fall outside expected ranges.
For operations/COO: Lead with efficiency. "This quality problem causes Z hours of rework per week across the team."
For marketing: Lead with performance impact. "This quality problem is reducing our email deliverability by X% and our conversion rate by Y%."
The One-Page Issue Summary
For any quality issue requiring stakeholder action, prepare a one-page summary:
- What is the problem? (1 sentence)
- What is the business impact? (2–3 bullets)
- How many records / what % of data is affected?
- What does fixing it involve and what does it cost?
- What's the risk of not fixing it?
This format works for an email, a Slack message, or a 5-minute meeting agenda item. It gives stakeholders what they need to decide without more information than they need.
