Skip to main content
Data Quality Dimensions

Data Consistency: Why the Same Information Means Different Things in Different Systems

Inconsistent data produces contradictory reports and broken integrations. Here's what data consistency means, where inconsistencies come from, and how to detect them.

Key Takeaways
  • Intra-dataset consistency (internal rules) and cross-system consistency are different problems requiring different solutions
  • Failed syncs and manual overrides are the most common causes of cross-system inconsistency
  • Different format standards (phone with vs. without hyphens) prevent system matching even when values are semantically the same
  • Cross-system comparison — exporting the same entity from two systems and comparing field by field — is the most direct detection method
  • Consistency violations compound: a 2-day sync failure creates a 2-day window of contradictory reports

The Two Types of Consistency

Data consistency has two related meanings:

Intra-dataset consistency: Within a single dataset, values should be internally consistent. An order with a ship date before its order date is internally inconsistent. A total that doesn't equal the sum of its line items is inconsistent. These violations are logical contradictions within the same file or database.

Cross-system consistency: The same entity should be represented consistently across systems. A customer's email in your CRM should match their email in your email platform. A product's price in your ERP should match its price in your e-commerce store.

Sohovi finds gaps, duplicates, and format errors in your CRM data — so your team is working from records they can trust.

Both types cause problems. Cross-system inconsistency is typically harder to detect and more disruptive — because by the time you discover it, multiple systems and the reports built on them have already diverged.

How Cross-System Inconsistencies Happen

Asynchronous updates: A customer updates their email in your e-commerce store. The CRM is synced nightly. For 24 hours, the two systems disagree. If a support ticket is opened during that window, it may be routed to the wrong record.

Failed syncs: The nightly sync failed on Tuesday because of a timeout. The CRM is now two days behind. Every system that relies on the CRM has stale data — and unless someone is monitoring for failed syncs, no one knows.

Sohovi tracks quality trends across runs and alerts you when a metric — null rate, duplicate count, score — moves outside its normal range.

Different formatting rules per system: The CRM normalizes phone numbers to 10-digit format. The marketing platform stores them as entered. Now "5551234567" and "555-123-4567" look different to any system trying to match them — even though they're the same number.

Manual overrides: A sales rep updates the account name in the CRM. The billing system still has the old name. Reports pulling from both systems now show two different names for the same customer. Finance can't reconcile the numbers.

Migration errors: During a platform migration, customers and orders are migrated in separate batches. Some customer records fail to migrate due to data errors. The orders that referenced them are now orphaned — pointing at customer IDs that no longer exist in the new system.

Sohovi validates your dataset before it enters the warehouse — catching format errors, nulls, and duplicates at the source.

The Business Consequences of Inconsistency

Inconsistency problems tend to surface at the worst possible times: during a board presentation when two reports show different revenue numbers, during a client renewal call when your team has different data than the client's portal, or during a regulatory audit when records across systems don't match.

The consequences aren't just embarrassing. They're operationally expensive:

  • Duplicate customer outreach: If your CRM and email platform have diverged, you may reach out to the same customer twice with different messages — or miss them entirely if the records don't match.
  • Wrong pricing: A price update applied in one system but not another creates customer-facing discrepancies. A customer who checked the price in your catalog finds a different price in their checkout.
  • Inaccurate reporting: Any report that joins data from two inconsistent systems produces numbers that can't be trusted. Decisions made on those reports carry hidden risk.

Detecting Inconsistencies

Cross-system comparison: Export the same entity from two systems (same customer ID, same product SKU). Compare field by field. Any difference is a consistency violation. Start with your highest-value records — top customers, top products — and work down.

Internal consistency rules: Write rules that should always be true: order_date ≤ ship_date ≤ delivery_date. Calculate revenue as units × price and compare to recorded revenue. Check totals against their components. Any record that violates a rule is inconsistent.

Change log audits: If a field has been updated in System A but not System B since the last sync — and the sync is supposed to keep them aligned — you have a consistency problem. Most integration tools log sync errors; many teams never check those logs.

Regular cross-system spot checks: Once a month, manually compare five to ten records across your most critical systems. It's low-tech but surprisingly effective at catching drift before it becomes systemic.

Sohovi lets you upload a CSV export and instantly profile it for internal consistency violations — values that contradict each other within the same file. No setup, no code required.

Measuring Consistency

For cross-system consistency:

Consistency = (Records consistent across systems / Total records checked) × 100

For internal consistency:

Consistency = (Records satisfying all internal rules / Total records) × 100

A score of 100% means no violations found. In practice, most systems that haven't been actively monitored will score between 90-98% — which sounds high until you calculate what the remaining 2-10% means for a 50,000-record database.

Building Consistency Into Your Processes

Consistency problems are much cheaper to prevent than to fix after the fact. A few practices that prevent the most common causes:

Monitor your syncs: Set up alerts for failed integration jobs. A failed sync that goes unnoticed for a week creates a week of inconsistency to untangle.

Standardize formats at entry: Enforce consistent phone number and address formats when data enters your systems. Format differences are one of the leading causes of failed cross-system matching.

Document your canonical source: For each important data field, define which system is the "source of truth." When two systems disagree, the source of truth wins. This doesn't prevent inconsistencies, but it makes resolution faster and less political.

Run consistency checks before reporting: Before pulling a report that joins data from multiple systems, run a quick check to verify the key fields that the join depends on are consistent. Catching a consistency problem before the report runs is far better than discovering it after the report is already in front of leadership.

If you're ready to get a clear picture of where your most important data is inconsistent, Sohovi will walk you through an instant data quality report. Upload your CSV free — no credit card, no code, no data leaves your machine.

Frequently Asked Questions

How do I choose a 'source of truth' when two systems disagree?

Define the authoritative source for each data domain before a conflict occurs. Typically: the system of record for that domain (CRM for customer data, ERP for financial data, warehouse system for inventory). Document this decision and communicate it across teams.

What's the difference between consistency and accuracy?

A value can be consistent across systems but consistently wrong. Consistency means systems agree; accuracy means they correctly represent reality. Both can be violated independently. Consistency problems are usually easier to detect than accuracy problems.

Can I automate consistency checks?

Yes. Schedule a nightly job that joins key tables across systems and flags records where values differ. Tools like dbt (data build tool) support data testing including cross-system consistency checks. For smaller teams, a scheduled SQL query or Python script achieves the same result.

Sohovi Team

Data quality, for people who ship

The Sohovi team writes practical guides on data quality, profiling, and governance to help teams ship better data.

Start for free

Stop guessing. Start knowing your data quality.

Sohovi profiles your datasets in minutes — surfacing completeness gaps, type mismatches, and duplicate patterns before they reach production.

No credit card required · Free forever plan