Nine Data Quality Problems

Aaron Reese: 10th October 2025

Introduction

Housing organizations collect huge amounts of data about tenants, properties, repairs, and business operations. This data comes from formal systems (Housing Management, Repairs, Finance, HR), emails, spreadsheets, and documents. Making sure all this data is correct, up-to-date, and useful is a constant challenge.

This guide explains nine common data problems with examples from tenant records, property management, and repair jobs. While we suggest ideal fixes, we know that many organizations can't change their main computer systems. That's where a Data Quality System like Project Overwatch helps - it can spot and track these problems even when your main systems can't be updated to prevent the issues occuring in the first place.

1. Missing Information

Sometimes important information is missing from records. This could be:

  • A tenant's birth date or emergency contact
  • Property energy ratings or safety certificates
  • Repair completion dates or contractor details
  • Health conditions requiring special property adaptations or interaction methods

Ideal fix: Set up your computer systems to require important information before saving a record. Train staff to check for missing details and give them context as to why the information is needed.

When you can't change your main system: A Data Quality System can scan your data on a regular basis, create reports of missing information, and send alerts to the right teams for follow-up. Speed is important because other processes may depend on this information to work correctly.

Robot Scratching Head

2. Conflicting Information in Different Systems

Robot looking at conflicting data

The same information might be stored differently across multiple systems:

  • Tenant phone numbers different in housing vs repairs systems
  • Property addresses formatted differently.
  • Contractor contact details varying between finance and repairs systems

Why this happens: Information gets entered at different times but there is no automatic update between systems, or the information is captured for one purpose and then used for something else without checking it is still correct (e.g. temporary phone numbers for a one-off repair).

Ideal fix: Use one main database where possible. Set up automatic data sharing between systems. Bring data into a data warehouse for reconcilation and push updates back to source systems.

When systems can't be integrated: A Data Quality System can compare information across different systems, highlight conflicts, and track which discrepancies have been resolved.

3. Information That Doesn't Make Sense When Combined

Sometimes individual pieces of information look OK, but they don't work together:

  • A repair job that ends before it starts
  • Sheltered accommodation with a tenant marked as "Market Rent"
  • A contact with a tile of "Mr." but a gender as "Female"

Ideal fix: Set up system checks that compare related information when data is entered.

When validation rules can't be added: A Data Quality System can run these checks regularly on existing data, flag conflicts for review, and track resolution time and trends.

Robot juggling data

4. Information That Could Be Right But Seems Strange

Robot comparing data

Some information might be technically possible but very unusual:

  • A tenant under 18 years old as the time of signing the tenancy
  • A repair job costing £50,000 for a simple task
  • A property vacant for over 2 years
  • A medical condition or support need lasting much longer than typical

Ideal fix: Build review workflows into your main systems with supervisor approval for unusual cases.

When workflow changes aren't possible: A Data Quality System can flag unusual cases for review, allow staff to mark them as "confirmed correct," and suppress future alerts for verified exceptions.

5. Information That Becomes Wrong When Other Information Changes

Data records connect to each other. When you change one record, it might make another record incorrect or break a business rule:

  • Deleting someone's "main" phone number without setting a new main number
  • Marking a property as "disposed" but leaving active tenancy records
  • Removing a contractor but leaving them assigned to ongoing repair jobs

Ideal fix: Set up automatic rules that update related records when changes are made.

When automatic rules can't be implemented: A Data Quality System can identify broken relationships between records and create cleanup reports for manual review.

Robot Scratching Head

6. Information That is Time Sensitive

Robot looking at his watch

Some information has dates attached. What was correct yesterday might be wrong today:

  • A "provisional" tenancy that should have started by now
  • Safety certificates that have expired
  • Repair jobs still marked as "in progress" that are now outside the Service Level Agreement levels
  • Tenant's vulnerability warning for a broken leg 6 months ago.

Ideal fix: Set up automatic status updates and deadline alerts in your main systems.

When automatic updates aren't available: A Data Quality System can run frequent checks for date-related problems and send alerts to the right teams when action is needed.

7. Old Information That Hasn't Been Updated, Deleted or Verified

Life moves on and we don't always get to hear about it. There are many examples of data that is collected once but needs a periodic review to ensure it is still correct. Often your systems will have a capture or change date but won't be able to tell you the last time that informations was verified if it didn't change:

  • Tenant phone numbers, addresses, and income
  • Property condition ratings and energy efficiency scores
  • Contractor contact details and certification/insurance status
  • Family size and health conditions requiring property adaptations or bedroom allocations/li>

Ideal fix: Schedule regular check-ins and set automatic review reminders in your systems.

When regular reviews can't be automated: A Data Quality System can identify records that haven't been updated for set periods and generate review lists for different teams. It can act as an external control to log that the review has been completed and suppress the warning for a period of time.

Robot looking at crumpled data

8. Information That Becomes Illegal/Invalid to Hold Due to Changes in Law or Business Needs

A Robot putting data in the bin

Laws and the the business environment change over time. Information you used to collect might now be:

  • Illegal to keep (credit card numbers)
  • No longer required for compliance or reporting (ethnicity and gender after completing CORE returns)
  • Require a different level of granularity (permission to contact)
  • Outdated due to changes in social housing requirements

Ideal fix: Build compliance checks into your systems that automatically flag or remove outdated data.

When systems can't be updated: A Data Quality System can identify data that may no longer comply with current regulations and create action plans for review and cleanup across multiple systems.

9. Information You No Longer Need

Some information is only needed for a specific task. Once that task is done, the information should be deleted:

  • Bank details collected for one-time refunds
  • Temporary contact numbers for repair appointments
  • Names of council officers in closed ASB cases
  • Contractor payment details after job completion
  • Demographic data for statutory returns

Ideal fix: Set up automatic deletion schedules and workflows in your main systems.

When automatic cleanup isn't possible: A Data Quality System can identify temporary information that should be removed, create deletion schedules, and track compliance with data retention policies.

A retired robot with a cane

Key Takeaways

Prevention is Best

  • Set up good data entry rules
  • Train staff properly: Explain why the data is important
  • Use automatic checks where possible

When Systems Can't Be Changed

  • Use a Data Quality System to monitor
  • Create regular reports and alerts
  • Track problems and resolution times

Fix Problems Quickly

  • Review data quality reports regularly
  • Assign clear ownership for different data types
  • Measure and improve resolution times

Remember: Many organizations can't significantly modify their main housing, repair, or financial systems. A dedicated Data Quality System like Project Overwatch works alongside your existing systems to identify, track, and help resolve these common data problems without requiring expensive system changes.