Fixing 100,000 Incorrect Addresses in a Bullhorn Database: A Case Study

The Hidden Cost of Bad Data

When recruiters look to Bullhorn to support sourcing efforts or efficiently process inbound applications, they need the information in each record to be accurate. But what happens when the data itself is unreliable or completely wrong??

A Bullhorn customer discovered that a misconfiguration with a job board had injected bad data into their system over several years, corrupting over 100,000 candidate addresses. Instead of having accurate location data, all of these candidates were assigned the same address in Texas—even if they lived in completely different states or countries.

This data issue wasn’t just an inconvenience; it blocked recruiters, automation tools, and sourcing strategies from working effectively.

How This Happened: The Root Cause

The problem stemmed from a job board integration that had been incorrectly configured when it was first set up. Whenever a candidate applied for a job through this board, the system assigned them a default address—an address in Texas—if their actual location wasn't parsed from their resume.

Over time, this issue snowballed. Tens of thousands of candidates from across the country had the same incorrect location. Often, these root causes are not immediately known, and once discovered, they can trigger an unexpected fire drill—especially when recruiters need to start sourcing from that location and have time-sensitive candidate submission requirements.

The impact?

• Any recruiter or automation tool sourcing for candidates in Texas was flooded with irrelevant results—blocking them from reaching actual local candidates.
• Search and sourcing efforts using location-based filters were completely unreliable.
• Compliance risks arose because location accuracy matters for tax, employment laws, and remote work classification.

The Scale of the Problem: Time and Cost Impact

Fixing these records manually wasn’t an option. Let’s break down why:

• Through analysis, they determined that correcting a single record—finding the right address, validating it, and updating it in Bullhorn—would take between 3 to 5 minutes per record.
• With 100,000 bad records, that meant:
  • 300,000 to 500,000 minutes of manual work
  • That’s 5,000 to 8,333 hours
  • Or roughly 2.5 to 4 full-time employees working for a YEAR

Even if a company outsourced this data cleanup at $20 per hour, they’d be looking at a $100,000 to $160,000 cost just to fix the mess.

How We Fixed It Efficiently

Instead of throwing time and money at a manual process, we deployed an automated, structured approach that restored Bullhorn’s data integrity without disrupting operations.

1. Bulk Identification & Issue Isolation

• We analyzed all candidate records in Bullhorn to identify which records contained incorrect addresses.
• We determined whether other information within the record or resume could provide an accurate location.
• If no location data was available, we sourced the correct address using external verification services.

2. Data Enrichment & Validation

• For candidates with resumes in Bullhorn, we extracted address details from the original application.
• For candidates without address information, we used external data services to enrich missing fields based on available contact details.

3. Automated Updates at Scale

• Once validated, we pushed corrected addresses back into Bullhorn in bulk, ensuring that recruiters and automation tools immediately had access to clean, usable data.

Results: Immediate and Measurable Improvements

• 100,000 records were fixed without any manual entry.
• Recruiters searching for Texas-based candidates finally got relevant results—restoring sourcing efficiency.
• Automated screening tools and job-matching algorithms started working correctly again.
• The company saved over $100,000 in labor costs compared to manual fixes.
• Bullhorn became a trusted data source again—improving recruiter confidence in searches.

Key Takeaways: Why Data Quality Matters

This case highlights a common but overlooked issue in recruiting databases: small misconfigurations can create massive problems over time.

• Bad data blocks automation. If a sourcing tool can’t trust location data, it can’t do its job.
• Bad data costs money. A seemingly minor integration issue led to a six-figure cleanup cost—avoided only through automation.
• Proactive data maintenance is essential. Regular audits and automated enrichment prevent problems like this from compounding over years.

Recruiters rely on accurate data to source and place talent. Without structured data management, issues like these will surface—and they will impact revenue.

‍