Most property tax firms hit the same wall. The pipeline fills up, the deadline gets closer, and the only way to take on more properties is to hire more people. So growth stalls at whatever the team can physically file by hand.
One Texas property tax protest firm broke through that wall. It went from 700 properties to more than 20,000 parcels, raised average revenue per property from $200 to $500, and filed its entire peak season with zero technical errors. It didn’t hire its way there. It rebuilt the appeal process around one idea: evidence should be built automatically, and built for the specific appraisal district reviewing it.
We know how it worked because we built the system alongside them. This is the inside account of what changed, in what order, and why it worked.
The Short Answer
The firm’s revenue didn’t grow because it found a better sales channel. It grew because three constraints were removed, one after another:
- Data. Missing and incomplete property records were replaced with a verified baseline pulled directly from appraisal district (CAD) websites.
- Throughput. Every manual step, from lead creation to filing to invoicing, became part of one automated pipeline.
- Evidence quality. A proprietary comparable-analysis tool built CAD-specific evidence packages automatically, which drove bigger reductions and higher fees per property.
Remove only one of these and a firm grows a little. Remove all three and the business stops being limited by headcount.
The Starting Point: A Fully Manual Firm in a Moving Market
Before the rebuild, this firm looked like most growing Texas protest shops. The team was capable, demand was there, and nearly everything ran on people’s time.
The firm faced four core problems:
- Incomplete portfolio data. Large gaps in property records and missing key fields weakened the firm’s position before a case was even built. You can’t argue a value down when you don’t have the full picture of the property.
- Manual bottlenecks at every stage. Lead creation, evidence building, and filing were all done by hand. Every new property added hours, not minutes.
- Volatile CAD environments. Appraisal districts changed how they shared data and calculated values, often without notice. Any workflow built on last year’s format could break mid-season.
- Fragmented client communication. Property owners got updates through manual check-ins, and managers had no single place to see where each protest stood.
That third problem is specific to Texas and is often underestimated. Texas has an appraisal district in each of its 254 counties. Each one publishes data differently, evaluates evidence differently, and has its own history of what reductions it will actually approve. A standardized process that works in one county can quietly underperform in the next.
Why Generic Automation Wasn’t Enough
The obvious fix would have been off-the-shelf property tax appeal software: pull comps, generate a packet, file. The firm had seen that approach, and it runs into the problem we documented when we analyzed 10,000 property tax appeals: about 40% of winnable cases get rejected because the evidence is generic, not because the facts are weak.
As our CTO put it when we published that analysis:
“A generic strong case and a CAD-specific strong case can use the exact same property data and get completely different results.” — Sargis Ghazaryan, Chief Technology Officer
Automating a generic process only means filing weak evidence faster. The firm needed automation that worked with each CAD’s own standards, not around them.
“The turning point was watching strong cases come back rejected from software-built packets. The tool did its job. It just didn’t know who it was arguing to.” — Vlad Evoyan, Head Project Manager
How the Appeal Process Was Rebuilt, Phase by Phase
The rebuild didn’t happen in one launch. It was sequenced so each phase removed the bottleneck that would have limited the next one.
| Phase | Focus | What changed |
|---|---|---|
| 1 | Core infrastructure | Mapped every manual workflow, audited missing data, built the first CAD data scraper |
| 2 | Core automation | Automated filing pipeline, live before peak season, starting from 700 properties |
| 3 | Evidence engine | Proprietary comparable-analysis tool; 100% of core evidence packages automated |
| 4 | Yield optimization | Tuned evidence to outperform off-the-shelf tools; revenue per property moved from $200 to $500 |
| 5 | CRM and communication | Client sign-up flow, centralized CRM, exception handling for outlier properties |
| 6 | Enterprise hardening | Scaled to 20,000+ parcels, full-pipeline stress testing, 0% technical error rate |
Phase 1: Fix the Data Before Automating Anything
The first instinct in most automation projects is to automate filing. We started with data instead, because every later step depends on it.
We deployed adaptive scrapers that pull property records directly from CAD websites. They filled the gaps in the firm’s portfolio and created a verified baseline. “Adaptive” is the important word. When a district changes its site structure or data format, the scrapers are updated so the pipeline keeps running rather than failing silently in the middle of protest season.
Lesson for other firms: if your property data has holes, automation will spread those holes across every case.
Phase 2: Automate the Pipeline End to End
With reliable data in place, we built the core pipeline: lead creation, filing, completed protests, and invoiced outcomes in one connected flow. It went live ahead of peak tax season so the firm could handle its first surge of volume without adding staff.
This phase also introduced system-wide logging. For the first time, every property’s status lived in one place instead of in spreadsheets and inboxes.
Phase 3: Build Evidence the Way Each CAD Evaluates It
This is the phase that changed the economics.
We built a proprietary comparative analysis tool that lets agents evaluate properties side by side and build targeted cases. More importantly, it tracks each appraisal district’s calculation methods in real time and adjusts the evidence as those methods change.
The result: 100% of core evidence package aggregation became automatic. Analysts stopped assembling packets and started reviewing and negotiating them.
This is where Vlad Evoyan, our Head Project Manager, saw the difference in casework first:
“Once we looked closer, the comps were technically accurate but not the comps that specific reviewer wanted to see. It wasn’t a data problem. It was a personalization problem.”
Automated, CAD-aware evidence solved the personalization problem at a volume no manual team could match.
Phase 4: Turn Better Evidence Into Higher Revenue per Property
Better evidence doesn’t just win more cases. It wins bigger reductions, and most protest firms charge a percentage of the savings they secure.
We refined the evidence tool until it consistently outperformed the off-the-shelf software the firm had compared it against, and stress-tested it on complex, multi-parcel valuations. Average revenue per property rose from $200 to $500, a 150% lift per unit, without changing the fee structure. The savings were simply larger.
Phase 5: Keep Clients Informed Without Manual Check-Ins
With volume growing quickly, client communication became the next bottleneck. We built a custom sign-up and onboarding flow connected to a centralized CRM that syncs property data with automated client updates.
We also added exception-handling rules. Outlier properties that don’t fit the standard model are automatically sent to a person for manual review. That’s what makes high automation safe: people handle the exceptions, not the routine.
Phase 6: Harden the System for 20,000+ Parcels
The final phase focused on reliability at scale. We stress-tested the full pipeline end to end, scaled the infrastructure to process a 20,000+ parcel portfolio, and put continuous adaptation procedures in place for ongoing CAD changes.
During peak season, the system recorded a 0% technical and filing error rate across all 20,000 property protests. In a business where one missed deadline or formatting error can forfeit a client’s savings for the year, that reliability matters as much as the growth.
The Results
| Metric | Before | After |
|---|---|---|
| Properties managed | 700 | 20,000+ (2,757% growth) |
| Average revenue per property | $200 | $500 (150% lift) |
| Technical and filing error rate | Manual risk every season | 0% across 20,000 protests |
| Core evidence packages built automatically | 0% | 100% |
Together, those results produced the 20x revenue growth in this case study’s title. The more important point is how it happened: the firm didn’t need a proportionally larger team to absorb the growth. Manual review is now reserved for exceptions, and analysts spend their time on strategy and negotiation.
In the client’s own words:
“From missing data and fully manual files to a nearly fully automated pipeline — this represents five years of continuous engineering built around real, explosive growth. Our platform automatically aggregates and builds 100% of the core evidence packages, allowing our tax analysts to focus entirely on final strategy and negotiation rather than manual data entry.”
What Other Firms Can Learn From This Case Study
You don’t need 20,000 parcels for these lessons to apply. Any firm trying to scale a property tax appeal practice runs into the same constraints, just at a different size.
1. Fix data before you automate. Automation only moves problems faster when the inputs are incomplete. A verified property baseline comes first.
2. Revenue per property is a lever, not a fixed number. Most firms focus only on adding properties. This firm got a 150% lift from each existing property through better evidence alone.
3. CAD-specific beats generic, at scale. Tailoring evidence to one appraisal district by hand is manageable for ten cases. Doing it for thousands requires a system that tracks each CAD’s behavior automatically.
4. Build for change, not for this year. Texas appraisal districts change formats and methods. A system that can’t adapt quickly will fail at the worst possible time: peak season.
5. Automate the routine, escalate the exceptions. The goal isn’t to remove people. It’s to make sure people only spend time where their judgment actually changes the outcome.
What This Means If You’re Evaluating Property Tax Appeal Software
If you’re comparing tools for Texas property tax appeal automation, ask these questions of any vendor:
- Does it build evidence for the specific CAD reviewing the case, or does it apply one method everywhere?
- What happens when an appraisal district changes its data format mid-season? Who fixes it, and how quickly?
- Can it show a measurable lift in revenue per property, not just time saved?
- What is its error rate at peak volume, and is that number measured across a real season?
- Was it built inside a working protest firm, or designed without that operating experience?
The last question matters more than it seems. Every feature in this system exists because someone on a real team needed it on a real deadline.
Frequently Asked Questions
How did the Texas firm achieve 20x revenue growth?
By removing three constraints in sequence: incomplete property data, manual filing workflows, and generic evidence. Volume grew from 700 to 20,000+ properties, and average revenue per property rose from $200 to $500, without proportional hiring.
What is CAD-aware evidence automation?
It’s evidence generation that accounts for the standards, calculation methods, and approval history of the specific county appraisal district reviewing a protest, rather than applying one template to every jurisdiction.
Can property tax appeal software really reach a 0% error rate?
In this case study, the system recorded zero technical or filing errors across 20,000 property protests during peak season, supported by end-to-end stress testing and exception handling for outlier properties.
Does automation replace property tax analysts?
No. Automation handles data collection, evidence aggregation, and filing. Analysts focus on final strategy, negotiation, and the exception cases that need human judgment.
Is this approach only for large firms?
No. The constraints of incomplete data, manual bottlenecks, and generic evidence affect firms of every size. The difference is that the cost of solving them manually grows with every property added.
See the System Behind the 20x
This platform was built inside a real, operating property tax protest firm, not designed in a vacuum. If your pipeline is growing faster than your team, it’s worth seeing how it works.
Take on more properties, without taking on more people. See It in Action →