How I Did a Property Tax Protest and Saved 36%

How I Did a Property Tax Protest and saved 36%.
Disclaimer – this is how I got these savings; do not take this as a guarantee that you can get similar results, I am describing my experience and how I did things, your mileage may vary!

Cutting expenses is the way to minimize inflation at the household level. Property assessments are one of the key drivers at the household level, but the time and cost of fighting is also high. So here is the walkthrough of how I fought the county and won. It wasn’t fun, but I learned a lot. I would estimate that on average it might take 20 hours to do well. I easily spent upwards of 40 hours, perhaps 80 – but my case was very complicated due to multiple lots being involved in a unique way, so I was able to cut 36% off of the tax assessor’s price. Most people won’t see savings this extreme I think.

Step 1: The county is supposed to show evidence proving their number. I got their “proof,” asking for it online through a portal. Procedures may vary by county.

My edge is that the county has a lot of parcels to justify and I did not. Counties often rely on “mass appraisal” techniques – they just vacuum up key house statistics for houses in the region and average it all together, then adjust a few key figures for my house. Beating the comparable sales evidence generally is not too difficult I thought, because these were based on houses sold. Houses for sale have likely been renovated within the past year or two, and because my house was sold “as-is”, my house was visibly worse than the houses for sale.

See the county’s comparable sales analysis below.


Step 2: I identified each county “proof” property on the online real estate websites. I saved every picture of every property possible (I was able to find ~2/3 of the properties that the county presented, with detailed pictures) into Word, then PDF. I learned to check every website for a property, because some of them maintain pictures for a property for longer times than others, meaning I was able to get older pictures of some houses that were not available on other sites at all. This helps because if the county claimed that the pictures were too old, I could truthfully claim that the pictures I found were the most up-to-date information available, and that it was better information than what the county had.

Here’s an example of what I did, along with notes.


I did this for my house and for each of the comparable houses. I added the houses in the county provided equity analysis into a combined valuation here. I did as many as possible to bulletproof my argument. I did not find every one online, but making a good faith effort showed. In doing this, I turned the county’s evidence against them – they couldn’t say “I cherry-picked these properties” – because THEY picked these houses. What I suspect but can’t prove is that THEY cherry-pick properties to a degree, and this analysis helped put perspective into the picture.

Step 3: I then estimated how much each “feature” increases the value of each of the “proof” houses. I searched online for “how much does it cost to add <new windows in Dallas >”, then used a median price, for every feature I saw. I then subtracted the value from the “proof” house. My goal is to TRUTHFULLY knock down the county’s comparables by pointing out just how much better their comparable houses were than mine, to the point where the assessor had to agree with me.

Step 4: Alternately, I could have gotten quotes from contractors for work on my house to bring my house up to the level of the other houses. The advantage of this is that contractors are now unrealistic and that I did not need to actually pay the amount quoted. For example, I received a quote for $8K to paint the interior of my ~1200 square foot house. In terms of materials, I could get this done for between $400-600; it just takes time! I chose not to use inflated quotes, but I could have, and this would have been acceptable, as long as I had the written quote.

Step 5: I created an Excel spreadsheet where each comparable house is its own column, and each row is a line item for the house (kitchen island, outside deck, new wood flooring, etc). I reviewed each photo of each comparable house, looking for differences in value, and marked down “Y” or “N” for each house feature visible – maybe it’s a chimney, new lighting, hardwood flooring, new windows, etc. Finally, I used Excel formulas to calculate the value difference for each house.

Adjustment to house

Special Adjustment



Step 6: By doing all of the above work, then subtracting the value of the improvements for each house, I got the “true” average value of the comparable houses. During the protest, the assessor claimed “a lot of those items are maintenance type / not suitable for adjustment”, I wish I had reminded the board that market value is dictated by buyers, and buyers care about the house condition and cost to bring a house up to standard,  not which particular bucket of spending.

Step 7: Now, I did the same thing for land – I looked at each property’s pictures, and found out unique features of each piece of land – for example, was the county’s comparable piece of land closer to a high-value location, like waterfront property, parks, shopping areas, amenities, etc. I factored in distance and assigned some values that make sense, based again on online estimates. I realized my house had a higher percentage of slope compared to the county property’s base lot value. Had my house been in a flood zone, or if I had less flat ground, or an irregular shaped property, etc, I would have noted it.

Step 8: I got ready for the protest. I signed an unsworn declaration and shipped the data online, then did a notarized one in case I had connection problems (online notaries are available for $25 when I did this). In the protest itself I only had around 5-7 minutes to make my case. The pictures sealed the deal for me – I was able to show that my property was overvalued against the comparables. I did not argue market conditions this year, but next year with a down market I likely would. It all worked out in the end.