Two cost segregation studies on the same building can produce very different numbers, and the gap almost always comes from how the study was built, not from the building. A study that estimates the component split from a few inputs tends to leave property on the 39-year schedule that belonged on a faster one. A fully engineered and accounted study works through the building component by component, documents each classification, and typically moves 20 to 35 percent of depreciable basis into five-year and fifteen-year property. With 100 percent bonus depreciation, that reclassified amount can be deducted in the first year, if you have income to deduct it against. A low number is not automatically safe, and a high number is not automatically right. The right number is the one your study can defend line by line.
Tom Alvarez owns a small commercial building in Greenville, South Carolina. Two storefronts and an office suite upstairs. He bought it in May 2025 for $2,000,000, and his closing statement and his CPA's allocation put $400,000 of that on the land. That left $1,600,000 of depreciable basis.
In October his CPA, Walt Presser, told him to get a cost segregation study before the return was filed. Tom did what most owners do. He asked two firms for proposals.
The first came back in a day. The fee was low. The projection was a reclassification of about 12 percent of basis, and the cover note said the study would be completed from the purchase documents and a questionnaire.
The second took a little longer. The fee was higher. The projection was about 28 percent, and the proposal described a component-by-component analysis of the building, with a direct review of the property in whatever way fits the building.
Tom called Walt. "They cannot both be right," he said. "One firm says I get a deduction of about $190,000. The other one says $450,000. It is the same building."
"Tom, they are not looking at the building the same way," Walt said. "One firm is estimating from a template. The other is working through your actual building. On a property this size, that difference can be worth close to a hundred thousand dollars in tax."
If you are comparing proposals, here is how to read them.
When you buy a commercial building, the tax code starts you at a default. The whole building, less land, depreciates straight-line over 39 years. A residential rental depreciates over 27.5 years. That default treats a roof, a load-bearing wall, a parking lot, and the cabling behind a data wall as one lump.
The code does not actually treat them as one lump. Some parts of a building are structural components of the building. Other parts are tangible personal property that happens to sit inside the building, and some are land improvements outside it. The tax code gives those categories shorter recovery periods, most commonly five years for certain interior property and fifteen years for site improvements such as paving, fencing, and certain landscaping.
A cost segregation study separates one into the other. It identifies the components, assigns each a cost, and classifies each under the tax rules. Your CPA then depreciates each class on its own schedule. If you want the full introduction, our plain guide on what cost segregation is covers the basics. This article is about the part most owners never see, which is how the study gets built.
Whiteco Industries v. Commissioner (65 T.C. 664, 1975) sets out six factors for deciding whether an asset is an inherently permanent structure or tangible personal property, and HCA v. Commissioner (109 T.C. 21, 1997) is another decision studies cite on how components inside a building get classified. A good study classifies each component with that case law in mind. We walk through one of them in our post on the Whiteco six-factor test.
The IRS Cost Segregation Audit Technique Guide, which is the manual IRS examiners use when they review a study, describes several ways a study can be prepared. The two most rigorous work from actual cost records, or from a detailed cost estimate built for the building. Cost Seg America performs studies using IRS Approach 1 and IRS Approach 2. Other methods lean more heavily on models, surveys, or rules of thumb.
In a detailed study, an analyst starts with what is physically in the building. Drawings, invoices, closing documents, photographs, video, and where the property warrants it, an in-person inspection. Every component gets identified and priced. The roof, the storefront finishes, and any dedicated electrical run to equipment each get their own line. The result is a ledger that shows what each component is, what it cost, and why it landed in its class.
In an estimated study, software or a template takes a few inputs, such as the property type, the square footage, the price, and sometimes a short questionnaire, and produces a split by applying benchmarks from other buildings. Benchmarks are fine as a sanity check. They are a poor substitute for the study. Your building is not the average building. If you put in a tenant improvement package, a specialty build-out, a long run of site paving, or a large amount of dedicated wiring, an estimate has no way to see it.
The question is whether a smaller, faster tool is the right one for a building where the deduction is worth six figures.
Here is Tom's building with both proposals run through the numbers. This is an illustration, and the figures are simplified so you can see the shape of it. Your result depends on your property and your return.
Depreciable basis is $1,600,000. That is the $2,000,000 purchase price less $400,000 of land.
Without any study, the building depreciates over 39 years. That is about $41,000 of depreciation in a full year, and a bit less in the first year because the building was placed in service partway through 2025.
With the estimated study at 12 percent, $192,000 of basis moves into shorter-lived property. Because the property was acquired and placed in service after January 19, 2025, the reclassified property qualifies for 100 percent bonus depreciation under the law signed on July 4, 2025, so that $192,000 can be deducted in the first year. At a 37 percent federal rate, that is about $71,040 of first-year federal tax reduction, assuming Tom can use the deduction.
With the fully engineered and accounted study at 28 percent, $448,000 of basis moves into shorter-lived property. At the same 37 percent rate, that is about $165,760 of first-year federal tax reduction.
The gap between the two studies is $256,000 of additional first-year deductions and roughly $94,720 of first-year federal tax, before the effect of recapture on a later sale. Even if the fully engineered and accounted study costs several thousand dollars more, the gap is large.
Why would the second study find more?
For a wider view of when the return on a study makes sense, see our breakdown of when a cost segregation study pays for itself.
The numbers separate in about five places.
A questionnaire asks whether the building has a tenant build-out. It does not ask what the build-out contains. A restaurant build-out might have a hood system, dedicated electrical, grease-handling plumbing, specialty finishes, and a walk-in cooler pad. Each has its own classification question. If the study does not look, those costs stay inside the 39-year bucket by default.
Tom's building has a paved lot, striping, parking-lot light poles and a landscaped frontage along the street. Many of those are land improvements that depreciate over 15 years, not 39, though each item has to be classified on its own facts. They are also easy to miss when a study starts from the building and never looks at the site. For more on this, see our post on how parking lots and landscaping become 15-year property.
A study is only as good as the numbers behind each component. If the building was constructed from scratch and you hold the contractor's invoices, the study can price components from actual cost. If you bought the building, there are usually no construction invoices to work from, so the analyst builds a detailed cost estimate from unit costs, quantities, and the construction drawings or the property review. A shortcut here is the most common reason numbers drift. A detailed estimate uses quantities and current unit costs for the specific scope of the building. A shortcut applies a percentage.
Land does not depreciate, and the amount allocated to it comes off the top before any study begins. A study that begins with a sloppy land allocation starts from the wrong basis. Our post on why the land number comes out first shows how much that one input can move the result.
Some components sit right on the line. Is a given wall structural or is it a movable partition tied to the tenant's business? Is a certain piece of wiring part of the building's general electrical system, or is it dedicated to specific equipment? A careful analyst makes those calls one at a time and writes down the reason. A fast study makes them in bulk, and usually rounds toward the safe, low answer. That is how a 12 percent result and a 28 percent result come out of the same building.
A study that comes in too low leaves deductions on the table. Those are real dollars you could have used in the first year. And a missed opportunity is not always easy to fix, though the tax rules do allow a catch-up for missed depreciation, either through a change in accounting method on Form 3115 with a Section 481(a) adjustment or, in some cases, an amended return. We covered that process in our post on catching up on missed depreciation without amending returns.
A study that comes in too high, with no documentation behind it, creates a different problem. If the classification of a component cannot be tied to a cost source and a reason, an examiner can challenge it, and the reclassified amount can be pushed back to the 39-year schedule along with interest and possible penalties. A big projection from a firm that cannot show its work is a risk, not a benefit.
The goal is the number that is both complete and supportable. That is what a fully engineered and accounted study is built to produce, and it is why we back every study with lifetime audit support at no additional cost.
The Audit Technique Guide lays out the features an examiner expects to see in a well-prepared study. In general terms, those include a preparer with real engineering and construction knowledge, a clear description of the methodology, documentation of where each cost came from, a consistent and explained classification of each component, and a report that someone else could follow and check.
You can read our deeper breakdown in the 13 elements of a quality study and our overview of the Audit Technique Guide. Our post on audit triggers and how to avoid them covers the other side, which is what tends to draw attention.
Here is the practical test. Picture an examiner opening your report. Can they see what each component is, what it cost, where the cost came from, and why it was classified the way it was? If the answer is yes, the report works. If the answer is a table of percentages, it does not.
Tom went back to both firms with the same questions. You can use them too. Our post on twelve questions to ask a cost segregation provider goes further.
Which IRS approach will you use? A good answer names the approach and explains it. A weak answer is that the firm uses proprietary software.
How will you review my property? There are three honest answers. An in-house analyst does an in-person inspection, which suits larger properties. The firm runs a tele-engineering review over live video. Or you get precise photo and video instructions and the team reviews what you send. Whichever one it is, the firm should be able to say why it fits your building.
Who does the engineering work, and where are they located? Many owners never think to ask this, and not every firm volunteers it. Ask the question, because the answer is not always obvious from a proposal. Do you want your engineering done overseas? At Cost Seg America, all of the engineering is done in the United States.
What will the report show? Ask to see a sample report, and look for a component-level ledger with cost sources, not a summary page of percentages.
What happens if the IRS asks questions? Look for audit support that lasts for the life of the study, with no cap and no fee. Ours does.
How long will it take? Our studies are delivered in 30 days, and rush studies are available in under 30 days.
The first firm's answers thinned out by the third question. The second firm's stayed specific. Tom signed with the second.
A week after the signing, Walt called Tom again with the part owners tend to forget.
"A deduction only helps when it meets income," Walt said. "Tell me what you earn."
Tom's building throws off modest rent from unrelated tenants. Most of his income is from his plumbing supply business, which he runs through an S corporation. The rental activity is passive under the tax rules, which means a rental loss does not automatically offset his plumbing income. Passive losses generally offset passive income, with exceptions for taxpayers who qualify as real estate professionals, for qualifying short-term rental operators who materially participate, and for a $25,000 allowance for owners who actively participate, which phases out between $100,000 and $150,000 of modified adjusted gross income.
That did not mean the study was a mistake. Unused passive losses carry forward, and they are released against passive income in later years or when the property is eventually disposed of in a taxable sale. It did mean Walt needed to model when Tom could use the deduction before anyone promised him a first-year check.
This is the single most common gap between what a proposal projects and what an owner experiences. A cost segregation study creates the deduction. Your income and your tax status decide when you can use it. Three of our other posts go into the detail: real estate professional status, material participation rules, and what happens to suspended passive losses when you sell.
One more thing Tom got wrong, and many owners do. He assumed that because he bought the building in May, he had to complete the study before December 31. He did not. The study supports a depreciation claim on a return. It is the filing of that return, and the placed-in-service date of the property, that drive the tax year. Our post on why the deadline is the filing date, not December 31 explains it. The related question of what counts as placed in service is covered in this post.
For the background on the bonus rate, see our post on 100 percent bonus depreciation under the 2025 law.
Choosing on fee alone. On Tom's building, the cheaper study would have cost him up to $94,720 in first-year federal tax, once he can use the deduction. Compare the deduction, not just the fee.
Choosing on the biggest projection alone. A high number from a firm that will not show its method is a reason to ask questions. Ask how the number was built.
Assuming every study is the same product. A software-generated split and a fully engineered and accounted study share a name, but they differ in method, in documentation, and in how they hold up under examination.
Forgetting the land. If the land allocation is wrong, every percentage downstream is wrong too.
Ignoring the property's minimum size. A study works best when there is enough depreciable basis to reclassify. Our minimum is $180,000 of depreciable basis after land is removed. Below that, the math rarely favors a study.
Ignoring what happens at sale. When you sell, gain tied to the depreciation you took is generally taxed as recapture. That does not make a study a bad idea, but it belongs in your plan. See depreciation recapture and cost segregation.
Because they are usually produced by different methods. One may estimate the component split from a few inputs. The other analyzes the property component by component and prices each one. The second method sees tenant improvements, site work, and dedicated systems that the first cannot see, and it documents each classification.
No. A higher projection only helps if the study can support it. Each reclassified dollar should tie to an identified component, a cost source, and a reason for its classification. A number that cannot be defended can be pushed back on examination.
Cost Seg America typically finds a reclassification rate of 20 to 35 percent of depreciable basis, and we typically identify $200,000 to $450,000 in additional first-year federal deductions per $1 million of property value. Some property types, such as certain manufacturing and cold storage facilities, can exceed that range. Every property is different, which is why the study is built from the property and not from a formula.
It applies to qualifying property acquired after January 19, 2025, under the law signed July 4, 2025. Property acquired earlier may be subject to a different bonus percentage. Your CPA confirms which rate applies to your facts.
Often, yes. A study on an earlier purchase can support a catch-up for depreciation you did not take. Your CPA decides whether that is done with a Form 3115 change in accounting method, which carries a Section 481(a) adjustment, or in some cases an amended return. Form 3115 support is available from Cost Seg America for an additional fee.
A study by itself is not a reason for the IRS to examine your return. A well-documented study, prepared using a recognized approach and reviewed against the Audit Technique Guide, is built to stand up if a question comes. We provide lifetime audit support at no additional cost.
It depends on the building. On larger properties, an in-house analyst does an in-person inspection. For others, we run a tele-engineering study over live video, or we give you precise photo and video instructions and our team reviews what you send. Whichever method fits, the work is a direct engineering analysis of your property, not a template.
Thirty days from the signed proposal. Rush studies are available in under 30 days.
We look for at least $180,000 of depreciable basis after the land value is removed.
Tom's building is an example, and your property will have its own numbers. If you have a proposal in hand and want to understand what is behind it, or if you want to see what a fully engineered and accounted study would show on your building, send Jim Dougherty and his team the property address, the purchase price, and the date it was placed in service. You will receive a free proposal in 24 hours showing the projected first-year deduction, the method, and the fee, with no obligation. Call 1-888-365-5023 or email info@costsegamerica.com.
This article is for general educational purposes and is not tax, legal or accounting advice. The examples are illustrative, and tax outcomes depend on your facts. Confirm current thresholds and your own situation with your CPA before acting.
Use the calculator, see your number, and request your free, no-cost proposal - delivered in 24 hours, with your flat fee quoted upfront and no obligation.