How to Read a Cost Segregation Proposal Before You Sign It

The short version. A cost segregation proposal is a scope document with a price on the bottom, and the price is the least useful thing on the page. Two firms quoting the same building can be selling two different work products. What separates them is method, and method is visible in the proposal if you know where to look: whether anyone walks the building, which of the six IRS-recognized approaches is being used, what will actually be in the report, and whether the fee moves with your deduction. Publication 5653 gives you the standard to hold each one against.

Most people compare cost segregation proposals the way they compare flights. Two numbers, take the lower one.

That works when the product is identical. It is not identical here.

The IRS has never set requirements or standards for preparing a cost segregation study. Nobody is licensed to do it. Nothing is mandatory. Which means the word study covers a physical inspection with component-level documentation, and it also covers a spreadsheet built from six answers on an intake form, and both proposals can land in your inbox the same afternoon looking like siblings.

So read the proposal instead of the price. Five places to look.

Does the proposal include an in-person site visit?

This is the first thing I check, and it disqualifies more proposals than anything else on this page.

The IRS Cost Segregation Audit Technique Guide, Publication 5653, revised February 2025, is direct about it: "A field inspection is recommended for all quality studies, whether the studies are for new or used properties."

Not recommended for new construction. Not recommended for large properties. All quality studies.

A site visit means a trained professional standing inside your building. A questionnaire and a photo request is not a field inspection, and I will not treat the two as interchangeable no matter what the scope section calls it. If the proposal is vague about who shows up, or whether anyone shows up at all, the vagueness is the answer.

Which of the six approaches is the proposal actually using?

Publication 5653 recognizes six methodologies, and it ranks them. A proposal that never names its method is asking you to assume it uses the good one.

On the detailed engineering approach built from actual cost records: "In general, it is the most methodical and accurate approach, relying on solid documentation of the construction costs and minimal cost estimating."

On the rule-of-thumb approach: "An examiner should view this approach with caution since it lacks sufficient documentation to support its allocation of project costs."

The residual estimation approach is the one most often sold at the low end, and it deserves a closer look. It prices the short-lived assets, adds them up, subtracts that total from project cost, and calls the remainder the building. The guide's assessment: "It should be recognized that this method generally does not reconcile project costs." And on which direction the error runs: "Different estimation techniques for short-lived assets can produce a skewed result in favor of § 1245 property."

Read that second quote again. The IRS is saying the cheap method tilts toward the answer the buyer wants, which is the answer an examiner is trained to test. Cheap and aggressive show up together.

Engineering-based is the only kind I will put my name on. Desktop studies, database allocations, rule-of-thumb percentages, and AI-generated outputs are cheaper, and they are worse, and I am not going to write about them as if they were one option among several. What an engineering-based study actually looks like walks the process end to end.

Can the firm show you where the thirteen principal elements live in your report?

Publication 5653 lists thirteen principal elements of a quality study. That list is the closest thing to a standard that exists, it is public, and you can hold it against a report that has not been written yet.

Ask the firm to point at where each element will appear in your deliverable. A firm doing engineering work walks you through it without breaking stride. A firm selling an output goes quiet somewhere around element nine, usually on unit cost determination and on reconciliation to actual costs, because those two require the underlying records rather than a model.

Preparer qualifications sit in that same guide, and the language is worth knowing before you read anyone's bio page. Publication 5653 states that "there are no prescribed qualifications for cost segregation preparers," and, in the same document, that "in general, a study by a construction engineer is more reliable than one conducted by someone with no engineering or construction background." Nothing is required. Everything is evaluated.

How is the fee calculated, and does it move with your deduction?

I push back on any proposal priced as a percentage of your tax savings, and my objection is not about fairness.

The engineering work on your building does not change based on the size of your deduction. The hours are the hours. When the fee floats with the reclassification, the person deciding which assets qualify as § 1245 property earns more by deciding aggressively, on precisely the allocations an examiner tests hardest. You have built a financial incentive into the one judgment call that needed to stay conservative.

Plenty of pages will tell you contingency pricing is a red flag. Fewer explain why. That is why.

A flat fee quoted against a defined scope has the opposite property. The firm gets paid for the work no matter where the components land, so the classification is free to be correct. For what those flat fees run and how to tell whether a study clears its own cost, see what a study costs and whether it pays for itself.

What five questions should you ask before you sign?

The answers will tell you more than any fee comparison.

1. Does this include an in-person site visit by someone trained to perform one, or does it rest on a questionnaire and a photo request? Publication 5653 recommends field inspection for all quality studies. A remote walkthrough is not a field inspection, and I will not treat the two as interchangeable.
2. Who performs the analysis, who reviews it, and what is their construction or engineering background?
3. Will the final report include photographs, cost allocations tied to actual records, and written reasoning for each reclassification?
4. How many studies has the firm completed in this specific asset type? A firm that has done four hundred self-storage facilities knows where the components live in a self-storage facility.
5. What happens if the IRS asks about this study four years from now, and is that support included or billed separately?

A provider who answers all five plainly is selling a different product than one who redirects you to the savings estimate. Keep the answers. The compliance folder post covers what to hold onto and for how long.

When should you walk away from a proposal?

When the savings number arrives before anyone has seen the building.

A firm that leads with your projected deduction is selling the number. A firm that leads with scope is selling the work. The second one is the only one that has anything to hand an examiner, and here is the part that gets skipped in every sales conversation I have watched: you carry that risk, not the firm that prepared the study. You sign the return.

Walk, too, when the proposal has no answer on what happens later. Audit support that is neither included nor priced is audit support that does not exist.

And walk when the property is wrong for a study at all, which no proposal will ever tell you. A short holding period is the big one, because accelerated depreciation on § 1245 property comes back as ordinary income recapture at sale, so three years and a sale is a different calculation than thirty years and an heir. If you plan to hold until death, a stepped-up basis changes the recapture question entirely. A shell building with almost nothing in it will not reclassify enough to pay for the work. A basis too small will not justify the engineering hours. And with no income to shelter and no path to using the deduction, the study buys you a report.

What should a proposal on an older building say?

It should say look-back, and it should say Form 3115.

A look-back study captures every year of missed depreciation in the current tax year through a Section 481(a) adjustment, without amending a single prior return. There is a procedural step people miss, and Publication 5653 states the Service's position directly: "a change in depreciation method, recovery period or convention for depreciable property constitutes a change in accounting method. Therefore, the use of a cost segregation study to reclassify property and/or reallocate costs requires the consent of the Commissioner." That consent comes through Form 3115.

A proposal on a building you have owned for years that never mentions the accounting method change was written by someone who has not done many of these. The look-back guide covers the mechanics.

What you are actually buying

The number on the proposal is not the thing to compare. What you are buying is the documentation behind the number, and the documentation is the only part the IRS ever looks at.

My father, a CPA, told me that tax evasion is a crime but tax avoidance is mandatory. That line has always meant taking every deduction the code allows and being able to prove you were entitled to it. A thin study can cost you both halves of that at once.

If you have a proposal in hand and want a second read on it, or you have never had a building-specific analysis done at all, I offer a no-cost analysis delivered through my relationship with CSSI. No obligation and no pitch. Do not go pull documents together first; that is a barrier, not a service. We will have a conversation, and if the numbers say a study will not pay for itself, I will tell you that.

Two ways to start:

• Book a time: calendly.com/david-wiener/cs
• Call 770-224-8504, option two

Frequently Asked Questions

What should a cost segregation proposal include?

At minimum it should name the methodology it uses, state whether an in-person site visit is part of the scope, identify who performs and who reviews the analysis, describe what the final report will contain, and quote a fee against that defined scope. A proposal that leads with a projected savings number and stays vague on all five is selling the number rather than the work.

Does a cost segregation study require a site visit?

It is not required, because the IRS has not established requirements or standards for preparing cost segregation studies at all. Publication 5653 does say field inspection is recommended for all quality studies, whether the studies are for new or used properties. A proposal that skips the building departs from the guide your examiner is reading.

How can I tell whether a proposal is engineering-based or a desktop estimate?

Look for the method and the inputs. An engineering-based study assigns costs to specific components using construction documents, invoices, and a physical inspection. A desktop or database study estimates from tables and industry percentages and needs very little from you. If the firm can complete the work from an intake form alone, it is not doing the first kind.

Should I pay a percentage of my tax savings for a cost segregation study?

I push back on it. The engineering hours do not change with the size of the deduction, so a fee tied to the reclassification pays the preparer to be aggressive on exactly the allocations the IRS examines hardest.

What are the thirteen principal elements of a quality cost segregation study?

They are a list published in Publication 5653 describing what a quality study contains. You do not need to memorize them to use them. Ask a firm to show you where each element will appear in your report, and watch what happens at unit cost determination and reconciliation to actual costs, since those two require the underlying records rather than a model.

Should a proposal on an older building mention Form 3115?

Yes. Publication 5653 states that using a cost segregation study to reclassify property or reallocate costs requires the consent of the Commissioner, and that consent comes through Form 3115. A proposal on a building you have owned for years that never raises the accounting method change was written by someone who has not done many of these.

Can I perform a cost segregation study myself?

Nothing prohibits it, since Publication 5653 states there are no prescribed qualifications for preparers. The same guide states that a study by a construction engineer is generally more reliable than one by someone without that background, and a self-prepared study will struggle on several of the thirteen principal elements, particularly unit cost determination and reconciliation to actual costs.