How Does a Cost Segregation Study Cut My Buildout Taxes in 2026?

The outcome you should expect

The honest framing is this: a cost segregation study does not reduce the total amount of depreciation you will eventually claim. It reduces *when* you claim it. Every dollar of basis in a building still gets deducted exactly once. What the study buys you is sequencing — pulling deductions forward from years 20 through 39 into years one through five. That sequencing is worth real money because a dollar deducted today is worth more than a dollar deducted in 2059, and because most owners who just finished a buildout are cash-poor at precisely the moment the study delivers.

Run the arithmetic on a concrete case. Say you own the building and put $1,000,000 of depreciable improvements in service. Without a study, the entire amount sits in the 39-year straight-line bucket and generates roughly $25,600 of depreciation per year — about $2,100 a month of shelter against income. With a study that reclassifies $300,000 into 5- and 15-year property, that $300,000 becomes eligible for bonus depreciation, which applies to any asset with a recovery period of 20 years or less. Depending on the bonus rate in effect for your placed-in-service year, somewhere between a large fraction and the entire $300,000 lands as a deduction in year one, with the remainder depreciating on its accelerated MACRS schedule rather than crawling across four decades.

Convert deduction into cash. At a 37% federal marginal rate plus a state rate in the mid single digits, a $300,000 first-year deduction is worth somewhere in the neighborhood of $120,000 to $130,000 of tax you do not write a check for this April. Even at a 24% bracket the number is roughly $80,000 including state. That is not a permanent tax cut — you will give some of it back through smaller deductions in later years and through recapture at sale — but it is an interest-free loan from the Treasury sized in six figures, drawn at the exact moment your bank account is recovering from a construction project.

The right metric for judging the study is net present value, not gross deduction. Discount the accelerated deduction stream against the 39-year baseline stream at whatever rate reflects your real cost of capital — 8% is a common planning assumption, and if you carry construction debt at 8% or 9%, that is not theoretical. Under those assumptions, accelerating $300,000 of depreciation out of a 39-year drip and into the first year or two generates an NPV benefit in the range of $40,000 to $70,000 per million dollars of basis. Compare that to a study fee of $5,000 to $15,000 on a project under $2 million and the return ratio is usually somewhere between four and eight to one.

There is a secondary outcome worth naming: cash flow smoothing at the worst point in the ownership cycle. A buildout is a capital event. You paid the general contractor, you funded the FF&E, you covered the two months of double rent while the space was dark, and you are now ramping revenue in a new location. The year-one deduction lands in that trough. Owners who model the study only as a tax number tend to undervalue it; owners who model it as working capital in the first twelve months of occupancy tend to run it every time.

One more expectation to set correctly. A study is not a self-serve spreadsheet exercise. The IRS is explicit in its own audit guidance that the credible approach is an engineering-based cost allocation performed by someone qualified to identify and cost the components. A rule-of-thumb allocation scribbled by a bookkeeper is the kind of position that unravels under examination. Budget for the real thing, or do not do it at all.

What drives that outcome

The engine is component classification. Your closing statement or contractor's final application for payment shows one number. The tax code sees dozens of assets with different useful lives, and the study's whole job is to walk the property, read the drawings and pay applications, and put each component in its correct class.

Five-year property under MACRS is where the leverage concentrates. This bucket generally captures carpet and other non-permanent floor coverings, decorative lighting that serves a decorative rather than a general illumination function, dedicated electrical circuits and wiring that serve specific equipment rather than the building, data and communications cabling, removable partitions, accent millwork and casework tied to the operation, breakroom appliances, and signage. Seven-year property picks up certain furnishings and fixtures. Fifteen-year land improvements cover the site work most owners forget entirely: parking lots, curbing, sidewalks, landscaping, exterior lighting, fencing, and site drainage. Everything left — the shell, structural walls, roof, building-wide HVAC, core plumbing — stays 39-year.

The percentage that moves is driven almost entirely by use type, and the spread is wide enough that it should shape whether you commission a study at all. A generic open-plan office buildout, where most of the money went into drywall, ceiling grid, and building-standard mechanical work, typically reclassifies 15% to 25%. A medical or dental fit-out, loaded with dedicated power runs, specialty plumbing for operatories or sterilization, lead-lined imaging rooms, and casework, usually lands in the 25% to 35% band. A restaurant is the outlier on the high end — heavy electrical service to the kitchen line, hood and exhaust interfaces, decorative front-of-house lighting and finishes, dedicated gas and plumbing runs — and commonly reclassifies 30% to 45%. Retail with a heavy fixture package behaves similarly. A warehouse with a big paved yard often gets more of its benefit from the 15-year land improvement bucket than from the interior at all.

Bonus depreciation is the multiplier that turns classification into a year-one number. Cost segregation alone just changes the schedule — a reclassified item now depreciates over five years instead of thirty-nine, which is meaningfully better but not dramatic. Bonus depreciation under Section 168(k) permits an immediate deduction of a percentage of the cost of qualifying property with a recovery period of 20 years or less. Every bucket the study creates qualifies; the 39-year shell never does. So the causal chain is: the study creates eligibility, and bonus converts eligibility into immediate cash. Without the study, those assets are invisible inside the shell and no amount of bonus depreciation reaches them.

Section 179 expensing is the second lever, and it reaches assets that bonus historically could not. Section 179 permits direct expensing of qualifying property up to an annual dollar cap, with a phase-out that begins once total qualifying purchases for the year exceed a threshold. Critically, the provision was expanded to cover roofs, HVAC, fire protection and alarm systems, and security systems on nonresidential real property — all of which are 39-year items that a cost segregation study cannot reclassify. The sophisticated play is to stack: Section 179 against the long-life building systems, bonus depreciation against the short-life assets the study surfaced. Confirm the current-year cap and phase-out threshold with your CPA, because both are indexed annually and both have moved.

Bonus depreciation rates have been genuinely volatile. The statutory schedule stepped down from 100% for property placed in service through 2022, to 80%, then 60%, then 40%, then 20%, with full phase-out thereafter — and Congress has repeatedly intervened to restore or extend higher rates. This is not a detail you can carry forward from an article you read two years ago. Ask your tax advisor for the rate that applies to your specific placed-in-service date before you model anything, because the difference between 40% and 100% bonus on a $300,000 reclassification is roughly $180,000 of first-year deduction.

There is one more driver that rarely gets discussed: documentation quality determines how much of the benefit survives contact with scrutiny. A study built from a site walk, the architect's drawings, the contractor's schedule of values, and the actual invoices produces a defensible allocation. A study built from a purchase price and a percentage table produces a number. Both give you a deduction on the return; only one gives you a position you can hold. Ask any firm you interview whether their report ties each reclassified line to source cost documentation, and ask to see a redacted sample report before you sign.

Benchmarks and realistic ranges

Start with the threshold question, because most of the bad decisions here are decisions to study a project that was too small. Below roughly $500,000 of depreciable basis, the economics get thin: a $5,000 study against a $500,000 basis reclassifying 20% is moving $100,000, which at an 8% discount rate produces maybe $15,000 to $25,000 of NPV benefit — still positive, but not obviously worth the coordination cost and the recapture bookkeeping you inherit. Between $500,000 and $750,000 it becomes a judgment call that turns on your bracket, your use type, and whether you have income to absorb the deduction. Above $750,000 with a use type in the 25%-plus reclassification band, it is usually a clear yes. Above $2 million it is close to malpractice not to at least run a free feasibility estimate.

Study pricing tracks complexity and square footage, not just basis. For a straightforward buildout under $2 million, expect $5,000 to $15,000. For larger or multi-property engagements, $15,000 to $50,000 is a normal range, and firms will often quote per-property discounts across a portfolio. Most reputable providers will give you a no-cost feasibility analysis first — a one- or two-page estimate of the likely reclassification percentage and dollar benefit against the fee. Take it. If a firm will not model the benefit before you pay, that is information.

Timeline benchmarks: two to six weeks from engagement to final report is typical, with the site visit usually happening in week one or two. Complex properties, multi-building sites, or engagements where the contractor's cost records are disorganized run longer. The binding constraint is almost never the engineering — it is how fast you can produce the schedule of values, the change order log, the architect's drawings, and the final invoices. Owners who have that package assembled before the kickoff call routinely cut a week or two off the timeline.

On the benefit side, the durable rule of thumb is that every $100,000 you move from the 39-year schedule into a year-one deduction is worth roughly $13,000 to $25,000 in net present value at an 8% discount rate, with the spread driven by your marginal rate and the bonus percentage available. Scale that: a $2 million buildout reclassifying 30% moves $600,000, which lands somewhere between $78,000 and $150,000 of NPV against a study fee under $20,000.

Now widen the lens, because the buildout study is one node in a larger fixed-asset strategy and the adjacent moves often carry comparable value. Qualified Improvement Property — interior improvements to nonresidential real property made after the building was first placed in service, excluding enlargements, elevators and escalators, and internal structural framework — carries a 15-year recovery period and is bonus-eligible in its own right. If you are a tenant, or an owner making improvements to an existing building rather than constructing new, QIP is frequently the more relevant path, and a good advisor will evaluate both rather than defaulting to a full study.

The partial asset disposition election is the most commonly missed adjacent move. When you gut a space, you are throwing away components that are still on the depreciation schedule from a prior improvement — the old ceiling, the old lighting, the old HVAC distribution. The regulations let you elect to recognize the remaining undepreciated basis of those retired components as a loss in the year of disposition, and to deduct the demolition costs rather than capitalizing them. The election is generally made on a timely filed return for the year of disposition, so it is easy to lose permanently by simply not thinking about it in time. A cost segregation study on the *prior* improvement is what makes this election quantifiable, because you need a component-level basis to write off a component.

The tangible property regulations create a third adjacent lane. The repair-versus-capitalization analysis, the routine maintenance safe harbor, the de minimis safe harbor for low-cost items (with a per-item threshold that differs depending on whether you have an applicable financial statement, and which requires a written capitalization policy in place at the start of the year), and the small taxpayer safe harbor for buildings all let you expense items outright rather than capitalizing and then accelerating them. Expensing beats accelerating. A well-run buildout tax strategy runs the repair analysis first, applies the safe harbors, then studies what is left.

Then there is the energy overlay. Section 179D provides a deduction for energy-efficient commercial building property — lighting, HVAC, and envelope improvements that meet specified efficiency standards — and Section 45L provides a credit in the residential context. These interact with cost segregation, because the same lighting retrofit can appear in the study's 5-year bucket and in the 179D analysis, and the sequencing matters. Coordinate them; do not run them in separate silos with separate advisors who never speak.

Finally, a note on where this connects to operational finance and RevOps practice. Teams that run disciplined revenue operations already model deferred revenue, cash conversion, and unit economics on a rolling basis. Depreciation timing belongs in that same model. When you evaluate whether a new location clears its hurdle rate, a first-year deduction worth $100,000 of retained cash changes the payback period materially — often by a full quarter or more on a location that was already marginal. Sites get approved or killed on that difference, so the tax timing assumption deserves to be an explicit line in the pro forma rather than an afterthought handed to the accountant in March.

Risks, edge cases, and failure modes

Recapture is the headline risk and the one owners most often discover too late. When you sell, the depreciation you accelerated comes back. Personal property in the 5- and 7-year buckets is subject to Section 1245 recapture, taxed as ordinary income to the extent of depreciation taken — at your full marginal rate, not the capital gains rate. Real property depreciation falls under Section 1250, with unrecaptured gain generally taxed at a higher rate than long-term capital gains. The net effect: you converted future capital-gain-rate dollars into present ordinary-income-rate dollars. If your ordinary rate is 37% and your capital gains treatment would have been meaningfully lower, part of your NPV gain is eaten by that rate arbitrage running the wrong direction.

This is why hold period matters so much. The time-value benefit compounds the longer you hold; the recapture cost lands all at once at sale. If you plan to sell in two or three years, run the analysis with the sale in the model, not just the year-one deduction. Owners who intend to hold indefinitely, or who plan to roll into a 1031 like-kind exchange and defer recapture, capture nearly the full benefit. Owners who flip inside a short window sometimes find the study was a wash after fees.

The passive activity loss rules are the second trap, and they catch a lot of people who own the building through a separate entity. If your rental real estate activity is passive to you, the accelerated deductions may be suspended and carried forward rather than offsetting your active business income this year. That does not destroy the benefit, but it can delay it by years, which is precisely the thing you paid to avoid. Material participation, real estate professional status, and grouping elections all bear on this, and they are fact-specific enough that the answer has to come from your tax advisor before you commission the study — not after.

A related failure mode: no income to absorb the deduction. If your entity is generating losses, a large year-one deduction just enlarges a net operating loss carryforward. NOL utilization is subject to limitation against a percentage of taxable income in later years, so the deduction becomes a deferred asset of uncertain timing rather than cash in hand. In that situation the correct move is often to wait — commission the study in a year when you have income, using the look-back mechanism described below, rather than burning it against nothing.

Timing discipline is where good plans die quietly. To claim bonus depreciation on reclassified assets for the placed-in-service year, the study needs to be complete and the position taken on a timely filed return for that year, including extensions. Miss it, and you are into Form 3115 territory — recoverable, but with added complexity and cost. The better practice is to engage the firm during construction rather than after. An engineer walking an active site can photograph and document components before drywall closes them in, which produces both a cheaper study and a stronger record.

Quality risk is real and asymmetric. The IRS publishes an audit techniques guide for cost segregation that describes what a credible study looks like and flags the methodologies it considers weak. Rule-of-thumb allocations and studies performed by people with no engineering basis are specifically identified as problematic. The asymmetry: a good study costs a few thousand dollars more than a bad one, and a bad one can cost you the entire deduction plus interest and penalties. Vet the firm on engineering credentials, on whether the report ties to source cost documentation, and on whether they provide audit defense as part of the engagement fee.

Two more edge cases worth flagging. First, if the building was acquired rather than constructed, the study allocates purchase price across land, building, and components — and land is not depreciable at all. A study that produces an aggressive land allocation downward to inflate the depreciable base is a different kind of risk than one that simply classifies components correctly. Second, state conformity varies. A number of states decouple from federal bonus depreciation and require an addback, which means your state benefit may be materially smaller than your federal benefit or may follow a different schedule entirely. Model federal and state separately; do not blend a single effective rate and call it done.

A practical rollout plan

Sequence the work so the decision points come before the money does. The first gate is feasibility, and it should cost you nothing. Pull the total depreciable basis, identify the use type, and get a no-cost feasibility estimate from two or three engineering-based firms. What you want out of that step is an estimated reclassification percentage, an estimated first-year deduction at the current bonus rate, and a fee quote. If estimated NPV benefit does not clear the fee by at least three to one, stop here and put the effort into the repair-versus-capitalization analysis instead.

The second gate is absorption, and it is the one most commonly skipped. Before you sign an engagement letter, confirm with your tax advisor that you can actually use the deduction: that you have taxable income in the placed-in-service year, that passive activity rules will not suspend it, and that your entity structure passes the benefit to whoever needs it. This is a thirty-minute conversation that prevents a five-figure mistake.

Third, assemble the document package before the kickoff call. The engineering team will want the architect's drawings and specifications, the contractor's schedule of values, the change order log, final applications for payment and invoices, the certificate of occupancy or placed-in-service documentation, and any FF&E purchase records. Owners who hand over a complete package on day one routinely finish faster and cheaper than owners who dribble documents out over three weeks.

Fourth, schedule the site visit while the space is still accessible and, ideally, while construction is still open. Walk it with the engineer if you can. You know which circuits serve which equipment and which millwork is decorative versus structural, and that knowledge tends to surface reclassification opportunities the drawings alone would miss.

Fifth, review the draft report against your own understanding before it is finalized. Check that the site work is captured in the 15-year bucket — paving and landscaping are the most commonly under-claimed items. Check that dedicated equipment power is separated from general building electrical. Check that the report cites source documents rather than percentages.

Sixth, coordinate the filing. The study output has to flow into the depreciation schedules, any Section 179 election, the bonus election or opt-out by class, any partial asset disposition election on retired components, and the state adjustment if your state decouples. Get the study firm and the CPA on one call rather than emailing a PDF and hoping.

That covers the forward-looking case. The backward-looking case is the look-back study, and it is the most underused tool in this whole area. If you completed a buildout one, three, or even ten years ago and never segregated it, you are not out of luck and you do not need to amend anything. You file Form 3115, Application for Change in Accounting Method, and take a Section 481(a) adjustment in the current year. That adjustment is a single catch-up deduction equal to the cumulative difference between the depreciation you actually claimed and what you should have claimed under the corrected classification. On a multi-million-dollar improvement placed in service several years back, that catch-up can be a six-figure deduction landing entirely in the current tax year, with no amended returns and no penalty for the earlier treatment.

The look-back mechanism also solves the absorption problem elegantly. If you had no income in the placed-in-service year, you can simply wait and run the study in the first profitable year, capturing the full catch-up then. That is a legitimate planning choice, not a workaround — the accounting method change procedure exists precisely for this.

Last piece of the rollout: build the retention file and keep it with the property records, not with that year's tax file. You will need the component-level detail when you make a partial asset disposition election on a future renovation, when you compute recapture at sale, and if the return is ever examined. Studies get lost in bookkeeper turnover and CPA changes with depressing regularity, and reconstructing one five years later costs more than the original.

Related questions

Can a tenant run a cost segregation study, or only an owner?

Tenants can. If you paid for the improvements and hold the depreciable basis, you can segregate them. The analysis differs — leasehold improvements and Qualified Improvement Property rules come into play, and the lease term interacts with the recovery period — so have an advisor evaluate both paths.

Does a cost segregation study increase audit risk?

Not inherently. It is a standard, IRS-recognized approach, and the agency publishes its own audit techniques guide describing acceptable methodology. Risk comes from weak studies — rule-of-thumb allocations without engineering support or source cost documentation. A properly documented engineering-based study is a routine, defensible position.

What happens to the accelerated depreciation when I sell the building?

It gets recaptured. Five- and seven-year personal property is subject to Section 1245 recapture at ordinary income rates; real property falls under Section 1250 treatment. You keep the time-value benefit already banked, but model the recapture before you sell. A 1031 exchange can defer it.

Is it too late if the buildout was finished years ago?

No. A look-back study paired with Form 3115 and a Section 481(a) adjustment captures the entire missed depreciation as a catch-up deduction in the current year. No amended returns are required, and there is no penalty for the prior treatment.

How does this compare to just electing Section 179?

They are complements, not substitutes. Section 179 has an annual dollar cap and a purchase-based phase-out, and it reaches 39-year systems like roofs and HVAC that a study cannot reclassify. Bonus depreciation on segregated assets has no dollar cap. Use both.

FAQ

Does a cost segregation study work for any type of commercial buildout?

It applies to most commercial improvements — office, retail, medical, industrial, hospitality, and restaurant. What varies is how much moves. Use types with heavy dedicated electrical, specialty plumbing, decorative finishes, or significant site work reclassify a larger share. A plain office fit-out with mostly drywall and building-standard mechanical work reclassifies the least, though even there 15% to 25% is common.

How much can I actually save?

Model it as net present value rather than gross deduction. A useful benchmark is $13,000 to $25,000 of NPV per $100,000 shifted from the 39-year schedule into a year-one deduction, at an 8% discount rate. On a $1,000,000 buildout reclassifying 30%, that is roughly $40,000 to $70,000 of benefit against a fee typically between $5,000 and $15,000.

Do I need to run the study before construction starts?

Not required, but earlier is better. Engaging during construction lets the engineer document components before they are concealed, which improves both the allocation and the supporting record. What does matter is finishing the study in time to take the position on a timely filed return, including extensions, for the placed-in-service year.

How long does a study take and what does it cost?

Two to six weeks from engagement to final report is typical, with the site visit early in that window. Fees generally run $5,000 to $15,000 for a buildout under $2 million and $15,000 to $50,000 for larger or multi-property engagements. Most reputable firms provide a free feasibility estimate before you commit.

Can I claim bonus depreciation on the reclassified assets?

Yes — that is the point of the study. Bonus depreciation applies to property with a recovery period of 20 years or less, which covers the 5-, 7-, and 15-year buckets. The applicable percentage has changed repeatedly under the statutory phase-down and subsequent legislation, so confirm the rate for your specific placed-in-service date with your CPA.

What is the smallest project where this makes sense?

Roughly $500,000 of depreciable basis is the conventional floor, and below that the fee often eats too much of the benefit. Between $500,000 and $750,000 it depends on your bracket and use type. Above $750,000 in a high-reclassification category, it is usually a clear win — but always start with a free feasibility estimate.

Sources

• https://www.irs.gov/businesses/cost-segregation-audit-techniques-guide
• https://www.irs.gov/publications/p946
• https://www.irs.gov/forms-pubs/about-form-3115
• https://www.irs.gov/newsroom/new-rules-and-limitations-for-depreciation-and-expensing-under-the-tax-cuts-and-jobs-act
• https://www.irs.gov/businesses/small-businesses-self-employed/tangible-property-final-regulations
• https://www.aicpa-cima.com/
• https://www.journalofaccountancy.com/
• https://www.thetaxadviser.com/
• https://www.cbre.com/insights

Related on PULSE

• [How Do I Depreciate Leasehold Improvements (QIP) to Save on Taxes?](/knowledge/q13793)
• [How Do I Value-Engineer a Buildout to Cut 20% off the Cost?](/knowledge/q13684)
• [How Do I Avoid Double-Paying Property Taxes in an NNN Lease?](/knowledge/q13660)
• [How Do I Sublease My Space to Cut My Rent?](/knowledge/q13673)
• [What vendor consolidation strategies are GTM teams using to cut tech stacks by 40% in 2027?](/knowledge/q16473)