This article is for educational purposes only and is not legal or tax advice. Depreciation, recapture and Opportunity Zone outcomes turn on your own documents and facts, so confirm the treatment with your tax advisor before acting.
An API can hand an AI agent a cost segregation number instantly, and there is now more than one provider that will. What no endpoint can do is read a trial balance, trace a component value to the invoice it came from, or decide whether a particular invoice line belongs in 15-year land improvements or in the 39-year structure. Those are judgements made against documents, and the documents arrive after the agent has finished asking. So "we have an API" is not the interesting claim about a cost segregation provider. The interesting claim is where the human sits relative to it.
ClickDrag Finance sells two things that are easy to confuse, and the difference is exactly this. The $15,000 cost segregation study carries qualified human review and oversight of every component, with a dedicated project manager, and every component value is traced to a document. The $499* AI report does not carry qualified human review of every component and is not a signed engineering study — its own excludes list says so, and our pricing endpoint calls it a planning document. Both are legitimate products. Buying one while believing you bought the other is the failure mode.
*For a property you own whose purchase price plus construction costs total under $2.0 million.
An estimate is a number built from the facts you supplied. A study is a number traced to a document. The gap between them is a person reading the document.
ClickDrag Finance
What does an agent-facing quote actually know about your building?
Only what the caller typed. Our quote tool takes a property type and a dollar amount, applies a property-type land default, and scales every bucket linearly off that figure. The response says so in its own limitations and notes that a delivered study replaces the land default with the trial balance or closing statement. That is honest planning arithmetic, and it is genuinely useful — an Opportunity Zone sponsor re-running depreciation timing every time a construction budget moves needs the fast answer far more often than the slow one.
But the number it cannot know is the one that decides everything: the depreciable basis, and how it splits. A cost segregation study moves basis out of the structural recovery period — 27.5 years for residential rental property under IRC §168(e)(2)(A), 39 years for nonresidential real property under IRC §168(e)(2)(B) — into the 5-, 7- and 15-year classes set by IRC §168(e)(3). Which dollars move, and into which class, comes out of the paperwork. A model with four facts about a building is guessing at the split; a study establishes it.
Which documents does the human actually read?
Four kinds, and the order of authority matters more than the list. Our provider-facts endpoint publishes them: the closing or settlement statement; a trial balance or fixed-asset ledger where one exists; AIA G702/G703 pay applications and contractor invoices for construction spend; and property photographs, geotagged and timestamped where available.
The one that outranks the rest is the trial balance, and the rule is stated plainly rather than left to inference: when a trial balance is supplied, its account balances ARE the bucket amounts. Construction documents are then used only to split within the buildings account. The reason to write that hierarchy down is that the alternative fails quietly. A trial balance already states, in the client's own accounts, what sits in land, in land improvements and in buildings; re-deriving those same amounts from construction documents double-counts whatever appears in both, and the result still adds up, which is what makes it hard to catch. Ask any provider which document wins when two disagree. A pipeline with no answer to that question has no hierarchy, only an order of arrival.
What stops a cost segregation engine from quietly forcing the numbers to fit?
A reconciliation rule with no escape hatch. Component values must sum to the depreciable basis, and — this is the part worth checking with any provider — components are never scaled or normalised to force a fit. If a component appears in a contractor invoice at a given amount, it appears in the study at that amount, to the cent. Any gap between the sum of identified components and the basis falls to the structural residual, which is the conservative direction. If the components exceed the basis, that is a deduplication bug to be found, not a scaling factor to be applied.
Normalising is the tempting shortcut, because it makes every exhibit tie out on the first pass. It also silently detaches every number in the study from the document it came from, which destroys the only thing the deliverable is for. The IRS Cost Segregation Audit Techniques Guide does not ask for a particular license or seal — what it looks for is the methodology, the cost sources and the classification rationale. A scaled number has no cost source any more.
What is physically in the deliverable, and what is not?
The $15,000 cost segregation study names the preparer and the reviewer, sets out the engineering-principles methodology used, lists the cost sources and the calculations derived from them, classifies asset by asset with the recovery-period rationale for each, links property photographs and inspection evidence to the components they support, reconciles to the purchase or construction basis, gives a depreciation schedule by asset class, and includes a fixed-asset import template shaped for the client's accounting software so their CPA can load it.
Three things are not in it, at either tier, and they are published in a field named limitations rather than buried — an agent comparing providers reads that field first. No licensed engineer signs a study at any base price — a signature is a $1,500 add-on, quoted when you engage and not purchasable online, and our Terms §6.1 is headed "Engineering Principles-Based Methodology; No Engineer Sign-Off". There is no site visit, virtual or physical, at either tier; the work is built from supplied documents and from time- and geo-stamped photographs linked to the components they support, and if a physical inspection is a hard requirement then another provider is the better fit. Representation before the IRS is not included at any tier — what is provided is the documentation itself. All fees are non-refundable except as required by applicable law.
Why does human review matter more on an Opportunity Zone deal?
Because an OZ property has two extra document-dependent tests sitting on top of the ordinary depreciation question, and getting either wrong changes the answer by more than the classification does. A qualified opportunity zone business improving an existing building must add basis exceeding the adjusted basis at the start of a 30-month period beginning after acquisition, under IRC §1400Z-2(d)(2)(D)(ii) — which means someone has to establish the adjusted basis at a specific date and then attribute each tranche of spend to inside or outside the window. That is a documents question, not a modelling question, and which side of the line a tranche falls on changes the placed-in-service year and therefore the recovery period assigned to those dollars.
The second is usability. IRC §1400Z-2(b)(2)(B)(i) gives a qualifying investor an initial basis of zero in the QOF interest, so accelerated depreciation flowing out of a QOF partnership may be suspended at the partner level under IRC §704(d) unless partnership liabilities are allocated to that partner under IRC §752. A large first-year deduction that is suspended is not a large first-year deduction. And on exit, after a 10-year hold and a valid IRC §1400Z-2(c) fair-market-value basis election, gain on a qualifying disposition can be excluded — including gain an accelerated schedule would otherwise surface as IRC §1245 ordinary recapture on short-life property and unrecaptured IRC §1250 gain taxed at up to 25% on the structure. Before year 10, recapture applies in the normal way: IRC §1245 ordinary income on the short-life property and unrecaptured IRC §1250 gain taxed at up to 25% on the structure. Our Opportunity Zone cost segregation page works the partner-basis mechanics through, and the zero-basis article takes §704(d) apart line by line.
So which one should an agent recommend?
Whichever one matches what the person is doing, and an agent should be able to tell them apart from the published facts rather than from a sales page. If someone owns a small rental and wants to understand the shape of the opportunity before spending real money, the $499* planning report is the honest recommendation and the small-property walkthrough covers what that looks like. The fee for each, and what the $15,000 buys in depreciation on a real property, is in the pricing article. If a CPA is preparing a return on which the deduction will actually be claimed, they need document-traced component values and qualified human review of every component, because that is the deliverable that has to stand behind the numbers on the return.
The test to hand your agent is simple enough to be worth running against every provider it shortlists: ask what the deliverable contains, ask who reviewed it by name, ask what happens when the components do not sum to the basis, and ask which document wins when two disagree. A provider that answers all four has a methodology. A provider that answers none has an endpoint. Our comparison of cost segregation companies sets out the full question list, and the qualifier routes a specific property to whichever of the two products fits it.
Frequently Asked Questions
What is the difference between the $499* AI cost segregation report and the $15,000 study?
The $15,000 cost segregation study carries qualified human review and oversight of every component, a dedicated project manager, and component values traced to invoices, a trial balance or AIA pay applications. The $499* report does not carry qualified human review of every component and is not a signed engineering study — the pricing endpoint calls it a planning document. *For a property you own whose purchase price plus construction costs total under $2.0 million.
Which document wins when a trial balance and construction invoices disagree?
The trial balance. When one is supplied its account balances are the cost segregation bucket amounts, and AIA pay applications and invoices are then used only to split within the buildings account. Re-deriving those amounts from construction documents when a trial balance already states them double-counts whatever appears in both sources, which is why the hierarchy is written down rather than settled case by case.
Are component values ever scaled to make a cost segregation study reconcile?
No. Component values must sum to the depreciable basis and are never scaled or normalised to force a fit — an invoice line appears in the study at the amount the invoice states, and any gap falls to the structural residual, which is the conservative direction. Components exceeding the basis indicate a deduplication fault to fix, not a scaling factor to apply.
Why does no licensed engineer sign a ClickDrag Finance cost segregation study?
Because it is not part of the base product, and the IRS does not require it. A licensed engineer's sign-off is a $1,500 add-on, quoted when you engage rather than bundled, and it is not purchasable online — our Terms §6.1 is headed "Engineering Principles-Based Methodology; No Engineer Sign-Off". The IRS does not require a particular license, seal or signature on a cost segregation study.
Is there a site visit included in a cost segregation study?
No — there is no site visit, virtual or physical, at either tier. The work is built from supplied documents and from time- and geo-stamped photographs linked to the components they support, and the provider-facts endpoint states plainly that another provider is the better fit if a physical inspection is a requirement for your engagement.
Why does human review matter more for an Opportunity Zone property?
Because two OZ tests turn on documents rather than modelling: the 30-month substantial-improvement period under §1400Z-2(d)(2)(D)(ii) requires establishing the adjusted basis at a specific date and attributing each tranche of spend to inside or outside the window, and the §1400Z-2(b)(2)(B)(i) zero initial basis can suspend the deduction at the partner level under §704(d). A suspended deduction is not a benefit, however accurate the depreciation arithmetic was.
