A chain's real estate team rarely lacks sites. Brokers send pads, franchisees send corners, the field team sends a lead. The work is the screen: who owns the site, what can be built, whether the prototype fits, what is in its past, and what land costs nearby. The six questions below are the ones site teams ask before a visit. Each is answered in the first paragraph, then shown on one finished Titleman run on three real vacant pads.
Prepared by Titleman AI, on behalf of the Titleman team. The worked example is run 26177, made on a Titleman QA account on 16 October 2026. The three addresses, owners and the metro's name are withheld, and the pads are called A, B and C. A feasibility run is a screen for your team to review. It is not an appraisal, a design or a go-or-no-go.
How do restaurant and retail chains screen a new site fast, and what does a site feasibility check include?
A site feasibility check answers six things for each candidate: who owns it and since when, what the rules let you build, whether your prototype fits, what drainage and ground cost you, what land is worth nearby, and what is still open. Titleman produces all six from addresses in one run. The run read for this page took 48.7 minutes.
What the run was given, in one message and no files: three vacant pads from brokers on one corridor of a large Texas metro, each 1 to 2 acres, for a freestanding drive-thru prototype. The prototype was a 2,800 SF building, one drive-thru lane with room to queue 12 cars, and about 35 parking spaces.
What it did and returned:
- Read: 25 searches, 84 pages and 161 records on owners and transfers, permit histories, plat and driveway rules, drainage criteria, flood maps, traffic counts and nearby drive-thrus.
- Worked: 94 figures, each with its source, in one side-by-side table. Also a test-fit of the prototype, pond sizes, and ground rent, build-to-suit rent and land residual math as live formulas.
- Flagged: 9 red flags and 7 open items, listed in the memo instead of smoothed over.
- Delivered: a memo in PDF, an Excel workbook, a site-rules report and a deck.
The speed comes from starting at the addresses instead of at three blank packets. The same screen works for a store, a convenience store or a clinic. The prototype row changes. The owner, rules, history and land rows stay.
How can a site selection team compare several broker pads side by side before anyone visits?
Put every pad through the same checklist and read the answers in one table: owner and holding period, size, frontage, driveways allowed, land left after the prototype, drainage, traffic and nearby competitors. Differences then show up as rows, not as three separate packets. In the run, the land left after the prototype ranged from 16,216 SF to 25,656 SF.
The run's side-by-side, with names left out:
| Measure | Pad A | Pad B | Pad C |
|---|---|---|---|
| Size on the tax roll | 1.8 acres | 1.5 acres | 1.38 acres |
| Frontage on the main road | 303 ft | 218 ft | 200 ft |
| Driveways the rules allow on that frontage | up to 3 | up to 2 | up to 2 |
| Owner since | July 2021, about 5 years | December 2022, about 4 years | April 2026, about 6 months |
| Tax-roll land value, 2026 | $470,448 ($6 per SF) | $653,400 ($10 per SF) | $600,000 ($10 per SF) |
| Nearest state traffic count (vehicles a day) | 33,450 (2021) | 28,991 (2021) | 32,774 (2022) |
| Drive-thru competitors listed within 0.75 mile | 4 | 2 | 2 |
| Mapped flood-hazard area at the parcel | none | none | none |
| Land left after the prototype | 25,656 SF | 19,254 SF | 16,216 SF |
All three were vacant on the 2026 tax roll. The land values are the roll's own figures, not asking prices: no asking price for any pad was in the material the run had. The flood row comes from a map copy. The official flood panel could not be reached, so the run listed it as open.
A table like this does not pick the site. It tells the team which pad to look at first and which question to ask each broker. For the method behind the comparison, see how we screen a comparable set.
Will our prototype fit on the lot, and how do you test that before a site visit?
Add up the program (building, parking, drive-thru lane and queue), then take away what the rules take first: the building line, stormwater detention and landscaping. What is left is spare land. In the run, a 2,800 SF prototype needed 17,930 SF and fit on all three pads on paper, with 16,216 to 25,656 SF spare. The harder test is geometry.
The program in the run was 2,800 SF of building, plus 35 spaces at 350 SF each, plus a queue of 12 cars at 20 ft each in a 12 ft lane: 17,930 SF in all. Three things come off each lot before anything else:
- A 25 ft building line along the frontage, because both roads are planned thoroughfares.
- A surface detention pond at 4 ft deep.
- 10% of the lot for landscaping.
Drainage took the most land. Stormwater detention is required on commercial sites in that jurisdiction with no exemption for 1 to 2 acre pads, at 0.75 acre-feet per acre when the pad drains to a storm sewer. A watershed rule adds 0.24 for commercial projects over one acre, so 0.99 acre-feet per acre in all. A roadside-ditch outfall needs more.
| Measure | Pad A | Pad B | Pad C |
|---|---|---|---|
| Detention plus retention, storm-sewer outfall | 1.782 acre-ft | 1.485 acre-ft | 1.364 acre-ft |
| As a surface pond | 19,406 SF | 16,172 SF | 14,854 SF |
On every pad the pond is larger than the parking field (12,250 SF). Underground storage frees the land, at higher cost.
Area is not the binding test. Geometry is: whether a 12-car queue can wrap the building without blocking the driveway entrance. Pads B and C have 218 ft and 200 ft of frontage and are about 300 ft deep, so that is the first thing a test-fit plan should draw. The run also noted that the jurisdiction has no zoning ordinance, so use is governed by deed restrictions, plat notes and building lines, and a left-turn lane is required at driveways on divided roads unless a variance is granted. For the zoning side of a screen, see parcel data and zoning check before underwriting.
What should we check about a site's history and title before spending money on a visit or a test-fit?
Check the past of the address, not only the lot: permits, business registrations, a fuel use next door, and how recently it changed hands. Then list what only a document can settle: deed restrictions, plat notes, easements and the official flood panel. On three pads that all showed as vacant, the run found nine red flags and seven open items.
Three of the run's nine red flags:
- A fuel station on the neighboring parcel, same owner as Pad A, permitted in 2018. Possible underground tanks next to the pad. The run's step: an environmental Phase I before Pad A.
- A commercial building registered at Pad B's street address while the tax roll shows nothing built on the pad. The run's step: confirm on site or from the air where it stands, and look for shared access or drainage easements.
- A permitted building project on Pad C in 2023 and 2024 that never reached the tax roll. The pad changed hands six months before the run and is unplatted acreage, so a plat is likely needed before permits.
"Vacant" on the roll had a past on every pad: the third, Pad A, shows tenant-suite permits at its address that the roll places on the neighboring parcel. That is the reason to read the address history before the visit.
A selection of the open items and next steps, with the document that closes each:
| Open item | What closes it |
|---|---|
| Flood zone letter, panel and date (the official map service did not answer) | The official flood map printout for each pad |
| Wetlands (the inventory returned an error) | A wetlands check |
| Deed restrictions, plat notes and easements, including any ban on drive-thru use | A title commitment for each pad |
| Asking price of each pad (none was in the material) | The broker, and any closed land sales the broker will share |
| Driveway locations, median openings, left-turn lane, outfall and detention | A pre-development meeting with the local engineering office |
| Build cost of the prototype on each pad | A contractor's estimate |
The run also listed what its sources do not cover: liens, code violations and historic designation in that jurisdiction. A blank that stays blank is better than a plausible number.
How do we get comparables and a land value for a new location when sale prices are not on record?
Say which kind of evidence you hold. Where sale prices are not disclosed, comparables are asking prices and tax-roll values, not closed sales. Then work the number from your own economics: a ground rent, a build-to-suit rent and a residual land value, each from stated inputs you can change. The run gave no price verdict. It gave the three methods.
The pads' own numbers in the run:
- Tax-roll land value: $470,448, $653,400 and $600,000, which is $6, $10 and $10 per SF. These are appraisals for tax, not market prices.
- Nearby land asking prices: 0.83 acre listed at $850,000, or $23.51 per SF, and a 1.837 acre corner marketed for quick-service restaurants at $900,000, or $11.25 per SF. These are asks, not sales. The state does not disclose sale prices, and the run held no recorded land sale near the pads. A small pad also lifts the per-SF figure.
The run then showed how each method is worked, with its own assumptions in the workbook as live formulas:
- Ground rent: the land cost times a ground-lease rate of 5.5% (tested from 5.0% to 6.0%).
- Build-to-suit rent: land cost plus development cost, at a developer's return on cost of 7.5%. Development cost was 2,800 SF at $375 per SF, site work at $10 per SF of lot, 12% soft costs and 5% contingency.
- Residual land value: the net rent a new drive-thru pays ($65 per SF), capitalized at 6%, less the same development cost and a developer profit of 10% of cost.
These are the run's assumptions, not market facts. Build cost moves the answer most: on Pad B, a hard cost from $300 to $450 per SF moves the residual by about a third either way, and site work from $8 to $14 per SF of lot by up to two-fifths. A contractor's estimate for your prototype is the input that narrows it. For how this works where sale prices are hidden, see non-disclosure states and underwriting models, and for the residual method, how to calculate residual land value.
What does the 72 hours mean, and what stays with our team?
The run above took about 49 minutes. The 72 hours is a different clock: the start-up job in which our team wires one workflow of yours into the Titleman engine, for example a broker's site email that comes back analyzed in the same thread. Our team checks each package before it reaches you. The choice of site, the offer and the prototype stay with your team.
The run in this page was typed into the workspace, from a message with three addresses and the prototype. Wiring your inbox so that a broker's shortlist goes in by email and the comparison comes back in the same thread is the 72-hour job. The principle is: your stack, our engine for real estate with 2,100+ sources in the middle, the answer where you work. Your mailbox, lease files and models stay where they are. The answer lands in the channel your team picks, set up during the start.
What stays with your team:
- The choice of site and the offer. The run states no price verdict and no go-or-no-go.
- The prototype and the assumptions. Change the build cost or the rent in the workbook and every result moves.
- The documents. Title, plats, the flood panel and the contractor's estimate come from outside, and a screen only tells you which to order first.
For the corporate real estate version of this offer, see AI for corporate real estate teams.
Run it on your next three sites
Send the addresses of three candidate sites and your prototype, and we will walk you through the side-by-side, the land left after drainage, the red flags and the open items on your own sites. Request a demo.