How to find off-market development land: a repeatable sourcing workflow

To find off-market development land, define the homebuilder's buy box, search parcels in the target market, and document why each candidate deserves more research. Keep unknowns visible. After a sourcing round, review distinct acquisition opportunities and the evidence that moved them forward or stopped them, so the next search starts with a specific question to resolve.

This guide is written by Pillar. It presents a suggested team method with a fictional worked example, not customer results or a test of any vendor's search coverage. The worksheet works in a shared document or spreadsheet without a paid tool.

1. Write a buy box that can rule sites out

“Land near a growing city” leaves too much room for interpretation. Write down your target geography, intended use, acreage range, acquisition budget, and timing. Separate hard exclusions from preferences.

For example, a homebuilder might start with a target county and a gross acreage range, require further investigation of access and utilities, and prefer proximity to an existing community. That last preference should not silently become a hard filter. A promising parcel could survive it.

Keep three columns in your brief: required, preferred, and unknown until investigated. Only filter on a field when you understand its coverage and meaning. Missing information belongs in the third column.

2. Search parcels, then inspect the context

Use your criteria to build an initial parcel list. Review the surrounding map before ranking it. An address alone can hide an awkward boundary, a neighboring use, or an assemblage question.

Pillar provides a conversational way to research sites and work with maps. Its AI agent guide explains how to start that workflow. A useful first request is narrow: name the market, describe the parcel criteria, and ask for a manageable set to review. Refine your request after inspecting the results.

A blank listing field does not establish marketing status. Record the listing sources or owner/broker information you reviewed and the date checked. "No listing found in the sources reviewed on this date" describes that observation; it does not establish that the property is available to buy or absent from every marketing channel.

3. Make every shortlisted parcel explainable

For each candidate, keep a short record:

• Parcel identifier and jurisdiction, with the source and date checked.
• Why it matches your buy box.
• The question most likely to change your decision.
• Ownership information and the next verification step.
• A named next action, its owner, and a review date.

Avoid turning every observation into a score. A high average can hide a single unresolved issue that determines whether the site works. Keep decisive questions visible beside any ranking.

4. Resolve identity before planning outreach

Start owner research with the parcel record and its jurisdiction. Check whether your research concerns one parcel, several adjoining parcels, or a portion of a larger holding. A mailing address and the property address answer different questions; preserve both when available.

When a record names an entity, do not assume that the first person found in a contact search can speak for it. Keep the recorded owner separate from a possible contact and document what connects them. Pillar's owner research workflow is a place to begin organizing that investigation.

5. End each review with a decision

Use three outcomes: investigate, hold, or pass. For an investigate decision, specify what would move the parcel forward. “Check utilities” is broad. “Identify the serving utility and request the service information needed for this proposed use” is actionable.

A hold needs a reason and a revisit trigger. A pass needs a short explanation so a teammate does not repeat the same research next month.

Review a sourcing round with a fixed population

Download the sourcing-round review worksheet before starting the next search. Save a round ID, geography, buy-box version, search date and observation cutoff. Record the sources, their dates, the fields you filtered, and any coverage or export limit. An incomplete result set describes the records you reviewed; it does not count every qualifying property in the market.

Keep these units separate:

• Exported rows are the records in the saved input, including any duplicates.
• Unique parcels are distinct identifiers within their jurisdictions after you resolve duplicate records. Retain the original input and document the match rule.
• Acquisition opportunities are the pursuits your team defines. One opportunity can require several parcels or owners, while the same owner can have separate opportunities.
• Outreach attempts are individual contact events. Several attempts can concern one opportunity, and one mailing can concern several parcels.

Assign a stable opportunity ID and list its constituent parcels and ownership dependencies. Adjacency or an identical owner name alone is not a sufficient grouping rule. Record why the parcels belong in one proposed pursuit; grouping them does not establish site control.

At the cutoff, give every opportunity one current investigate, hold or pass decision. Record a primary reason, its evidence, any unresolved question, and the person responsible for the next action. Keep secondary issues in the record, but do not add overlapping reason counts and call the result a count of opportunities. Preserve dated decisions when a case moves between states.

A pass can follow a documented buy-box mismatch or a team decision about timing and resources. Record which it is. Missing access information, a returned letter, or silence from an owner does not establish that the land fails the buy box or that the owner will not sell.

Worked example: count opportunities before judging the search

This invented example describes eight proposed pursuits in one fictional jurisdiction. It contains no real parcels, owners, delivery events or customer outcomes. The worksheet includes the full example so another analyst can reproduce the count.

Suppose the saved input contains 12 rows. Two rows repeat earlier parcel records, leaving 10 unique parcels. The team defines eight opportunities: A and B each contain two parcels, while C through H each contain one. At the observation cutoff:

• A, B and C are marked investigate, each with a specific research action.
• D is on hold for missing access evidence. E is on hold while the team reviews a returned-mail event and its address evidence. F is on hold with no owner reply recorded by the cutoff.
• G is passed because its documented gross acreage falls outside the frozen buy box. H is passed because its location is outside the agreed geography.

The decision counts reconcile: 3 investigate + 3 hold + 2 pass = 8 opportunities. The investigate share is 3 / 8, or 37.5%, for this fictional cutoff. Dividing by 12 exported rows would answer a different question. Neither fraction is an acquisition win rate or a market benchmark, and the three investigate cases have not been established as feasible sites.

The holds need different actions: retrieve access evidence for D, reconcile the postal and address record for E, and retain an appropriate follow-up decision for F. These counts do not establish that a broader acreage range would improve the search. Keep the original cohort and record any later change to the criteria as a new version.

Separate outreach evidence from seller interest

If the sourcing round includes direct mail, retain the actual postal event and date separately from an owner's expressed interest. USPS Domestic Mail Manual 507, section 1.4.1 lists nondelivery endorsements, including "Insufficient Address" and "Attempted—Not Known." The manual also explains that forwarding, return or other treatment depends on the mail class and applicable service.

Record the observed wording and service used. A returned piece alone does not establish whether your underlying address record is incorrect, whether the recipient can act for the owner, or whether the owner would sell. If the only observations are no return and no reply, delivery and interest remain unconfirmed. Reconcile conflicting parcel, owner and address records using dated sources before choosing the next permitted contact step. The LLC owner research guide explains how to keep the recorded entity separate from a possible contact.

Choose the next research change

Compare rounds with the same definitions and comparable observation ages. Record time since search and, where outreach matters, time since each first contact attempt. A round reviewed two days after mailing and one reviewed a month later have had different opportunities to produce a reply. If geography, source coverage, staffing, criteria or observation time changed, record that difference before attributing a result to the search method.

Use the documented reasons to choose one next question. Repeated missing access evidence may justify finding a better access source earlier in the review. Repeated verified acreage mismatches may call for checking the filter definition and source field. Where the evidence is still unresolved, keep an owner and due date rather than treating the case as a failure. Write down the change you will try and what observation would make you keep or reverse it.

Use the desktop site screening checklist to structure the next candidate review. When a site needs explicit scenarios, assumptions and specialist handoff, the land feasibility guide covers that later decision.