Research

HUD Digital Housing Advertising Guidance: A Record-Review Method for Community Marketing

A source-led method for reviewing housing ad inputs, platform settings, delivery evidence, approvals, and uncertainty under HUD's digital-platform guidance.

MobileHomeBiz Research Desk9 minutes read4 sources citedVerified 2026-09-18
HUD Digital Housing Advertising Guidance: A Record-Review Method for Community Marketing research illustration
Key finding: HUD guidance explains that Fair Housing Act risk can arise from ad targeting and delivery through digital platforms, including automated systems.

Key takeaways

  • Define the public dataset before extracting it.
  • Keep aggregate evidence separate from property records.
  • Record uncertainty and route decisions to the accountable owner.

Published September 18, 2026

Research question

Which records should a manufactured-home community preserve when digital platforms help target or deliver housing advertising?

Bottom line

A useful review covers both the advertiser's selections and the platform's delivery evidence. A compliant-looking ad image alone cannot show who was included, excluded, reached, or able to respond. The guidance is a legal-risk source, not a finding that a particular campaign violated the law. The operational value comes from using public evidence to frame a narrower local check, not from converting an aggregate or regulatory source into a property conclusion.

This distinction matters for manufactured-home businesses because one research request often crosses several units of analysis. A federal source may describe households, covered entities, insurance claims, establishments, advertisements, or communications. A community file may describe homes, spaces, residents, vendors, tasks, or transactions. Those units do not become comparable merely because they appear in the same spreadsheet.

What the primary sources establish

The first source listed below is the anchor. The remaining official materials supply definitions, method, interpretation, or an adjacent control. We reviewed the live source pages on September 18, 2026 and did not rely on a commercial summary for a factual claim.

The verified fact for this brief is: HUD guidance explains that Fair Housing Act risk can arise from ad targeting and delivery through digital platforms, including automated systems. That statement is intentionally bounded. It does not establish the condition, value, compliance status, performance, or future outcome of a particular manufactured-home community.

The source set also demonstrates why definitions must travel with every copied figure or rule excerpt. The publisher may specify a universe, covered party, reporting system, reference period, geography, revision status, privacy treatment, or suppression rule. Removing that qualifier creates a new claim that the source may not support.

Reproducible review method

Start with a written question and name the decision it might inform. For this brief, that decision is approval or renewal of a housing marketing campaign. Do not begin by searching for a persuasive number. Beginning with the decision reduces the temptation to keep a statistic merely because it looks precise.

Create a source log with publisher, exact title, direct URL, access date, product or regulation name, universe, unit, geography, period, revision status, and relevant warning. For a downloadable file, add the filename, release label, field names, filters, row count, and a checksum when practical. For a rule or guidance document, add the section, defined term, effective or issue date, and applicability question.

Next, reproduce the extract. Save filter choices in words rather than relying only on a screenshot. Keep estimates and margins of error together. Preserve suppressed, unavailable, not-applicable, or redacted values as published; never silently turn them into zero. If an interactive page and download disagree, record both versions and investigate instead of selecting the preferred value.

Finally, have a second reviewer answer four questions: Does the cited source support the sentence? Does the sentence retain the source's scope? Is analysis labeled as analysis? Is the proposed business action supported by local evidence? A review that cannot answer all four should stop before publication or decision use.

Evidence boundary

Evidence layer What belongs here What it can support
Federal source Published definition, table, dataset, rule, guidance, and method A statement within the publisher's stated scope
Research workpaper Filters, extraction date, calculations, caveats, and reviewer notes Reproduction and transparent analysis
Local operating record approved copy, images, audience settings, excluded audiences, campaign dates, delivery reports, lead-source exports, vendor terms, change history, and counsel-approved policy A fact about the named business or property, subject to verification
Professional judgment Legal, accounting, insurance, engineering, appraisal, or other qualified review A decision within that professional's engagement

Public evidence should not be promoted to a stronger layer. A national estimate does not become a local count. A dataset row does not prove causation. Guidance does not itself prove a violation. An absent record does not prove an absent event. An industry average is not a local price.

Operating workflow

A marketing assistant can archive the exact creative, settings, platform report, approval, and run dates. The designated decision owner should review targeting choices and escalate legal interpretation to qualified fair-housing counsel.

Use a queue with six statuses: requested, source confirmed, extracted, locally reconciled, exception raised, and owner approved. Attach the source log and workpaper before moving an item to locally reconciled. Record who approved the final use and which version they reviewed.

The assistant's role is evidence administration: find the approved source, preserve its context, perform documented transformations, compare defined fields, and surface exceptions. The role does not include making legal conclusions, altering an unfavorable result, filling missing facts by assumption, or presenting a research screen as professional advice.

For recurring work, set a refresh trigger based on the source rather than an arbitrary daily edit. Triggers can include a new annual release, changed guidance, revised dataset schema, new local document, changed vendor relationship, or upcoming owner review. Keep the earlier workpaper so a later reviewer can explain what changed.

Quality-control tests

Identity test. Confirm that the publisher and URL are official, the title matches the cited item, and the version is current for the question. Archived versions may still be appropriate, but label them.

Unit test. Write the unit in plain language. If the numerator is claims and the denominator is properties, or the source covers occupied homes while the local file includes vacant spaces, do not calculate a rate until the mismatch is resolved.

Time test. Separate the source's reference period, release date, extraction date, and decision date. A current webpage can display an older reference period.

Geography test. Record the exact geographic level and boundary. County, city, ZIP Code, tract, service area, and property address are not interchangeable.

Applicability test. For legal or program material, identify the covered entity, transaction, activity, and exceptions. Record uncertainty. Obtain qualified review instead of treating a broadly relevant source as universally applicable.

Local reconciliation test. Identify the record that would confirm the property fact. When the public and local evidence differ, investigate definitions, dates, and coverage before calling either one wrong.

Common failure modes

One failure is false precision: copying a value without its uncertainty, disclosure treatment, or revision status. Another is denominator drift: dividing a local count by an unrelated public total. A third is applicability drift: turning guidance for defined parties into a universal script. A fourth is recency theater: changing a visible date without repeating the review. A fifth is silent transformation: excluding inconvenient rows or changing missing values without a calculation note.

Avoid these failures by keeping raw evidence read-only, doing calculations in a separate workpaper, labeling every transformation, and retaining the review trail. If evidence is incomplete, the useful conclusion may simply be that a local verification step is required.

Decision checklist

  • Is the business question written before extraction?
  • Are universe, unit, period, and geography recorded?
  • Are estimates, uncertainty, redactions, and suppressions preserved?
  • Are verified facts separated from our analysis and inference?
  • Is the applicable local record identified?
  • Has a decision owner reviewed consequential assumptions?
  • Are professional questions routed to a qualified reviewer?
  • Can another person reproduce the extract from the workpaper?

If any answer is no, the item remains research in progress. It should not be promoted into a property claim, resident communication, campaign rule, staffing commitment, or financial decision.

Practical handoff for a virtual assistant

The handoff should contain the question, source log, saved extract or cited section, calculation notes, limitations, local-record request, exception list, reviewer, and next review trigger. Use access controls appropriate to resident, financial, insurance, or vendor records. Do not place private operating data in a public research file.

MobileHomeBiz provides marketing support and administrative support for teams that want repeatable evidence workflows. The service connection is administrative: organizing sources, records, queues, and handoffs. It is not a claim that public data can replace local diligence or professional advice.

Sources reviewed

  1. Guidance on Fair Housing Act advertising through digital platforms. HUD Office of Fair Housing and Equal Opportunity. https://www.hud.gov/sites/dfiles/FHEO/documents/FHEO_Guidance_on_Advertising_through_Digital_Platforms.pdf. Checked 2026-09-18.
  2. Fair Housing Advertising, 24 CFR part 109. U.S. Department of Housing and Urban Development. https://www.hud.gov/sites/dfiles/FHEO/documents/BBE%20Part%20109%20Fair%20Housing%20Advertising.pdf. Checked 2026-09-18.
  3. Fair Housing Act overview. U.S. Department of Housing and Urban Development. https://www.hud.gov/helping-americans/fair-housing-act-overview. Checked 2026-09-18.
  4. Fair housing advertising resources. U.S. Department of Housing and Urban Development. https://www.hud.gov/hud-partners/fair-housing-advertising. Checked 2026-09-18.

Conclusion

The decision-grade conclusion is deliberately narrower than a marketing headline. A useful review covers both the advertiser's selections and the platform's delivery evidence. A compliant-looking ad image alone cannot show who was included, excluded, reached, or able to respond. The guidance is a legal-risk source, not a finding that a particular campaign violated the law. A well-run research workflow records what is known, what is inferred, what remains uncertain, and which local evidence must be checked next. That boundary makes the work more useful because an owner can see exactly which claim is reproducible and which decision still requires judgment.