Evaluate virtual-staging software

Define one task that every candidate must complete

Write the intended output and recipient before starting a trial. For example, you may need a clearly labelled furnishing concept for a private seller discussion, or a reviewed image package for a particular listing channel. Those tasks can have different requirements. Record the room, source view, allowed changes, fixed features and person who decides whether the output is useful.

Give each candidate the same bounded brief where its documented workflow permits it. Do not conceal important differences in the input or revision request and then compare the results as though the tasks were identical. If a candidate cannot accept the required source or deliver the required file, record that mismatch. It may make the candidate unsuitable for this task without implying that it is a poor product for every use.

Establish source retention and edit boundaries

Keep an unchanged original in your own records and assign an identifier to each output. Ask how the candidate associates the original with revisions and whether the recipient can distinguish the current result from an earlier attempt. Test the actual retrieval and export steps that matter to you. Do not infer source retention from the fact that a thumbnail remains visible after generation.

Use a room with identifiable openings and fixed features, with appropriate permission for the trial. Compare walls, windows, doors, permanent finishes and condition against the source. Record a concrete difference and its consequence instead of writing only “looks realistic”. A convincing image can still change an opening or hide a defect. Request rework or reject such a result; a concept label does not correct a misleading physical change.

Keep visual fidelity separate from measured fit. If the task depends on furniture dimensions or clearance, obtain those dimensions independently and retain the layout assumptions. A rendered sofa appearing to fit is not a measurement. The linked image-quality guide supplies a more detailed review; use the same review standard for each candidate rather than making exceptions for the most attractive output.

Test a revision and a recovery path

Ask for one specific revision, such as a different movable furnishing choice while preserving the room’s architecture. Record the original brief, requested change and resulting version. Check whether an accepted fixed feature changes during the revision. If it does, the new version needs review even when the requested change itself was completed. Keep earlier files identifiable so a reviewer can reconstruct the sequence.

In a controlled trial, inspect what happens when an operation does not finish or the session is interrupted. Determine whether the request status is clear, whether a retry creates another chargeable operation and whether completed files remain retrievable. These are questions to test or obtain documented answers to, not capabilities assumed of every provider. Avoid interrupting a live client job to discover its recovery behavior.

Record whether a failure is recoverable within your workflow or requires support, another request or a new source. Ask who handles the issue and what evidence they need. Do not turn an advertised turnaround into a guarantee for your particular revision or interruption unless the actual agreement provides that commitment. An unresolved delivery dependency belongs in the decision record.

Read rights and service terms for the intended use

Confirm the rights to the original separately from the service’s terms for processing it. In the United States, the Copyright Office explains that the photographer is generally the initial copyright owner, with limited exceptions such as qualifying works made for hire. Commissioning a photograph or owning the property does not automatically settle every editing, transfer or publication right.

Record the current terms, agreement date and relevant clauses or provider answers for your intended use. Ask about permitted source material, reuse of outputs, sharing with clients and any restrictions on commercial publication. Do not assume that paying for a plan resolves all those questions. If the source licence and service terms appear inconsistent, hold real-property use and seek clarification from the responsible parties or an appropriate adviser.

• U.S. Copyright Office: photographer ownership and exceptions

Compare quoted units rather than headline prices

Request a quote for the same defined task and write down what the billable unit means. It might be an image, a request, a room, a credit or a period of access; do not equate those units without evidence. Record the quoted plan, currency, date, taxes if stated and what happens when a request fails or a result needs revision. Leave missing figures unquoted rather than inserting a market estimate.

Separate included revisions from additional attempts and distinguish an edit to an accepted result from a fresh design request. Ask whether output size, download access, extra users or commercial use affects the quoted scope. Compare the total requirements of your task against the written inclusions, not a promotional number detached from the service conditions. This guide supplies no current vendor prices or savings claim.

Open the delivered files in the recipient’s workflow

Specify the required image format, usable size and associated source or review files, then open a real trial delivery where the recipient intends to use it. Check that filenames, room identity, labels and version references survive the transfer. A download link is not the same as an accepted deliverable. Record who opened the file and whether anything had to be recreated manually.

Ask what remains accessible if a subscription ends or an account is closed, and obtain an answer for the actual plan being considered. Do not infer long-term access from a successful download today. Keep your own approved source and output record within your organization’s agreed process. If the recipient needs something other than the supported output, identify that gap before purchasing more work.

Check access, deletion and retention evidence

List who needs to upload, review, download and share the material, then check the candidate’s documented access controls and the operations available in your trial. Record whether a shared link reaches only the intended audience under the tested settings. Keep observations narrow: seeing one access setting is not an independent security audit or a claim that all sensitive information is protected.

Read the current retention and deletion information and ask what a deletion action covers. Distinguish removing a visible project from a stated retention policy for underlying files, backups or other copies. If the documentation does not answer the question, record it as unknown and ask the provider. Do not claim data is immediately or permanently deleted merely because it disappears from the interface. Use non-sensitive, authorized trial material while essential questions remain unresolved.

Keep publication approval separate from software selection

For REALTORS®, the NAR 2026 Code addresses truthful advertising and misleading online imagery. Its scope is professional conduct for REALTORS®, not a universal publication rule. NorthstarMLS’s page dated 10 July 2026 describes disclosure and original-image pairing, but calls the material proposed guidelines while using mandatory language in parts of the body. Confirm current applicability with that channel before relying on it for a particular listing.

Use the linked disclosure guide for editorial wording examples and the channel, jurisdiction, rule URL, check date and reviewer record. A provider describing its images as ready for listings does not establish acceptance by your destination. Test the actual presentation and retain the unchanged original. Do not choose software on the assumption that its default label settles every local or channel-specific requirement.

• NAR: 2026 Code of Ethics for REALTORS®
• NorthstarMLS: 10 July 2026 proposed virtual-staging guidance

Blank candidate comparison worksheet

Complete one column or sheet per candidate with the same task and review date. Identify whether each entry comes from official documentation, a written quote, a provider response or your own observed trial. Record an owner and next action for missing evidence. Do not add the states into a single score if that would hide an essential unresolved requirement.

Fictional comparison: neither candidate is ready

The two candidates below are invented and represent no real service. In this fictional exercise, Robin needs a reviewed furnishing concept and a file that a colleague can open. Candidate A’s sample file opens, but its additional-revision terms are unknown. Candidate B has a documented fictional revision scope, but its architecture review has not been completed. Neither is a winner or approved purchase.

Robin requests the missing evidence instead of assigning a reassuring score. The fictional table contains no quoted amounts, security conclusion or performance result. In a real evaluation, the next decision would follow the completed evidence for the same task. A candidate could remain on hold or be rejected if an essential condition cannot be met, even if its other results look useful.

Keep the Spaceoverlay comparison factual

If Spaceoverlay is part of your room workflow, distinguish its scan record from independent photography. New room capture requires a LiDAR-equipped iPhone and iOS 26 or later. A scan saves geometry and a model image, with a plan available after at least two walls are captured, not a room photograph. This guide does not promise in-app import of original property photos. Without LiDAR, existing studio rooms can still be used, but new ones cannot be created.

Apply the same evidence discipline to any proposed workflow: verify the actual source available, the result produced and the file handed to the recipient. Do not infer feature parity from a shared technology or from examples on a website. End the evaluation with a specific task decision, its evidence and remaining limits, not a claim that one application is universally best.

Sources

• U.S. Copyright Office: photographer ownership and exceptions — checked 2026-10-09
• NAR: 2026 Code of Ethics for REALTORS® — checked 2026-10-09
• NorthstarMLS: 10 July 2026 proposed virtual-staging guidance — checked 2026-10-09

Explore Spaceoverlay

• Calculate a budget from your own quote

• Record disclosure and channel requirements
• Review output against the original

• Virtual staging
• Floor plan app