After the photographer delivers, four jobs still stand between the image folder and an MLS-ready listing: produce the narrated tour, write the property description, review the text for fair-housing concerns, and preserve provenance for every AI-edited photo. Treat them as one controlled production sequence, because a mismatch between files can create more risk than any single unfinished deliverable.
In September 1999, NASA engineers were trying to determine why the Mars Climate Orbiter was approaching Mars on an unexpected trajectory. The spacecraft was lost during orbital insertion. NASA’s Mars Climate Orbiter Mishap Investigation Board later documented the underlying failure: one engineering team produced data in pound-force seconds while another system expected newton seconds.
Each component had done a defined job. The failure sat between them.
That is the useful analogy for listing production. Your photographer can deliver excellent source images, yet the launch can still break at the handoffs: the video describes a feature the listing text omits, the fair-housing review happens before the final revision, or an AI-staged image reaches a portal without durable disclosure and access to its original.
Start with a controlled set of source photos
Create one approved source set before generating anything else. Separate the images cleared for marketing from duplicates, alternate exposures, detail shots, and files that should remain internal.
For virtual staging, keep the original beside each edited version. Use a naming convention that preserves the relationship, such as `living-room-03-original` and `living-room-03-staged`. Do not rely on visual memory when several angles show similar rooms.
Review the staged result for changes beyond furnishings and decor. If the process appears to alter a wall, window, fixture, view, room dimensions, or property condition, stop and compare it against the source. What Should You Do When AI Staging Changes a Wall? covers that review in more detail.
This approved image sequence becomes the common input for the remaining work. The tour, description, disclosure record, and provenance page should all refer back to it.
Build the tour and description from the same facts
A narrated tour needs order. Start with the exterior or arrival image if one is available, then move through the property in a sequence a viewer can follow. Avoid jumping from a bedroom to the kitchen, back to the bedroom hallway, and then outside.
The narration should describe what the images support. An AI voice can connect the stills into a coherent tour, but it cannot verify renovations, school assignments, permitted uses, views, or neighborhood claims. Those details require the agent’s review and reliable listing information. For a practical example of sequencing still photos, see Listing Photo Video: How Marcus Turned 27 Stills Into a Coherent Tour.
Write the MLS description from that same verified fact set. Include property details that help a buyer understand the home, while avoiding claims the photos or records cannot support. Then compare the final description with the narration. Names, counts, finishes, and room functions should agree across both deliverables.
This is the listing equivalent of checking units at an engineering interface. Accuracy inside each file does not protect you when two files describe the property differently.
Review the final text for fair-housing risk
Run the fair-housing review after the description reaches its final substantive draft. Reviewing an early version and then adding new language creates a gap in the process.
Look for wording that describes or implies a preferred type of person, household, religion, disability status, or other protected characteristic. Focus the copy on the property. Describe the layout, materials, spaces, and documented amenities without guessing who belongs there.
A software check can flag wording for closer review. It cannot determine every legal issue, account for every local rule, or replace brokerage policy, counsel, and association guidance. The agent remains responsible for the final listing language.
The tour script deserves the same attention. Spoken marketing can carry the same problematic implication as written copy, even when the phrase never appears in the MLS description.
Preserve disclosure and provenance through publication
AI-edited listing photos need a disclosure process designed for the channels where they will appear. California AB 723 and rules in roughly 38 other states make this an operational issue for agents, not a decorative label added at the end.
Burn the applicable per-state disclosure into the staged image. Then preserve a public provenance record connecting that image to its original. This gives reviewers and viewers a way to inspect what changed, even if a portal crops the image or separates it from surrounding remarks.
Before upload, confirm that every staged file has the intended disclosure, the provenance page opens, and the original remains accessible through the approved record. Keep unstaged and staged versions clearly separated so the wrong file does not enter the MLS batch.
NASA’s 1999 loss came from an interface nobody controlled tightly enough. Listing production has smaller stakes, but the same operational lesson applies: define the handoffs, use one verified source set, and inspect the finished package as a package.
The last pre-MLS check should be concrete. Play the narrated tour with sound, read the final description, review the fair-housing result after the last edit, and open the provenance record for every staged image. Only then move the files into the publishing queue.
Comments
No comments yet.