When a platform should stop accepting a partner's offers
Let’s explore this page together
A practical standard for deciding when repeated unverified terms and weak correction processes make continued partner listings a risk to readers.
This article reflects the named expert’s practical perspective. See NovAsia’s editorial policy for how material is prepared and reviewed.
A partner can send the wrong area, an old price or an outdated furnishing list without being dishonest. Property information changes, files are revised and people make mistakes. A rule that treats one discrepancy as proof that a company should never be listed again is too crude for me.
The harder question begins after the platform has seen the same kind of material problem more than once. At that point, continuing to publish is no longer just reliance on somebody else's data. It becomes a platform decision about what level of uncertainty is acceptable for the reader.
The consequence matters more than the embarrassment
Not all errors deserve the same response. A minor typo and a missing condition that changes the buyer's payment obligation are both inaccuracies, but they do not create the same risk.
The first thing I want to know is what the reader could reasonably do with the information. Could the error merely create annoyance, or could it cause the buyer to compare two offers on a false basis? Did the partner identify the problem and correct it openly, or did the issue only surface after repeated questioning? Is the source of the corrected value now clear?
This avoids two bad extremes. One is treating every imperfect record as evidence of misconduct. The other is normalising material errors because they can be edited later. The standard should be connected to the buyer's decision, not to how awkward the mistake feels internally.
A correction is weak if the next record can fail in the same way
The most useful question after a serious error is not “has this card been fixed?” It is “what prevents the next card from repeating it?”
Consider a hypothetical feed in which several prices arrive without a compulsory component of the payment. The partner corrects each property after the platform flags it, but the underlying method remains unchanged. The old cards are cleaner, yet tomorrow's import can reproduce the same problem.
At that stage, another polite reminder is not much of a control. The platform may need a different source field, a confirmation step, a temporary publishing hold, or manual review of that specific condition. The right mechanism depends on the case. The response has to address recurrence rather than just appearance.
A vague promise to “be more careful” can be sincere and still be operationally insufficient.
Manual rescue cannot become the permanent partnership model
A platform can compensate for weak incoming data by checking more and more itself. That may be sensible for a limited period. It becomes questionable when every new offer has to be reconstructed from messages, screenshots and follow-up questions before its important terms can be understood.
At that point, the apparent scale of the catalogue may be misleading. Readers see polished records, while the platform carries a large hidden burden just to establish basic facts. The relationship is no longer supplying usable offers; it is supplying material from which the platform must repeatedly rebuild them.
I would rather publish fewer properties with a traceable basis than preserve volume by treating systematic uncertainty as an editing problem. Catalogue size is useful only when the entries remain dependable enough for the purpose in which they are shown.
Stopping should be a documented decision, not a mood
Ending or pausing a partnership also needs discipline. The reason should be specific: repeated material terms that cannot be substantiated, refusal to correct a known discrepancy, an inability to identify the source of key information, or continued delivery of records known to be outdated.
That record should describe the state of the relationship rather than label the partner's character. What failed? What was requested? What changed? What remained unresolved? Why would further publication create an unacceptable risk for the reader?
Sometimes a pause can be lifted if the process genuinely changes and the improvement can be observed. Sometimes the sensible conclusion is to stop accepting the offers. The weak option is carrying on unchanged because ending the relationship is commercially inconvenient.
A property platform cannot guarantee that every outside party will be perfect. It can, however, decide what crosses its own publication threshold. Once the team knows that threshold is repeatedly being missed, continuing to publish without a stronger control is no longer somebody else's mistake alone.