The honest starting point
Land AI's qualification system routes leads with 94 to 96 percent accuracy across its client base. That means roughly 4 to 6 percent of leads are misclassified on the first pass before the QA review catches them.
The QA review catches most of those errors before they reach the client. But some misclassified leads do reach the client's CRM, either because the error was subtle enough to pass QA review or because a seller's circumstances changed between the Maya conversation and the time the operator called them back.
Here is what happens in those cases.
What "bad lead" can mean
A bad lead is not always a misclassification. There are a few distinct situations that operators sometimes group under the same label.
A misrouted lead is a seller who reached Pre-Qualified status but should have been classified as Nurture. The most common version is a seller who described flexible timeline and below-market willingness on the call but whose language was ambiguous enough that the routing logic and the QA review both passed it through.
A changed-circumstances lead is a seller who was genuinely qualified when Maya spoke with them but whose situation shifted by the time the operator called. They found another buyer, their family situation changed, or they decided to hold the property. This is not a system failure. It reflects the reality that seller motivation is a point-in-time condition.
A data quality issue is a lead where the underlying property information was inaccurate: the wrong owner contacted, a property that does not match the county criteria, or contact information that reached someone unrelated to the target parcel. These come from skip-tracing limitations rather than from Maya's qualification process.
How clients flag issues
Every lead in the managed CRM is fully accessible to the client. When an operator calls a Pre-Qualified Lead and finds that the seller's situation does not match the brief, they can flag it directly in the CRM by reclassifying the lead, leaving a note, or messaging the Land AI team through the client's Slack channel.
The Slack channel is the primary channel for issue escalation. Operators message the Land AI team with the lead details, what they found on the call, and what they believe the correct classification should be. The team reviews the transcript, the QA notes, and the operator's feedback and responds with a resolution.
The resolution is typically one of three things: a confirmation that the lead was correctly classified and the situation changed after qualification, an acknowledgment of a misclassification with a correction in the CRM, or a note that the issue reflects a data quality problem that will be addressed in the next data pull for that county.
What happens when it is a genuine misclassification
When Land AI's team confirms a misclassification, the lead is corrected in the CRM and the correction is documented. The operator's pipeline reflects the accurate classification. The misclassified lead does not count toward the operator's Pre-Qualified lead total for that month.
This matters because Land AI tracks pre-qualified lead delivery against the plan threshold. If a client is on Scale Engine (100 leads per month), a confirmed misclassification reduces the delivered Pre-Qualified count by one, and Land AI makes up the difference in subsequent weeks.
The QA team also reviews confirmed misclassifications as training inputs. If a particular pattern of seller language is producing consistent errors, the prompting and routing logic gets updated so the same error does not repeat across other campaigns.
What to do when leads are coming in but not closing
Some operators reach month two or three with Pre-Qualified leads in their CRM and no signed contracts. Before attributing this to lead quality, it is worth looking at the data.
Land AI's average close rate on Pre-Qualified leads is around 5 percent. On 20 Pre-Qualified leads per month, that is one deal per month at the average. If the close rate is lower than that, the most common causes are follow-up timing (calling Pre-Qualified leads more than 48 hours after delivery reduces conversion significantly), offer structure (offers too far below seller expectations create friction that closes the conversation), and market fit (some counties produce Pre-Qualified leads at reasonable volume but the seller price expectations in those counties make deals difficult to make work at acceptable margins).
Land AI's team reviews performance data with clients who are not hitting expected deal flow and works through those variables together. The process is not hands-off after onboarding. Clients who are not closing at expected rates get active support in diagnosing whether the issue is in the leads, the follow-up, the offer structure, or the county selection.