KE Holdings
a dedicated human role, assisted by AI, inserted between the inbound lead and the sales agent
From May to July 2026, the AI-assisted client manager role at KE Holdings (Beike, Lianjia) handled more than 50,000 real estate leads with a 7.4% lead to showing conversion rate, versus the 5% market average cited by CEO Stanley Peng.
Key points
- KE Holdings created an AI-assisted client manager role, placed between the inbound lead and the real estate agent.
- The AI organizes the data, a human assesses the customer's stage and needs, the agent receives a fully profiled customer.
- From May to July 2026, more than 50,000 leads handled and 7.4% lead to showing conversion, versus 5% for the market.
- Evidence level A: figures stated by CEO Stanley Peng on the Q2 2026 earnings call, on August 21, 2026.
Objective
Convert more inbound leads into property showings, by no longer losing track of the prospect at the point of handoff to an agent, and by assigning qualification to a role paid outside transaction commission.
The deployment
KE Holdings operates Beike and Lianjia, the largest real estate transaction platform in China, with 60,274 stores as of June 30, 2026. On the second quarter 2026 earnings call, CEO Stanley Peng described a new role in the lead handling chain: the client manager. It answers a follow-up gap he names explicitly, the platform losing information as soon as a lead reaches an agent. The division of labor is described in these terms: the AI organizes the data, a human assesses the customer's journey stage and needs, and the agent receives a fully profiled customer. The structuring point is not technical but about pay: the client manager is not paid on transactions, which Peng repeats twice in the same passage, and that is what is presented as the guarantee of objectivity when the role routes or sets aside a case. Over the May to July 2026 window, the role handled more than 50,000 leads with a 7.4% lead to showing conversion rate, versus 5% for the market according to Peng. Two caveats matter when reading that figure. The measurement window runs past the calendar quarter, which ends in late June. And KE Holdings publishes neither the definition of its market average, nor the composition of the leads assigned to the setup, nor the share of its total flow that these 50,000 leads represent. The same quarter shows net income up 100.8% year over year and a non-GAAP net margin at a three-year high: those results are the company's, attributed by management to cost optimizations and a favorable business mix, and are not presented as an effect of the client manager.
Results Proof A
Figures stated by the Chairman and CEO of KE Holdings during the Q2 2026 earnings call of August 21, 2026, as part of financial reporting, and verifiable in verbatim transcripts from two publishers. The volume, the rate and the measurement window are tied explicitly to the client manager role, not to the group's overall results.
How it works
Inferred typical approachThe internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.
The stack in detail
- plateforme Plateforme Beike / Lianjia KE Holdings' proprietary real estate transaction platform. It is what collects the lead and what, until now, lost visibility on the prospect once that prospect was handed to an agent.
How it runs, concretely
For ops teams-
1Lead arrives customer
A buyer or a seller makes contact through the Beike platform or the Lianjia network.
-
2Structuring of the prospect data AI
The AI layer organizes the available data on the prospect and the prospect's project.
-
3Human qualification client manager
The client manager assesses the customer's journey stage and needs. This role is not paid on transactions, which leaves it free to set a case aside or to wait.
-
4Handoff to the agent real estate agent
The real estate agent receives an already profiled customer and no longer has to redo the framing work.
-
5Conversion measurement data team
The platform tracks what becomes of the lead up to the property showing and compares its rate to the market average.
The move from lead to property showing. That is the published metric (7.4% versus 5% for the market). It only exists if the prospect is still tracked after the handoff to the agent: that is precisely the gap Peng describes before the role was introduced.
How your customers perceive this type of use
Sourced studiesLe pricing algorithmique est le terrain le plus inflammable : 68% des consommateurs disent se sentir leses quand les marques utilisent le pricing dynamique et 80% jugent plus dignes de confiance les marques aux prix constants (Gartner, 2024). L'equite percue varie selon le secteur : le pricing dynamique n'est juge juste que par 33% a 40% des repondants selon qu'il s'agit de concerts ou de cinemas (YouGov, 17 marches). Le prix personnalise par les donnees individuelles est le plus rejete : 47% des Americains s'y opposent fermement (Consumer Reports, 2024).
Acceptance conditions
- La constance des prix comme signal de confiance : 80% jugent plus fiables les marques aux prix stables (Gartner 2024)
- Le secteur conditionne l'equite percue : le pricing dynamique est mieux tolere pour les cinemas (40% le jugent juste) que pour les concerts (33%) (YouGov 2024)
Red lines
- Le pricing dynamique percu comme abus : 68% se sentent leses (Gartner 2024)
- Le prix individualise a partir des donnees personnelles : 47% d'opposition ferme (Consumer Reports 2024)
- Les frais caches et hausses imprevues, vecus par 79% des consommateurs sur un an et associes a la perte de confiance (Gartner 2024)
Sources: Gartner 2024 · YouGov 2024 · Consumer Reports 2024
How to replicate
Inference, not sourcedData prerequisites
- History of inbound leads with their outcome all the way to the showing, and not only up to assignment to a salesperson
- Usable prospect project data tied to a stable identifier across the whole journey
- A market or internal conversion benchmark to compare the setup against, with its definition written down before measurement starts
Org prerequisites
- Accepting a role inserted between the lead and the salesperson, and therefore sharing a contact that the sales teams see as theirs
- A pay model for the qualification role that is decoupled from transaction volume
- A written rule for what counts as a profiled case that can be handed over
- A right for the platform to keep tracking the lead after the handoff to the salesperson
Possible stack
- Automatic structuring layer for prospect data, plugged into the CRM
- Qualification workstation giving the human role a consolidated view of the prospect
- Conversion measurement by lead cohort, from first contact to the chosen mid-journey event
The plan, step by step
- Step 1Reconstruct the inbound lead journey up to the event that precedes the sale, and identify where tracking breaks down.Deliverable: A quantified baseline conversion rate, and the point where tracking breaks.
- Step 2Define the qualification role: what it decides, what it hands over, and what it is paid on.Deliverable: Job description and a pay rule decoupled from closings.
- Step 3Automate the structuring of prospect data upstream of that role, without taking the decision away from it.Deliverable: Consolidated prospect file, presented to the qualifier on open.
- Step 4Open the setup on a subset of leads and compare its conversion to the baseline, at comparable lead composition.Deliverable: Cohort-level reading, with lead composition documented from the start.
- Step 5Extend to other lead sources while keeping the conversion measurement and the handoff rule to the salesperson.Deliverable: Extended setup and conversion tracking sustained over time.
First step: Measure the current conversion of inbound leads up to the mid-journey event that matters for you, and check whether that tracking survives the handoff of the lead to a salesperson. If the information stops there, that is the gap to close before any AI.
The same pattern, proven in other industries
Choice Hotels
AI response to group requests for proposals
Xometry
AI-optimized B2B marketplace (quoting and recommendations)
Sources
- S1 KE Holdings (BEKE) Q2 2026 Earnings Call Transcript Primary archive pending
- S2 KE Holdings Inc. Announces Second Quarter 2026 Unaudited Financial Results (Form 6-K, Exhibit 99.1) Primary archive pending
- S3 KE Holdings (BEKE) Q2 2026 Earnings Call Transcript - The Globe and Mail Established press archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.