Renting is stressful: what millions of renter conversations reveal (EliseAI x a16z)
In an X Article published in July 2026, Minna SongDealroom has a profile for this one. Try Dealroom →, CEO and co-founder of EliseAI, shared a demand-side read of the US rental market for a16z's chartpost series, drawing on aggregate data from millions of renter conversations handled by EliseAI's AI leasing agent. EliseAI now powers renter interactions for roughly one in six US rental apartments.
A rising "rent-anxiety index." Classifying the share of opening messages that lead with money (rent, fees, "any specials?"), the index has climbed relentlessly since its low point in August 2024. San Francisco entered the window among the least price-anxious big markets, then added 15 points in fourteen months to overtake New York in late 2025.
A narrative violation on deals. Checked against Zillow's count of listings advertising concessions, renters ask about deals most in the places where the fewest are posted — listings show what landlords advertise, conversations reveal what renters want but cannot find. New York renters carry about 19 points of unrelieved stress; Denver landlords offer 38 points more deal than renters request.
The affordability squeeze bites hardest on cheap stock. Class C (workforce) units draw the longest lines yet sign fastest (10.8 days to a lease versus 12.3 for luxury). Even Class A buildings answered only about half of mystery-shopper inquiries, and reply times grew as rent shrank — the customer does the chasing.
Attendance, not paperwork, is the contest. Once an application starts, about four in five renters finish it; the biggest leaks are getting renters to show up for a tour and to begin the form. Nationally, only 16% of booked tours end in a signed lease (8% in both San Francisco and New York). Tours cram into lunchtime (52% between noon and 3pm) while inquiries roll deep into the evening.
AI adoption at scale. The median renter journey grew from about 9 days and 3 messages in early 2024 to 14-15 days and 7 messages by late 2025; 61% of signing renters now use more than one channel (email, text, webchat, voice), and median AI-handled call length rose from 46 seconds in mid-2023 to about 1 minute 40 seconds — as long as human calls. "The talkers are the signers": conversations of five or more messages are 14% of traffic but 39% of leases.
As a16z GP Alex Immerman put it, "The strongest evidence of AI adoption is what people actually trust AI to help them do. This data shows millions of renters using AI to ask questions, compare options, and navigate the process of finding a home." EliseAI notes it is hiring a Lead Economist to build more reports like this.
Read more: Minna Song on X
