01
The problem
PropNest was a regional agency sitting on real inventory and real demand, losing both to a slow, search-unfriendly site. Listings took 6-8 seconds to load, search was a flat keyword box with no map, and every enquiry landed in a shared inbox with no routing or attribution.
Their definition of success was concrete: cut time-to-first-response, stop losing leads to faster competitors, and make the catalogue searchable the way buyers actually shop: by location and budget, on a phone.
02
What we changed
We rebuilt the front end on Next.js with server-rendered listing pages, so a buyer (and Google) sees a fully-formed listing in under a second instead of waiting on client-side hydration.
Search moved to a Haversine-distance query over indexed coordinates: 'within 3km of here, 2BHK+, in my budget' resolves server-side in milliseconds. We deliberately cut a planned 3D-tour feature from v1: it wasn't what was losing leads, and shipping the search fix six weeks sooner was worth more than the flourish.
Every enquiry (web form, listing CTA or WhatsApp click) now flows through a single lead pipeline with source attribution and automatic routing to the right agent.
03
The build
Two-week paid discovery to map the catalogue, the lead flow, and the must-win queries. Then eight build sprints with a Friday demo each, shipping search, listings, and the WhatsApp funnel in priority order so value landed early.
A small team: the founder leading engineering, a senior front-end partner on the listing UI, and a designer on the search and enquiry flows.
04
Results
Within 11 months: 400% revenue growth, 4x lead conversion, and $50K+ in monthly sales. Listing load time dropped from 6-8s to under a second, and median time-to-first-response on a lead fell from hours to minutes.
Twelve months on, the WhatsApp funnel is still the single biggest source of qualified enquiries, and the catalogue has scaled past 500 active listings across 12 cities without a performance regression.
05
What we'd do differently
We under-instrumented the lead pipeline at launch and spent the first month inferring attribution from logs. Next time, analytics and event tracking ship in sprint one, not after the first 'where did this lead come from?' conversation.
