Top Real Estate Platform Finds Higher-Value Customers at Lower Cost
By the numbers
Overview
Company
Industry
Campaign Type
Snapshot
A leading real estate platform had built a reliable strategy, running Google Search campaigns optimized for users who communicated urgency to sell and often closed at significantly higher rates.
However, self-reported intent only told part of the story. Urgency didn't always mean action, and hesitation didn't always mean no. The team was over-investing in some users and overlooking others, with no way to tell which was which.
With Voyantis, the team shifted to value-based bidding based on predicted conversion likelihood, growing conversion volume while improving both acquisition costs and deal values.
Impact at a glance
Nearly doubled conversion rate
from initial engagement to closed deal
Reduced CAC by over 35%
while increasing average deal value nearly 10%
Delivered 74% ROAS uplift
across non-brand campaigns
Expanded addressable audience
by surfacing high-intent users that rules-based targeting missed
Scaled to 100% of Google non-brand search
with expansion to additional campaign types underway
The challenge
Homeowners engage with the company for all kinds of reasons. Some are ready to sell, while others are just curious about their home's value, casually exploring options, or testing the market. At first interaction, motivated sellers often look identical to casual browsers.
That ambiguity is expensive. The company invests real resources in every potential customer. When someone decides to wait or was never serious in the first place, that investment doesn't pay off.
Value-based bidding could help them focus on users most likely to convert, but there was a timing mismatch. Google optimizes best with fast feedback, but conversions happen months later - long after the algorithm has made its bidding decisions. The team needed a way to predict those outcomes early enough for Google to act on them.
The solve
The team integrated first-party data with Voyantis's predictive growth infrastructure to identify high-intent users earlier and translate those predictions into signals Google could optimize against.
Build predictive models for conversion intent
Voyantis worked with the growth team to identify the right prediction target: likelihood to convert within weeks of first interaction. This became the model's north star—representing actual value rather than early-funnel proxies like self-reported intent.
The platform analyzed behavioral and contextual signals - engagement patterns, early-funnel actions, user characteristics - to predict which users were ready to move forward. Predictions generate within an hour of first interaction and feed directly to Google, allowing the algorithm to learn from future conversion intent while its learning window is still open.
Surface audiences hiding in plain sight
The high-intent filter converted at significantly higher rates, but the team had no visibility into what it might be missing. Voyantis analyzed historical conversion data and surfaced a pattern rules-based targeting couldn't see: a meaningful share of users who converted had initially appeared lower-intent based on early responses. The platform now surfaces both high-intent users and the lower-intent users most likely to convert.
Engineer predictive signals for ad platform optimization
Google optimizes based on deterministic events, like clicks, form fills, conversions. It doesn't natively understand probabilistic predictions of future value.
Voyantis translates conversion-intent predictions into the signal format, timing, and value encoding Google needs. Instead of binary events, Google receives predicted values for each user—bidding aggressively for high-intent users, conservatively for unlikely converters, and strategically for lower-intent users likely to close.
Adapt to seasonal dynamics
User behavior shifts throughout the year. What predicts conversion in spring looks different in winter. Voyantis's always-on infrastructure keeps predictions tuned to current market dynamics - no manual recalibration required.
The platform also refines predictions as users engage. A day-one prediction based on initial interaction becomes a day-three prediction informed by engagement patterns, then a week-one prediction validated by deeper funnel progression. Each update sends more precise signals to Google while the learning window is still open.
Voyantis’ impact
Nearly 2x
conversion rate from first interaction to closed deal
Over 70%
more conversions
Nearly 10%
higher average deal value
Over 35%
reduction in CAC
Proven results across the funnel
The team tested value-based bidding campaigns powered by Voyantis signals against their existing tCPA campaigns. With Voyantis, the company was not only acquiring more users, but also better ones.
Based on sustained results, the team has scaled Voyantis signals to 100% of non-brand search and is expanding to additional campaign types. With Voyantis operating as an extension of the growth team, each new channel builds on a foundation that's already proven.
