AIM boosts application rate by 40% and cuts cost-per-application by 26% with Level Signal V2
THE EPIC WIN
Level Agency built a smarter lead scoring model that taught Google’s algorithms to prioritize the prospects most likely to apply, helping AIM generate 26% more applications at 26% lower cost.
THE CHALLENGE
AIM was generating strong lead volume through Google Ads but identified an opportunity to improve down-funnel efficiency across its 15-campus network.
While lead volume was healthy, application conversion rates varied by intent and quality. AIM sought to optimize toward higher-propensity prospects to increase application yield and improve cost efficiency.
THE ANALYSIS
AIM partnered with Level Agency to examine the full enrollment funnel, from ad click to completed application, and understand where the disconnect between lead volume and application volume was occurring.
Level’s data science team analyzed AIM’s offline conversion data, mapping which lead characteristics and behaviors were most predictive of a prospect actually completing an application. The team studied patterns across campuses, programs, and audience segments to build a comprehensive picture of what a high-value prospect looked like before they ever reached an admissions counselor.
The analysis revealed that Google’s bidding algorithms were optimizing for the wrong signal. Campaigns were being rewarded for generating form fills, but the leads that filled out forms weren’t necessarily the ones who moved through the funnel. AIM didn’t have a lead generation problem. It had a lead quality signal problem.
THE INSIGHT
By building a propensity model that could score leads based on their likelihood to apply, and passing those scores directly into Google Ads as conversion values, the team could fundamentally change what the algorithm was optimizing toward. Instead of chasing the cheapest lead, Google’s Smart Bidding would learn to chase the most valuable one.
This shift, from optimizing for volume at the top of the funnel to optimizing for outcomes at the bottom, became the foundation of AIM’s V2 lead scoring strategy.
THE STRATEGY
AIM and Level deployed a data-driven, value-based bidding approach built on Level’s proprietary lead scoring platform, Level Signal V2:
- Implemented a second-generation predictive propensity model trained on historical enrollment data to score leads by likelihood to apply.
- Integrated propensity scores into Google Ads as conversion values to enable Value-Based Bidding optimization.
- Conducted full-cycle validation tests to measure impact across the application funnel prior to scaling.
- Scaled deployment after validating improved application yield and cost efficiency.
Throughout the process, Level’s accounts team, spanning data science, media strategy, and campaign management, worked as an integrated unit with AIM’s marketing leadership to monitor performance and calibrate the model in real time.
THE RESULTS
After deploying Level Signal V2 with Value-Based Bidding across Google Ads, AIM improved down-funnel performance while strengthening overall marketing efficiency.
The model refined how media investment was allocated, prioritizing higher-intent prospects and driving stronger application outcomes across the full enrollment cycle.
Results validated that AIM could scale a more efficient acquisition strategy without compromising growth:
- 40% increase in lead-to-application conversion rate
- 26% increase in total application volume despite optimized lead targeting
- 26% reduction in cost-per-application, improving network-wide marketing efficiency
This initiative marked a strategic shift from volume-based acquisition to predictive enrollment optimization across AIM’s national footprint.
What AIM had to say
“Level Signal V2 enabled us to align media investment with enrollment intent. We are now prioritizing the prospects most likely to become successful students.”
— Joanna Lawner, Director of Performance Marketing, Aviation Institute of Maintenance
Want results like this value-based bidding higher education enrollment case study?
