UniLodge achieves 33% lower CPA with SegmentStream AI-Driven Attribution

About the Client:

UniLodge is a prominent provider of student accommodation in Australia and New Zealand. With over 130 properties across major cities like Sydney, Melbourne, and Auckland, UniLodge aims to enhance the university experience by creating vibrant student communities. These communities are designed to help students live comfortably, focus on their studies, and enjoy an active social life.

Goals and Challenges:

UniLodge faced a significant challenge in understanding which marketing channels or campaigns were effectively driving bookings. The reliance on last-click attribution was inadequate due to the complex nature of the user journey in their business. Prospective students often explore multiple accommodation options over several weeks, using various devices and browsers, making it difficult to attribute bookings accurately.

Specific Challenges:

• User Journey Complexity: Prospective students’ decision-making process involves multiple interactions across different devices and browsers. This multi-touch journey made it hard to track and attribute the initial point of contact or influence accurately.
• Ineffective Attribution Models: The last-click attribution model was not providing an accurate picture of which channels were driving bookings. This model failed to capture the influence of upper-funnel activities such as Paid Social, Display, and Performance Max campaigns on Google Ads, which play a crucial role in the decision-making process.
• Marketing Agility: Before implementing SegmentStream, both UniLodge and their media agency teams had to manually compare multiple data sources, each with different attribution models. This process was time-consuming and made it challenging to make informed and weighted decisions on budget allocation.
• Upper-Funnel Activity Evaluation: Evaluating the effectiveness of upper-funnel activities like Facebook and TikTok ads was particularly difficult. These channels are essential for creating brand awareness and engaging potential customers early in their decision-making process, but their impact was not visible in last-click attribution reports.

Goals

• Enhance Attribution Accuracy: Shift to a more accurate, AI-driven attribution model that is able to capture the true incremental value of every marketing channel or campaign even when the conversion doesn’t happen on the same device, visit, or cookie.
• Optimize Budget Allocation: Identify the most effective channels and campaigns to ensure optimal use of marketing budgets, ensuring every dollar spent contributes to filling accommodation slots cost-efficiently.
• Improve Marketing Agility: Streamline the process of data comparison and decision-making, allowing for quicker, data-driven adjustments to marketing strategies.
• Gain visibility on Top-Of-Funnel: Gain clear insights into the performance of upper-funnel activities, allowing for a fair assessment and comparison of all marketing efforts.

Partnership with SegmentStream:

SegmentStream — AI-powered analytics & marketing mix optimisation platform.

Unlike traditional attribution models, SegmentStream uses proprietary machine-learning models to measure the revenue contribution of each marketing channel and campaign, helping marketers gain a complete picture of their marketing effectiveness.

This, in turn, helps to find inefficiencies in the current marketing mix, make effective budget reallocation decisions, and improve overall revenue.

To enable its AI-driven Attribution, SegmentStream combines a customer’s first-party data, such as sales and user behaviour, with third-party advertising data (ad impressions, clicks, costs).

• Website Behavioral Data: GA4 BigQuery Exports
• Ad Cost Data: Google Ads, TikTok, DV360, and Real Estate Ads uploaded through custom Google Sheets
• Conversion Tracking: Main conversion is Purchase. Additional events (Canceled, Tentative, Reserved & Not Canceled) are fetched from the CRM (StarRez) for reporting.

SegmentStream admin panel. Demo data.

As a result, the UniLodge team was able to assess their cross-channel marketing performance in one place and make quick, informed budget allocation decisions. Continuous interaction with the SegmentStream CS team during promotion periods, along with wrap-up sessions post-semester, helped further uncover the potential for revenue improvement.

Example of evaluation: TikTok Ads

UniLodge experimented with TikTok ads, which constituted about 1% of their total spend. Without SegmentStream, they couldn’t evaluate TikTok performance due to lack of conversion attribution. SegmentStream provided insights into TikTok’s contribution to overall conversions, allowing for a fair comparison with other channels.

Example of TikTok evaluation (Last Paid Click vs AI-Driven Attribution)

Business Results:

The partnership with SegmentStream has significantly improved UniLodge’s marketing efficiency and effectiveness. By leveraging AI-driven attribution, UniLodge optimized their ad spend, enhanced visibility on channel performance, and achieved substantial cost savings and revenue growth.

Performance Uplift

• 33% Reduction in Customer Acquisition Cost: Throughout 2022-2023 promotion campaigns, UniLodge decreased CPA by 33% and grew ROAS 2.3x. This trend continued with a further 6% CPA improvement and 34% ROAS increase in 2023-2024.
• Paid Social Performance Visibility: SegmentStream enabled UniLodge to assess TikTok ad performance effectively, integrating it with other paid media activities.
• Improved Agility: Previously, the UniLodge team had to spend many hours navigating through multiple platforms to calculate the true Return on Ad Spend. SegmentStream made it possible to unify all marketing data in an automated way, leaving this tedious work in the past.

Client Testimonial:

“SegmentStream has helped us identify where to spend our next best marketing dollar. SegmentStream AI-Attribution assists us in pinpointing the most effective marketing channels and eliminating underperforming ones. The customer service has also been exceptional, and our regular check-ins have cemented our marketing strategy.”

— Grant Waldeck, Chief Sales and Marketing Officer, UniLodge