Situation

AdventHealth (AH) is a non-profit healthcare system headquartered in Florida that operates facilities within nine states across the United States. Given AdventHealth’s continuous brand growth, their leadership team was focused on increasing their brand awareness in current markets without cannibalizing their efforts. As their media partner, Net Conversion faced the challenge of sustaining brand awareness growth by maximizing and leveraging search efforts to increase leads across service lines within each region. 

Our marketing objective was to drive incrementality in quality leads with the goal of decreasing CPA by -15%. To reach this goal, we decided to adopt AI initiatives within search campaigns, and lean into broad match campaigns – a newly elevated KW match type that would allow us to leverage Google’s advancement in machine learning and increase our efficiency in reaching our marketing objectives and goals.

---

Results

---

Our broad match category cancer campaign drove +112% more leads within our experiment period (5-week period), while our category cardiac campaign drove +34% leads compared to our baseline campaigns. Both campaigns exceeded our marketing objective and goal during this test. Broad match cancer campaigns dropped CPA by -42% with the same budget, while cardiac campaigns saw a -11% decrease in CPA. Within the 5-week experiment period, broad match search campaigns pushed +52% incremental leads compared to baseline campaigns. Exceeding all goals of increasing incremental leads while lowering CPA, our team will continue to implement and test broad match campaigns within our search efforts to ensure continued incremental leads to AH services lines and regions.

+52%

Incremental Leads

-42%

Cost Per Acquisition

Objective

---

Our marketing objective for every location were as follows:

• Drive incrementality in quality leads
• Decrease CPA by -15%

---

Implementation

After conducting thorough research, our team uncovered recent strides in AI, particularly within the realm of large language models (LLMs), which have bolstered the capabilities of Google Ads’ broad match feature. According to Google, broad match has undergone notable improvements in its capacity to delve into search queries at a profound level, discern the intent behind keyword searches, and even factor in the sequence of words within the queries. Additionally, it claims to have refined its ability to direct traffic towards the most pertinent keywords while adeptly recognizing queries in multiple languages. To put this newfound understanding into practice, we devised an experiment within our CFDS – Central Florida category campaigns in our Google Ads account, integrating broad match keywords as a match type for our biggest service lines – Cancer General and Cardiac General. 

Broad match acts as an assurance that our ads are activated by search queries encompassing searches that relate to our keywords such as keyword variations, synonyms, and closely related terms. As we implemented these keywords, our priority was to ensure the preservation of brand safety and overall quality. Consequently, we rigorously cross-referenced campaign control negatives and ad group control negatives, a process anchored by frequent search term reports. This practice enabled us to promptly sift out irrelevant and underperforming search queries, thus upholding our standards.

For this experiment all creative, messaging and campaign settings (aside from keyword match type) remained the same.