Ditch Last-Click: Attribution Secrets for ROI

The world of marketing is in constant flux, and the ability to accurately measure the impact of your campaigns is more critical than ever. Attribution, the process of identifying which touchpoints in the customer journey are most responsible for a desired outcome, is no longer a “nice to have” but a fundamental requirement for effective marketing. Are you still relying on last-click attribution and missing out on the insights that can transform your ROI?

Attribution in Action: A Campaign Teardown

Let’s dissect a recent campaign we ran for a regional healthcare provider, Piedmont Healthcare. They wanted to increase appointments at their urgent care facilities across metro Atlanta, specifically targeting individuals aged 25-54. This is an incredibly competitive market, with multiple hospital systems vying for the same patients.

The Challenge

Piedmont faced a common problem: they knew they were spending money on marketing, but they weren’t entirely sure which channels were driving the most valuable appointments. They were primarily using last-click attribution, which, as we all know, gives undue credit to the final touchpoint before conversion. This meant they could be overinvesting in channels that appeared effective but were actually just benefiting from the work of other channels earlier in the funnel. It’s like crediting the closer in baseball with the win when the starting pitcher did most of the work.

Our Strategy

Our approach was to implement a multi-touch attribution model, specifically a time-decay model, which gives more credit to touchpoints closer to the conversion but still acknowledges the influence of earlier interactions. We also planned to run incrementality tests to validate our findings and ensure we weren’t over-attributing conversions to specific channels.

Campaign Details

• Budget: $75,000
• Duration: 3 months (January – March 2026)
• Target Audience: Adults aged 25-54 in metro Atlanta
• Channels:
  • Google Ads (Search & Display)
  • Meta Ads (Facebook & Instagram)
  • Connected TV (CTV)

Creative Approach

We developed a series of ad creatives that highlighted the convenience and accessibility of Piedmont’s urgent care facilities. The messaging focused on common urgent care needs, such as flu symptoms, minor injuries, and COVID-19 testing. We also emphasized Piedmont’s commitment to providing high-quality care and short wait times. For example, one of our CTV ads showed a busy parent quickly getting their child treated for a fever, with the tagline “Piedmont Urgent Care: Get back to what matters.”

Targeting

Our targeting strategy was multi-faceted:

• Google Ads: We used keyword targeting to reach people searching for terms like “urgent care near me,” “flu treatment Atlanta,” and “COVID testing Marietta.” We also used audience targeting based on demographics, interests, and in-market segments.
• Meta Ads: We used demographic targeting, interest-based targeting (e.g., people interested in health and wellness), and lookalike audiences based on Piedmont’s existing customer database. We also leveraged Meta’s Advantage+ campaign budget feature to allow the algorithm to optimize budget allocation across ad sets.
• CTV: We used demographic and geographic targeting to reach households in metro Atlanta. We also leveraged contextual targeting to show ads on channels related to health, news, and family programming.

What Worked

Several elements of the campaign performed exceptionally well:

• Google Ads Search: This channel consistently delivered the highest conversion rate and ROAS. The intent-based nature of search made it a highly effective channel for capturing people actively seeking urgent care services.
• Meta Ads Lookalike Audiences: The lookalike audiences generated from Piedmont’s customer database outperformed other Meta Ads targeting options. This suggests that Piedmont’s existing customers are a valuable source of referrals and new business.
• CTV Contextual Targeting: Showing ads on health-related channels on platforms like Hulu and Sling TV proved to be effective in reaching a relevant audience.

What Didn’t Work

Not everything went according to plan:

• Google Ads Display: While display ads generated a large number of impressions, the conversion rate was significantly lower than search. This suggests that display ads were less effective at driving immediate action.
• Broad Meta Ads Interest Targeting: Targeting broad interests like “health and wellness” resulted in a high number of impressions but a relatively low conversion rate. This indicates that the targeting was too broad and not specific enough to the target audience.

Optimization Steps

Based on our initial performance data, we made several optimization adjustments:

• Shifted Budget from Google Ads Display to Search: We reallocated budget from underperforming display campaigns to high-performing search campaigns.
• Refined Meta Ads Targeting: We narrowed our Meta Ads targeting by focusing on more specific interests and behaviors related to urgent care needs. We also created new lookalike audiences based on different segments of Piedmont’s customer database.
• Improved Ad Creatives: We A/B tested different ad creatives to identify the most effective messaging and visuals. We also created new ad creatives that specifically addressed the concerns and needs of the target audience.
• Implemented Conversion Lift Testing on Meta: We used Meta’s Conversion Lift to measure the incremental impact of our Meta Ads campaigns. This helped us determine the true value of our Meta Ads spend and avoid over-attribution.

The Results

After three months, the campaign generated the following results:

Overall, the campaign generated 900 appointments at a cost per conversion of $100 and a ROAS of 3.0x. This represented a significant improvement over Piedmont’s previous marketing efforts, which relied on last-click attribution and lacked incrementality testing. What’s often missed is that attribution isn’t just about knowing what works, but also how much it works.

Attribution Model Comparison

Here’s a comparison of the conversion distribution across channels using last-click attribution versus the time-decay model we implemented:

As you can see, last-click attribution significantly overvalued Google Ads Search while undervaluing Meta Ads and CTV. The time-decay model provided a more balanced and accurate view of the customer journey. This allowed us to make more informed decisions about budget allocation and optimization.

The Power of Incrementality Testing

To validate our attribution findings, we conducted incrementality testing using Google Ads Experiments and Meta’s Conversion Lift. These tests involved dividing our target audience into two groups: a test group that was exposed to our ads and a control group that was not. By comparing the conversion rates of the two groups, we could isolate the incremental impact of our campaigns.

The results of our incrementality tests confirmed that Meta Ads and CTV were indeed driving more conversions than last-click attribution had indicated. In fact, we discovered that Meta Ads was responsible for 25% more conversions than we had initially estimated. This insight led us to further increase our investment in Meta Ads, which resulted in even greater returns.

The Future of Attribution

The future of attribution is all about precision and automation. We’re seeing advancements in AI-powered attribution models that can analyze vast amounts of data to identify the most influential touchpoints in the customer journey. These models can also predict the likelihood of conversion based on a user’s past interactions with a brand. Tools like Singular and Branch are leading the charge here.

However, even with the most advanced technology, it’s important to remember that attribution is not an exact science. There will always be some degree of uncertainty and subjectivity involved. That’s why it’s crucial to combine data-driven insights with human judgment and experience. I had a client last year who was so focused on the data that they completely ignored the qualitative feedback they were receiving from their customers. The result? A campaign that was technically “successful” but ultimately failed to resonate with the target audience.

The IAB (Interactive Advertising Bureau) has been instrumental in developing industry standards for attribution and measurement. A recent IAB report highlighted the importance of adopting a holistic approach to attribution, one that takes into account both online and offline touchpoints. This is particularly relevant for businesses like Piedmont Healthcare, which rely on both digital and traditional marketing channels.

Ultimately, the goal of attribution is to understand the customer journey and optimize marketing efforts accordingly. By embracing multi-touch attribution, incrementality testing, and a data-driven mindset, marketers can unlock the full potential of their campaigns and drive meaningful results. But here’s what nobody tells you: attribution is a never-ending process. The customer journey is constantly evolving, and marketers must continuously adapt their attribution models and strategies to stay ahead of the curve.

Don’t get bogged down in the weeds. Implement incrementality testing to validate your attribution model and ensure you’re not over-attributing conversions to specific channels. This will give you a clearer picture of the true impact of your marketing efforts and allow you to make more informed decisions about budget allocation. For more insights, consider how marketing performance mistakes can impact your attribution efforts.

To truly understand your customer journey, leveraging conversion insights is key. This allows for a more accurate attribution model. Also, it’s vital to ensure that your attribution model aligns with your marketing KPIs to measure what truly matters.

Don’t let perfect be the enemy of good. Start with a simple attribution model and iterate as you gather more data and insights. Even a basic multi-touch attribution model is better than relying on last-click alone. The key is to start measuring and learning, so you can continuously improve your marketing ROI.