How Media Agency Good Apple’s AI Tool Is Boosting Client Revenue and Winning New Business

Harvest, an artificial intelligence tool developed by media agency Good Apple, is redefining how the agency measures campaigns, ultimately helping it drive revenue for clients and secure new business.

By integrating data from multiple sources—client sales, out-of-home, TV, and digital ads—Harvest shows how a campaign performs across channels, such as search or social. This helps Good Apple’s clients understand where their ad dollars are most effective, enabling smarter budget decisions and more strategic planning.

“As a media agency, we don’t always think that our clients are going to believe what we say the ROI (return on investment) on a campaign is going to be,” said Mark Sturino, vice president of data and analytics at Good Apple. “What Harvest entails is the idea of answering a media question, how our media is impacting the bottom line, what is the end goal of any given campaign and across all channels. And if we have additional money, where should that get spent?”

Harvest, which sidesteps traditional tracking models by using probabilistic modeling, offers a privacy-safe alternative to tracking, said Sturino.

“It doesn’t tie directly to a tracking system,” he added. “From that standpoint, it’s been a very good sales tool.”

Nearly 25% of Good Apple’s clients pay to use Harvest, although Sturino wouldn’t share pricing details or how many use the tool.

Using AI to identify the best-performing channels

Good Apple’s home decor retailer client used Harvest to refine its paid media strategy, identifying which channels—search, social, or programmatic—were most effective and where to invest to boost sales across its 200+ stores.

In three months, Harvest identified paid media channels to end up driving 29% of sales—more than double the industry average of 12%, per the agency—resulting in an additional $7.6 million in sales, according to Sturino.

“We were able to understand how the different channels were performing, where the marginal ROI is, and where to invest the next dollar, ” he said.

Harvest saw an average return on ad spend of $4.06 for each channel it was used with, notably higher than the typical $3 ROAS that is common in the retail industry, per the 2023 U.S. Digital Ad Spending report by eMarketer.

Harvet also identified opportunities in underutilized channels like Pinterest and TikTok, where interest-based audiences offered higher ROAS. This meant the retailer could strategically allocate its budget to maximize for profit and growth.

“When you’re building a statistical model, it’s easy to build something that accurately predicts—that describes it in mathematical terms—but doesn’t give you the insight into what each channel is actually doing and how it’s contributing,” said Sturino. “The AI piece is helping in that area.”