Success Story: Nonprofit Reduces Donor Acquisition Costs 75%+

Advertising & customer acquisition: targeted campaign efforts generate viral buzz on Reddit and increased donations

Success Story: Nonprofit Reduces Donor Acquisition Costs 75%+

Advertising & customer acquisition: targeted campaign efforts generate viral buzz on Reddit and increased donations
In This Article

“Placer is unbelievable. Being able to find the exact demographic we want to target, or learn what their interests are, makes our job so much easier. And we can prove that it works... a 75% drop in acquisition costs and record web traffic tells me our Placer insights were spot on for this campaign!”

~ Hunter Shaffer, Senior Director & Strategic Advisor

The Challenge

A Blood Center in Need of Donors From a Specific Demographic

During the COVID-19 outbreak, the Stanford Blood Center faced a shortage of donors in their area, especially among males aged 25-45, an under-represented blood donor demographic who are able to donate larger amounts of blood. The blood center looked to the industry cooperative, Blood Centers of America (BCA), for help with increasing collections during a historically low time of year. Together, they launched a raffle promotion to target donors from that demographic, offering prizes that included rare NVIDIA computer graphics cards, a Sony PlayStation 5, a Lululemon Mirror, and other tech items. For that campaign, Hunter Shaffer, Senior Director and Strategic Advisor at BCA, and Stanford’s Marketing Manager, Kay Enriquez, along with the SBC Marketing team, needed to:
Identify the best areas to advertise the raffle to that demographic
Analyze the donor acquisition costs to ensure the advertising campaign and raffle were cost-effective.

A seasoned Placer user, Hunter knew that mobile location and people data could help answer these questions.

The Solution

Using Placer to Identify Zip Codes That Over-Index for Males 25-45

Due to its location in Silicon Valley and the numerous tech companies nearby (companies that hire men ages 25-45), Hunter knew that Stanford was surrounded by its target demographic. To find them, he mapped the trade areas of employees for major tech companies nearby, including Facebook, Apple, Uber, and more.

With the employee distribution mapped, Hunter then overlaid their prior donor information onto those trade areas and looked for zip codes that had both an abundance of tech employees and existing blood donors. Following Tobler’s Third Law of Geography, which posits that similar groups cluster around each other, Hunter had the team send ads to the zip codes with the highest concentrations of both the male tech cohort and past blood donors.

Learn more about Audience Profile

The campaign lasted one month, and used digital ads like the one above to solicit increased blood donations from new and returning donors.

The Outcome

SUCCESS: Donor Acquisition Costs Down 75%, Campaign Viral on Reddit

The campaign results far surpassed any expectations Hunter or the Stanford Blood Center had expected, such that it’s become the model that Hunter and Blood Centers of America use for other campaigns. Some highlights include:

Donor acquisition costs decreased by more than 75% due to the large number of donors.
First time visitors increased by 83% compared to the time directly before the campaign.
Male donors increased by 17% compared to the previous year.
Donors aged 20-40 increased by 51% compared to the previous year.

And, perhaps most impressive, the campaign went viral on Reddit and generated record setting web traffic to the Stanford Blood Center website.

Many of the new customers even drive farther than previously, as the zip codes targeted were significantly closer to the competitor’s store. In other words, the campaigns were a resounding success.

Going forward, the store team plans to continue using Placer to target marketing campaigns for their store, and Placer looks forward to working with them as they do.

Case Study

The Challenge

The Outcome

Case Study

The Challenge

The Outcome

Was this helpful?
Yes
 | 
No
Thanks, we'll pass on your feedback to improve our services.
Do you have suggestions for new topics?