In 2008, the PEDIGREE Foundation was established by Mars Petcare & PEDIGREE to drive dog adoption. Since then, they have made it their goal to end pet homelessness by 2050, ensuring every dog finds their forever home.

However dog homelessness has continued to grow, with 12 million dogs currently sitting and waiting in shelters around the world.

With a problem so huge, making another ad about adoption wasn’t enough. Instead, we made every ad Adoptable.

Adoptable is a stable diffusion machine learning model, an innovative approach that gives unprecedented control over outputs compared to generic AI models.

Most shelter dogs are mixed breed so don’t conform to a generic shape. Our unique solution creates an adaptable 3D model of the individual dog with grooming information applied to enable representation of the individual dog. This allows Adoptable to be rolled out at scale, requiring minimal effort from shelter with maximum benefit.

The fully augmentable rig means we can pose a dog in any position we desire, right down to micro expressions in the face, enabling unprecedented control and consistency over each output. This allows us to create a photo-real double in any pose or expression to suit any piece of PEDIGREE creative.

Adoptable will be rolled out across every PEDIGREE digital touchpoint by 2026. Turning every digital PEDIGREE ad into an ad for a shelter dog.

Data was collected through PEDIGREE's extensive shelter network, to understand the key reasons why shelter dogs are more likely to be returned. The data told us that mismatch of lifestyle and misunderstanding of the breed were the most significant markers in whether a dog would be re-surrendered.

A deep dive of data was conducted to find characteristics of the adoptable shelter dogs that could then be matched to suitable homes and owners (driven by geographic and 1PD consumer data). Instead of potential adopters needing to seek out the perfect dog for them, we can now introduce them through PEDIGREE's digital advertisements.

Adoptable was launched in New Zealand, partnering with local shelters to showcase real adoptable dogs as the stars of our campaign. Their photo-real outputs were used across adoption, retail, main meal and treat messaging.

Using geographic and 1PD data such as proximity to parks or average household size, the shelter dogs were shown in media placements with proximity to their shelter but also suitability to audience and area.

Audience data was also used to create dynamic copy with specific messaging relating to location, time of day and interests of the consumer. For example, commuters on at the end of the day say digital posters featuring a shelter dog with the headline, “Going home? Can I come?”. Each ad’s QR code took the viewer to a personalised landing page where they could learn more about the dog they just saw, and arrange to meet them.

Our objective is to get shelter dogs adopted. But before they can be adopted, they need to be seen.

Adoptable has already proven that showing shelter dogs in their best light gives them a better chance of adoption.

We’ve seen a significant uplift in the likelihood of adoption for dogs who have featured in an Adoptable asset, versus their original picture taken by a shelter.

Shelter site visits: 6x increase in traffic and profile views for featured dogs

4.5 x longer dwell time on the dog’s profile once on the site

50% of featured dogs adopted in first two weeks

12% more likely to adopt when the adopter comes into the shelter site via an Adoptable asset