INNOVATION THAT MAKES SENSE

Voice technology is on the rise and we only expect it to keep growing. Right now, one in four searches is conducted by talking, and comScore predicts that figure will reach 50 percent by 2020. Currently, Amazon has the largest share of voice-enabled assistants and as of May 2017, 10.7 million US Amazon customers owned an Amazon Echo.

When Amazon announced the launch of the new Echo Show device in June 2017, we saw an opportunity to leverage Allstate’s existing top-performing expert content while expanding their voice technology offering. The Echo Show device allowed us to augment voice content with helpful videos, reinforcing voice-first instructions with step-by-step visual direction.

Working with the Allstate Blog team, we identified the most popular “hack” content to bring to life first. Transforming proven, useful content helped lay the foundation for a skill that’s poised to garner high engagement and returning users. What followed was a fast-paced, four-month timeline to get a complete skill live and ready to help people better tackle everyday tasks in their homes.

ITERATIVE PROTOTYPING

Our creative team and development team collaborated to quickly bring a working prototype to life. Within a few weeks, our clients were able to interact with the prototype on their own Echo Show device and work with us directly to shape the development of the skill.

To keep things moving quickly, we filmed video content for 12 different hacks and developed the skill concurrently. Both the T3 and client teams were able to interact with iterative versions of the skill with placeholder content while the videos were being edited. As soon as the videos were complete, we had a fully functional skill.

LAUNCH PLAN

While voice technology is growing, driving initiations through paid media is not as seamless as driving app downloads. There is not currently a one-to-one ad unit that drives straight to skill initiations. In order to track users through ad exposure, we created a landing page to drive people to that would not only allow us to better track user interest, but would also allow us to educate people about the skill.

We identified target audiences using look-alike data modeling. Our social media targeted users with an affinity for Amazon Prime and smart homes as well as those interested in technology, DIY projects and hacks. From there, we refined the audiences further based on lifestyle. We then created unique, compelling content targeted to each of our audience segments to drive interest in the skill.