Objective

With in-store purchases making up a vast majority of KIND’s revenue, historically Media Mix Modeling (MMM) and offline sales studies became the source of truth for measuring the success of media campaigns. Unfortunately, neither of these metrics can be reported in real time and are often not delivered until well after the campaign concludes, which limits the ability to make informed optimizations. However, this presented an opportunity to explore alternative date sources to help maximize revenue for KIND Snacks.

Strategy

Tinuiti partnered with Nielsen Catalina Services (NCS) and deployed their Buyer Driven Bidding (BDB) on Meta. BDB was originally built as an offline purchase event for programmatic display buys. Tinuiti and NCS worked to come up with a new and unique way to leverage the BDB connection on Meta for the first time. KIND Snacks has a long-standing partnership with NCS, conducting frequent Offline Sales Studies across a number of social platforms. Traditionally, these studies run in the background and results take 6 to 8 weeks after the campaign ends, informing recommendations that won’t be useful until the next campaign.

The BDB tool is set up as an offline data set, with regular data refreshes in Meta, helping to attribute offline sales data back to our campaigns on Meta. KIND was positioned to be one of the first advertisers to utilize the new tool on Meta and outside of a programmatic display setting. 

However, shortly before launch, the ability to optimize for an offline event set in Meta was deprecated, halting the original plan dead in its tracks. Tinuiti x NCS x Meta quickly put heads together to devise an alternative solution. Never before had an advertiser utilized the BDB data feed as an optimization lever. Tinuiti designed a multi-cell test, measured by an NCS offline sales study,that tested if weekly optimizations based off of BDB data can drive a greater bottom line ROAS for KIND Snacks, compared to our business as usual strategy of monitoring platform metrics in upper- funnel campaigns. 

On a weekly basis throughout the 10 week campaign, NCS would push over updated data sets to Meta Ads Manager, so the Tinuiti team could understand how each campaign, audience, and creative asset was performing from an offline sales perspective, making optimizations along the way to maximize blended (offline and online) ROAS for KIND Snacks. Mid-campaign, offline-attributed ROAS was looking incredibly strong, affirming that we had strong data signals coming from NCS. As a result, Tinuiti up-leveled the strategy again, by modeling a lookalike audience off of the BDB data feed in an effort to capture more users outside the core targeting personas, that would be most likely to purchase KIND products offline. Early results from this tactic were incredibly encouraging.

Results

This campaign was not only trailblazing — a first-time use-case for a BDB data feed as an optimization lever — it was also incredibly effective. KIND Snacks was able to drive an incremental 6% return on investment over their BAU strategy, using Buyer Driven Bidding data to influence mid-campaign optimizations. This was the first time KIND was able to gain insight into offline sales performance mid-campaign, allowing Tinuiti and KIND to adjust and test audience and messaging strategies to maximize ROAS. Furthermore, this campaign laid a strong foundation for future testing with KIND, and product innovations in partnership with NCS, to further develop the product to be even more effective for Social.

Media