October 29, 2024
|

Yahoo Blueprint Performance, the optimal ghost in the machine

AI might seem like a spooky ghost in the machine, but it’s more like a friendly phantom optimizing your campaigns! Discover how Yahoo Blueprint Performance drives results across the entire funnel and explore one of its core features, Customer Value Optimization.

Kavisha Doshi

Senior Product Manager, Yahoo

In June of this year, we launched Yahoo Blueprint Performance, an enhanced iteration of our AI engine, Yahoo Blueprint. Since then, it has delivered significant improvements for our advertisers across all verticals, driving results with efficiency throughout the entire funnel, from top-of-funnel awareness to bottom-of-funnel conversions.

To summarize some of the performance we’ve been seeing:

  • An entertainment brand whose target audience is anyone who watches TV, which is almost everyone, wanted to achieve massive reach. After they activated Yahoo Blueprint Performance (YBP) they saw 56% more impressions resulting from a 44% lower eCPM when compared to historical campaigns that did not have YBP turned on.¹
  • CPG clients generally seek scaled awareness but are increasingly focused on driving more down-funnel engagement. A leading CPG client saw a 34% increase in clicks at a lower cost per click (CPC) after activating YBP.²
  • For insurance advertisers, a key indicator of prospect interest is when they take action and begin the quote process on their website. After activating YBP, a leading insurance advertiser achieved a 35% increase in quote starts, along with a 27% lower cost per acquisition (CPA) for those starts.³

Demystifying the optimal ghost in the machine
AI has become omnipresent, helping unlock numerous use cases across industries. 66% of advertisers believe AI will be extremely or very valuable for optimization to drive incremental performance.⁴ We agree! However, AI can sometimes feel like a spooky mystery—a 'ghost in the machine,' though an optimal one. As we approach All Hallows' Eve, we thought it would be fitting to demystify one way Yahoo Blueprint Performance enables incremental performance.

Leveraging AI to address disappearing IDs
As we covered before, although Google is no longer fully deprecating third-party cookies, increasing platform privacy controls and government regulations mean we are still moving toward an identity-constrained world. This means advertisers still need tools to effectively reach their desired audiences. One increasingly impactful strategy is for advertisers to take greater control of one of their most valuable assets - their first-party data - and maximize its application across various use cases. In fact, 80% of advertisers report leveraging first-party data to prepare for a cookieless future.⁵

Advertisers go to great lengths to maintain and optimize their first-party data, often using advanced scoring systems to sort customers based on various attributes such as lifetime value, number of accounts, historical sales, account value, and more. Yahoo Blueprint Performance unlocks the potential of first-party data through Customer Value Optimization (CVO). CVO allows advertisers to upload their first-party data along with their proprietary user scores. The Yahoo DSP then leverages AI capabilities across that dataset to find and optimize toward new, high value customers across the entire programmatic ecosystem.

CVO works at the user level by ingesting scores of each active customer. It takes first party scoring data as a model seed. Yahoo then uses the power of AI to model against the entire footprint of the Yahoo ConnectID universe. This then allows us to predict customer scores across fully addressable media. From there, Yahoo Blueprint optimizes to higher predicted scores. We don't stop there. We also provide our modeled predictions back to the advertiser for them to grade themselves.

Customer Value Optimization (CVO) casting its spell
And it’s working! An insurance company developed a sophisticated scoring system to categorize individuals seeking insurance quotes based on their expected lifetime value (LTV). The scoring system uses current customers behaviors (tenure, on-time payments, claims history) to determine who is a high-valued customer.

Then, the CVO model we built ties that to quotes. So, we're able to drive higher predicted LTV at time of quote to get the best quality users before they're even a customer.

After implementing Customer Value Optimization, the insurance company saw an 80% average lift in performance for display campaigns compared to using a standard CPA goal type, and a 30% improvement in CPA for video campaigns.⁶

Want to learn how you can get started? Reach out to your Yahoo DSP account team today to see what Yahoo Blueprint Performance can do for you!

Sources
¹⁻³Yahoo DSP internal data, Q2 2024
⁴Yahoo DSP AI & Optimization Omnibus, June 2024
⁵Funnel and Ravn Research, “Agency Acceleration,” Sep, 2024
⁶Yahoo DSP internal data, 2023

About Kavisha Doshi
Kavisha Doshi is a senior product manager at Yahoo DSP with expertise in machine learning, optimization models and advertising technology

Ready to Start?
Hey, So Are We.

Our sales team can help you truly connect to audiences with unified solutions scaled for growth.

Get in Touch