IBM Watson Advertising Accelerator ("Accelerator") is a performance-driven dynamic creative solution that uses AI to rapidly and continuously make sense of real time data signals to predict and serve the optimal combination of digital display elements to drive the highest engagement and conversion for a given audience.

Accelerator takes a new approach, beyond just dynamic creative optimization. It takes the manual work and human bias out of dynamic creative set up, and gives a clearer picture of an audience to identify hidden patterns that weren't initially apparent.

Additionally, with upcoming changes to traditional identifiers, Accelerator can function without third-party cookies.

Today, while many companies use AI, <25% use the full scope of AI capabilities. We estimate even less adoption within the advertising industry, which has primarily used AI to automate media buying and selling. We believe AI, given its cookie-less data processing and machine learning capabilities, can fulfill – and beat - the ambition of the cookie without the cookie.

What’s been missing from the last decade is the recognition that technology's job shouldn’t be to automate human action, it should be to help augment human thinking at scale.

So in January 2020, we set out to determine if and how AI could impact the creative process. What better place to test than with the most human part of our industry: creative. The offering we gave brands is called Accelerator.

With ad fraud rampant, finding the right audiences can be a challenge. Even with advances in targeting and automation technologies like dynamic creative optimization (DCO) that enable marketers to more precisely find and reach relevant audiences, often audiences are over-targeted with messages that fall short in driving action.

Additionally, current DCO strategies are time consuming and require in-depth reporting on KPIs, and media strategists are often tasked with generating insights to inform future strategies.

With this in mind, in January 2020 we launched Accelerator, a new solution that innovates and evolves on what was previously available in the marketplace.

Accelerator uses Watson AI to continuously learn and predict the optimal combination of digital display elements to drive the highest engagement and conversion for a given audience. In the process, Accelerator helps advertisers better understand the composition and preferences of their audience to inform future strategy.

January 7, 2020:

Accelerator was launched with an announcement at the 2020 Consumer Electronics Show (CES).

June 15, 2020:

As our country and world grappled with some of the most challenging issues of our time - the rise of COVID-19, the fight against racial injustice, and more - brands needed to pivot how they were reaching and communicating with their consumers. As an industry, we have a collective responsibility to develop and use technology that can help our society live our values.

With that in mind, IBM Watson Advertising announced Call for Creative, an initiative that offered 5 premium brands access to our IBM Watson Advertising Accelerator solution for free as long as they used it for purpose-driven campaigns. Through Call for Creative, IBM Watson Advertising sought to demonstrate how artificial intelligence (AI) can drive positive change when used in a purposeful way.

April 2021:

Recent enhancements to Accelerator include the ability to dynamically adapt video and OTT creative based on real-time consumer engagements and other data signals. This means advertisers are no longer limited to adapting ads based on contextual signals to capitalize on growing streaming audiences. Additionally, Accelerator is now available as a managed service through DSP partners across the ecosystem, making it easier for brands to apply this technology to their existing campaigns to help boost performance.

Accelerator takes a new approach to engaging and converting consumers by going beyond just creative optimization It gives a clearer picture of an audience and identifies hidden patterns that weren't initially apparent.

Unlike current DCO offerings, Accelerator leverages AI to predict the right combination of assets for each unique consumer. Accelerator uses Watson AutoAI to deliver ever-faster in-market data processing, model creation, and training with hyper-parameter optimization. Simply put, Watson AutoAI builds models faster, excels in prediction thanks to auto-feature engineering, and uses cutting edge approaches to optimization and self-learning.

First, brands first share their creative assets which are organized into templates and fed into Accelerator. Then, Accelerator leverages machine learning to analyze the hundreds of cookieless data signals that surround each consumer engagement to assemble user-specific ad variations that deliver on a brand’s KPIs.

Accelerator self-trains, learns, and predicts that best performing creative variation to improve campaign performance over time. It analyzes each campaign from a creative and audience perspective, clusters users into audience groups by creative resonance, and reveals robust insights that brands can use to inform future strategy. These insights can make a brand’s future campaigns not just smarter, but more efficient and effective.

With Accelerator, brands have experienced a minimum of a 25% increase in performance from the beginning to the end of campaigns, proving that AI has a growing and powerful effect on the advertising industry.

Results:

120% average increase in performance for participating 2020 Call for Creative brands (Mastercard, Duke Energy, Travelers Insurance, Kimberly-Clark’s Kotex brand, and the Ad Council)

69% lift in conversion rate (actions on site) for the Ad Council

10% lift in site visits for national hospitality brand seeking to show the variety of their accommodations, after only one week of Accelerator in action

143% lift in conversion rate (clicks to cart) for CPG brand over the course of the campaign.