I managed the delivery of hundreds of causal attitudinal studies at Twitter to assess the impact of paid campaigns on consumer perceptions. These studies utilized a bespoke on-platform intent-to-treat methodology, with the option to incorporate programming and analysis by Nielsen or Kantar. Delivering these studies required a deep understanding of the methodological intricacies involved, excellent survey design best practices, knowledge of SQL for data extraction, and proficiency in cloud-based data transfer techniques.

Twitter’s first-party Conversation Lift campaign measurement was built upon Twitter’s Intent-to-Treat methodological foundation and aimed to capture the leading mid-funnel indicators of favorability and top-of-mind awareness by evaluating the extent to which people exposed to paid campaigns on Twitter created more—or more favorable—content after seeing an ad. From auto brands launching new models to entertainment brands launching new shows and movies, my team and I deployed these studies to help clients understand the extent to which campaigns increased their cultural relevance.

In partnership with Placed Powered by Foursquare, these studies quantified the increase in retail visits resulting from Twitter campaigns by linking ad exposure with in-store foot traffic. Built on Twitter’s emerging measurement API and enriched by the vendor’s synthetic control group methodology, my team and I deployed these studies to evaluate the effectiveness of driving foot traffic to car dealers for in-market automotive and theatrical movie premiere campaigns.

I had the privilege of being among the first at Twitter to deliver Conversion Lift studies as part of the alpha testing team. Upon its eventual G.A. launch, the studies we developed enabled advertisers to measure the incremental purchases, sign-ups, and downloads resulting from their Twitter investments. In my role during the Alpha test, I spearheaded client communication and report delivery and compiled client feedback on the study's methodology and mechanics.

Representing the measurement function in this capacity required a deep dive into methodological intricacies, especially when the Alpha test highlighted that the platform's standard intent-to-treat methodology fell short in rigor. Specifically, we discovered that platform usage levels were imbalanced across cohorts. At the same time, our intent-to-treat design meant that the “exposed” users who may not have seen the ad were weighing down reported performance. While our product team pursued long-term solutions, I managed the delivery of the alpha studies in partnership with our Advanced Analytics team, which manually rebalanced flawed data.

I was directly responsible for the selection, implementation, launch, and ongoing refinement of Twitter’s Incremental Viewership measurement solutions. Originally referred to as “TV Tune-In” measurement in partnership with either Nielsen or SambaTV, our offering eventually grew to serve Television, Streaming, Sports League, and Premium Video On-Demand advertisers. This solution was vital to the Entertainment sales team’s strategy to move Twitter out of its experimental advertising budgets and into a position as a flagship partner proven to drive results.

By the end of my time at Twitter, our most popular incremental viewership offering was built on top of Twitter’s intent-to-treat methodological framework, offering a causal look at how exposure to promoted Tweets delivered incremental viewers to the advertiser and provided performance split by targeting creative and flighting strategies. In addition to these standard offerings, these reports integrated with our A/B testing rollout and even showcased the viewership behavior among those exposed to second-hand earned media generated by people exposed to the paid campaign (sometimes referred to as “paid-earned” exposure).

Powered by Oracle Advertising or Nielsen Catalina Solutions, Twitter's closed-loop sales lift reporting enabled advertisers to measure the Return on Advertising Spend (ROAS) from promoted Tweets. This approach involved matching users exposed to ads on Twitter with real-world purchase data—sourced either from loyalty card transactions or compensated research panels. The vendors then applied advanced modeling techniques to extrapolate the impact observed in the matched data to the overall purchase behavior of all exposed households. They then compared these findings with those from a synthetic control group. Successfully delivering this measurement demanded not only a grasp of the experimental design's intricacies but also a thorough understanding of the creative, targeting, and flighting strategies proven to boost offline sales.

While Acxiom's Clean Room solution was not widely used at Twitter, it served mainly as a workaround. While the company built a measurement API, advertisers who leveraged the partnership could quantify the impact of Twitter exposure on increasing conversions, such as requests for insurance quotes, trial sign-ups to a streaming service, or requests to schedule a test drive at a dealership. As a representative of the U.S. measurement leadership team, I presented and explained the methods and logistics of this solution to entertainment, e-commerce, and automotive clients to drive the adoption of the solution and demonstrate how Twitter was investing in lower funnel measurement.

Oracle Advertising’s Buy Through Rate (BTR) reporting offered an observational understanding of automobile purchases among households exposed to in-market automotive advertising on Twitter by matching exposed users to DMV registrations. Although these reports weren’t causal and did not have a control group, they did offer clients an understanding of which targeting techniques were best aligned with likely purchasers, discover the rates at which the reached audience purchased a car from a competitor, as well as other insights about the reached audience. The offering helped capture advertising dollars for Twitter by strengthening the platform’s position as home to affluent households who purchased “more cars, more often.”

I was part of the team that helped deliver Twitter’s first MTA integration with Neustar to key clients in the automotive and entertainment sectors, for which I was responsible. Although the Musk acquisition cut these initiatives short, I had the opportunity to influence the product roadmap, learn about my clients’ attribution models, and provide feedback on the fidelity of our measurement API.

Throughout my time at Twitter, we provided data to service clients’ Marketing Mix Modeling initiatives. Handling these requests often involved intricate data extraction and transmission processes. Yet, the key to success is client discovery and proactive objection handling. This involved understanding how clients structured their models, grasping the granularity at which publishers were analyzed, and assessing how closely clients' on-platform advertising adhered to best practices for driving sales. These efforts were critical for maintaining and expanding Twitter's share of the advertising budget.

Twitter's Amplify program, a critical strategic initiative, aimed to increase revenue by tapping into digital video advertising budgets, which were incremental to the social media advertising budgets to which Twitter was traditionally aligned. Amplify curated, relevant, high-quality, brand-safe video content—ranging from NFL instant replays to The Oscars' red carpet moments—and offered pre-roll ad space to premier advertisers. Despite its popularity, securing rights to high-quality content posed a significant, ongoing challenge.

In collaboration with SambaTV, I spearheaded the development of a custom audience measurement product for our Amplify content acquisition team, providing them with critical data and insights. These reports detailed the size and composition of the audience reached by each content provider and assessed the extent to which this audience was incremental to traditional live, linear viewership. For instance, we could demonstrate to a major sports league how their live instant replays predominantly attracted light TV viewers, boosting their audience size by 20%. Senior leaders at Twitter used these insights in discussions with C-suite entertainment executives during content acquisition and renewal negotiations.

Twitter established a custom server-to-server integration with Nielsen, facilitating cross-publisher Digital Ad Ratings (DAR) dashboards and cross-channel Total Ad Ratings reporting. This integration empowered advertisers to quantify deduplicated audience measurement metrics accurately, gaining deeper insights into campaign performance across various platforms. My team and I played a pivotal role in helping advertisers configure their campaigns for DAR measurement. We guided optimal flighting strategies to maximize results and addressed any objections. Our hands-on approach ensured advertisers leveraged the full potential of Nielsen's measurement capabilities, driving impactful outcomes for their campaigns.