The Always-On Approach: How to Continually Improve Your Streaming Advertising with Incrementality
2 likes181 views
The document discusses strategies for improving streaming advertising through incrementality testing, emphasizing its importance for modern marketers in decision-making. It covers topics such as experimental design, case studies on successful implementations, and the benefits of a continuous measurement approach termed 'always-on incrementality.' Key takeaways include fostering a testing culture and enhancing analytics literacy within organizations to optimize marketing efficacy.
The Always-On Approach: How to Continually Improve Your Streaming Advertising with Incrementality
- 1. The Always-On Approach How to Continually Improve Your Streaming Advertising with Incrementality FEATURING:
- 2. Today’s Logistics Kerry Mallett Senior Content Specialist, Webinar Recording & slides will be in your inbox shortly after the session Download relevant resources Join our chat and submit questions for live Q&A Log back anytime with same link Join our poll questions - including if you want to talk to our teams after!
- 3. WE’VE MASTERED THE MEDIA THAT MATTERS MOST Tinuiti is the largest independent performance marketing firm across Streaming TV and the Triopoly of Google, Facebook, Amazon. Employees Media managed Award-winning culture* $3B #1 1000+
- 4. Some of Our Clients Our Recognition For Commerce Leaders Of Today And The Future
- 5. Our Speakers PHILIP RUNDALL Senior Manager of Economics TINA MOFFETT Principal Analyst LAUREN WOLFEN VP of Client Strategy & Analytics WITH GUEST SPEAKER
- 6. Agenda ● Improve Your Streaming With Incrementality Testing ● Introduction To Randomized Controlled Experiments ● Incrementality Case Study ● Traditional And Updated Incrementality Testing ● What Is Incrementality & Always-On Incrementality? ● What This Means For Modern Marketers 6
- 7. What is your comfort level with experimental methods from other channels? ● Very comfortable ● Somewhat comfortable ● Not comfortable POLL
- 8. How much of a priority is building experiments to measure incrementality? ● A big priority ● Somewhat a priority ● Not a priority POLL
- 9. Improve Your Streaming With Incrementality Testing Tina Moffett Principal Analyst B2C Marketing Executives
- 10. 10 © 2022 Forrester. Reproduction Prohibited. Let me introduce two macro trends that are reshaping business. Consumer hyper adoption Digital disruption Consumers who are always willing to try new products or services
- 11. 11 11 © 2017 FORRESTER. REPRODUCTION PROHIBITED. B2C Marketing Executives have competing priorities 25% Improve ROI effectiveness. 24% Improve marketing alignment and collaboration with other departments. 23% focus on post-sale customer engagement.
- 12. 12 © Forrester Research, Inc. All rights reserved. Businesses have a lot to worry about. Data integrity Geopolitical fluctuations Empowered customers Supply chain stability Climate change Economic uncertainty Human health and wellness
- 13. 13 © Forrester Research, Inc. All rights reserved. Massive hurdles prevent data driven marketing decisions Source: Forrester Global Marketing Survey, 2022 (B2C) Base: 638 B2C Marketing decision-makers whose do not rate their marketing org's use of measurement and analytics as optimized Question: Please select the top challenges that prevent your organization from making use of measurement and analytics. 19% Data materializes too fast for us to manage. 21% Inconsistent levels of data quality. 18% Buy in on the value of measurement and analytics. “The biggest challenge our marketing measurement is the speed at which we're analyzing data. The lag time between getting the data and evaluating the results is too long.” --Automotive Marketing Analytics Executive
- 14. 14 14 © 2019 FORRESTER. REPRODUCTION PROHIBITED. Unified measurement blend strategic and tactical marketing measurement to enable customer obsessed strategies
- 15. 15 © Forrester Research, Inc. All rights reserved. Experimental design is the process of carrying out research in an objective and controlled fashion so that precision is maximized, and specific conclusions can be drawn regarding a hypothesis statement. Experimentation helps us decipher cause and effect Source: Science Direct
- 16. 16 © Forrester Research, Inc. All rights reserved. B2C marketing executives use experimentation to understand the incremental impact of marketing Marketers use experimental designs to analyze how a marketing campaign incrementally influences customer behaviors. • Provide simplicity and clarity. • Measures impact of multiple marketing variables, like creative and messaging.
- 17. 17 © Forrester Research, Inc. All rights reserved. Building experiments require careful development of test and control cells Total Populatio n Test Group Control Group Test Group Results Control Group Results Incremental Lift Measurement
- 18. 18 © Forrester Research, Inc. All rights reserved. Source: Forrester Global Marketing Survey, 2022 (B2C) Base: 638 B2C Marketing decision marketers 2021 What advanced marketing measurement approaches does your organization use to understand marketing efficacy?
- 19. 19 © Forrester Research, Inc. All rights reserved. Source: Forrester Global Marketing Survey, 2022 (B2C) Base: 638 B2C Marketing decision marketers 2022 2021 What advanced marketing measurement approaches does your organization use to understand marketing efficacy?
- 20. 20 © Forrester Research, Inc. All rights reserved. Naked Wines used incrementality testing to boost brand awareness and revenue • Summary: Wanted to increase brand awareness in new Australian market. • Approach: Developed testing strategy across key brand specific channels. • Results: Increase in brand awareness 59%, from 40%. ̶ Aggregate market payback increased.
- 21. 21 © Forrester Research, Inc. All rights reserved. Apply core best practices to upstart your incrementality measurement Build basic statistics and data literacy. 1 Use your analytics team to help construct or guide your incrementality tests. 2 Split test cells across homogenous populations, like segment level and ensure audience and geo-location areas consistency for precise incrementality results. 3
- 22. Putting Incrementality into Practice in Streaming
- 23. Experimental Methods in OTT Can Provide More Efficient Decisions 23 ● As omnichannel marketers, we need to understand incrementality throughout the funnel to drive optimal investment decisions. ● OTT can serve as a full-funnel marketing tool, broadening reach and filling the top of the funnel while producing efficient acquisition in its own right. Awareness Acquisition
- 24. There Are A Few Techniques To Understand Incrementality Throughout The Funnel Experimental Methods Quasi-Experimental Methods Observational Methods
- 25. Experimental Methods Are The Gold Standard 25 OTT Incrementality Tests benefit from being experiments. ● Test designed to meet requirements. ● Proper randomization is done. ● Accurately measures incrementality on an OTT publisher. The basic principle of the test involves creating two distinct impression groups: a control group, and an exposed group. TV lift tests can often be run with a conservative budget and a relatively short time frame to produce statistically significant results that equate to a measured lift in response. Measurement To understand the incrementality of TV, we measure and compare the responses of the control group vs. the group that saw the commercials over a distinct time frame and for a particular KPI. Test Structure Implementation
- 26. Test Overview Client industry: Wedding services KPI: Signups Test duration: 3 weeks Property: Hulu Reality Cluster Total test spend: $30k Spend ratio: 80% / 20% (exposed / control) Findings Based on the difference in response rates, the exposed group was ~6x more likely to register for an account than the control group. Control Group Exposed Group N= 159k N= 102 signups N= 36k N= 4 signups 4.0 3.8 x 10-5 t-Stat p-Value CASE STUDY Streaming Audience Holdout Audience Sample Sample Response Response Rate 1.1 x 10-4 6.4 x 10-4
- 27. Experiments Have Their Practical Limitations 27 Truth in the test Truth in real life Generalization Internal Validity External Validity ● Money spent on placebo ads. ● Limited OTT partner capabilities. ● Incrementality changes over time. ● Results to not apply to other publishers. Incrementality Test Drawbacks Problems With External Validity
- 28. Quasi-Experimental Methods Solve Drawbacks And Maintain Validity 28 ● Quasi-experimental methods are used when randomization is not possible. ● Results are still valid even if randomization is not used. ● Commonly used in the social sciences.
- 29. Always-On Incrementality Allows For Costless, Continuous, and Campaign-Wide Insights 29 Incrementality: Measuring a funnel step or action that would not have occurred without a specific interaction, such as seeing an advertisement. Always-On Incrementality: Constantly being able to measure the incrementality of a marketing campaign. Not just a static measurement based on a test that has been run in the past. Incremental Conversions Converted because they saw the OTT impression Background Responses Would have converted regardless of exposure to impression Video Impressions That Converted
- 30. How Always-On Incrementality Works 30 ● Impressions are served on the same publishers by multiple clients. ● Behaviorally match impressions to create control groups. ● Quantify incrementality. Daily Measurement Process
- 31. What Does This Mean For A Marketer? 31 ● With always-on incrementality, we can take the guesswork out of driving a campaign as a marketer. ● No longer need to apply a haircut to the results to get the “real” metrics, now know the full picture based on always-on incrementality.
- 32. Key Takeaways Tinuiti/Bliss Point Media: 1. Quasi-experimental methods in OTT can lead more efficient decisions. 2. Always-on incrementality allows for costless, continuous, and campaign-wide insights. Forrester: 1. Create a testing culture in your organization by sharing success stories and the economic value of marketing experiments. 2. Build a baseline of analytics literacy. Start with the constructs of what makes a good test: statistical significance, population representation, and clear testing parameters.
- 33. Schedule Your Consultation with a Streaming Expert
- 34. Q&A PHILIP RUNDALL Senior Manager of Economics TINA MOFFETT Principal Analyst LAUREN WOLFEN VP of Client Strategy & Analytics WITH GUEST SPEAKER
- 35. 35 Stay informed on the future of digital marketing Visit our content hub ➜
- 36. Thank you!