PPC Testing in Low Volume Accounts By Amalia Fowler

46 %
54 %
Information about PPC Testing in Low Volume Accounts By Amalia Fowler

Published on October 25, 2018

Author: SearchMarketingExpo

Source: slideshare.net

1. #SMX #21B @amaliaefowler Amalia Fowler – Account Director, Snaptech Marketing PPC Testing in Low Volume Accounts

2. #SMX #21B @amaliaefowler Amalia Fowler Amalia is an Account Director at Snaptech Marketing Moved to Vancouver, BC six years ago for a “summer” Enjoys coffee, is stoked that Canadians can now apply for Survivor, and wishes she owned a dog.

3. #SMX #21B @amaliaefowler

4. #SMX #21B @amaliaefowler How do you handle testing in low volume accounts? Be aware of your accounts’ specific situation when determining what and when to test Consider alternative testing methods (the how) Know the risks of testing in low volume accounts Communicate – with stakeholders and each other

5. #SMX #21B @amaliaefowler

6. #SMX #21B @amaliaefowler But… do I really need to reach significance?

7. #SMX #21B @amaliaefowler Share these #SMXInsights on your social channels! #SMXInsights There is a big difference between scientific decisions and business ones

8. #SMX #21B @amaliaefowler Preparation is key.

9. #SMX #21B @amaliaefowler Bid Strategies Ad Copy Ad Schedules Display Audiences Bid Adjustments

10. #SMX #21B @amaliaefowler What If? List

11. #SMX #21B @amaliaefowler How to test Get results & take action Wait for the data Set up the test Decide on the metrics Choose test type Create hypothesis

12. #SMX #21B @amaliaefowler Hypothesis We believe that restricting the ad schedule for XYZ campaign will increase conversion rates.

13. #SMX #21B @amaliaefowler Choose your testing type Single adgroup • High volume • Brand campaigns • Important campaigns Multi-adgroup • Excellent for lower volume accounts • Allows you to aggregate the data to find patterns

14. #SMX #21B @amaliaefowler Metrics Confidence Interval • How confident do you want to be with your results? So many metrics to choose from • CTR, Conversions, Conv. Rate, CPI, CPA, • Conversion per Impression (Conversions / Impressions) • Define your minimum necessary data

15. #SMX #21B @amaliaefowler Wait…

16. #SMX #21B @amaliaefowler Analyze Do the math • Excel pivot table to aggregate data Third party platforms • IE. Adalysis Statistical Significance Calculators • IE. Cardinal Path Google Ads Drafts & Experiments

17. #SMX #21B @amaliaefowler Low Volume Considerations - How Type of Test • Multi-adgroup v.s. single adgroup • Use third-party platforms or pivot tables to measure results Metrics that determine success • Lowering confidence level Waiting for the data • Don’t give up!

18. #SMX #21B @amaliaefowler Share these #SMXInsights on your social channels! #SMXInsights When applicable, use multi-adgroup testing to mitigate the fact that you have lower volume.

19. #SMX #21B @amaliaefowler Low Volume Considerations - When Don’t test over a period when you are expecting the impacts of seasonality

20. #SMX #21B @amaliaefowler Questions to ask yourself before testing in low volume accounts • How long will this take me? Is this the right time? • What metrics will I use to determine whether or not this is a success? • Is there a way I can make this a multi-adgroup or aggregate test? • How confident do I want to be in my results? • What tools do I have at my disposal? • Is this account worth the risk?

21. #SMX #21B @amaliaefowler Assess Risk & Address Assumptions

22. #SMX #21B @amaliaefowler Assessing Risk Know your surroundings Understand context Use the 10% rule Beware of over-optimization

23. #SMX #21B @amaliaefowler Addressing Assumptions Don’t bring your gut to a data fight

24. #SMX #21B @amaliaefowler Communicate

25. #SMX #21B @amaliaefowler Communication Considerations • Set proper expectations • Ensure there is active communication with new team members • Keep a record of what testing has been done • Don’t wait until something is broken to start

26. #SMX #21B @amaliaefowler Share these #SMXInsights on your social channels! #SMXInsights § There is a big difference between scientific decisions and business ones § When applicable, use multi-ad group testing § Communicate with team members to avoid overlap – time is precious


#smx presentations

Add a comment