When it's right to send Google the WRONG data

In today’s Paid Search world we’ve had a lot of levers taken away and an increased reliance on first-party data, ensuring we are optimising to the business’s goals, as a main lever for driving greater performance.

This has led us to give Google as much information as possible but when is too much information for Google a bad thing for performance? This talk goes through a case study where we at WeDiscover decided to test suppressing information, and intentionally give Google inaccurate conversion data, to drive better business performance.

Transcript

1. The best talk title in the world! Name Surname COMPANY NAME When it’s right to send Google the WRONG data Inderpaul Rai WEDISCOVER Speakerdeck.com/inders @inderpaul-rai @inderpaul_rai
2. Change is the only constant
3. Game changing releases 2015 Customer Match
4. Game changing releases 2015 Customer Match 2016 Right sidebar ads gone STAs deprecated ETAs launch
5. Game changing releases 2015 Customer Match 2017 Exact Match brutalised 2016 Right sidebar ads gone STAs deprecated ETAs launch
6. Game changing releases 2015 Customer Match 2017 Exact Match brutalised 2016 Right sidebar ads gone STAs deprecated ETAs launch 2021 PMax launch BMM deprecated
7. Game changing releases 2015 Customer Match 2017 Exact Match brutalised 2022 ETAs deprecated 2016 Right sidebar ads gone STAs deprecated ETAs launch 2021 PMax launch BMM deprecated
8. Control taken away With their iterative updates to Google Ads 2020 Search Terms Restricted 2017 Exact Match Killed 2021 BMM Killed
9. Increasingly relying on AI PMax Launched RSAs Launched Keywordles s Campaigns ? 2018 2021 2025 With new features
10. And how does this make us feel?
12. Our jobs have changed
13. We’ve been told that Data is the New Oil for years
14. And ever since we’ve been obsessed with sending the right data
15. Because increasingly that’s all that’s left for us
16. “But I’m doing everything Google has told us to do and not geing results”
17. Our jobs have changed PPC Teachers PPC Doctors PPC Pilots Unlevel the Playing Field: The Biggest Mindshift in PPC History - Fred Vallaeys
18. Client Problem
19. Focussed on growth
20. Suddenly wanted eiciency
21. ROAS CPC CPA
22. Without cuing volume
23. Option 1: Bid strategies
24. Option 2: Ad Copy
25. Option 3: Keyword Mix
26. We did something drastic
27. We edited the values of 14% of our revenue data
29. The Great Hanoi Rat Massacre of 1902
30. The Four Pests Campaign
31. These are known as Perverse Incentives
32. Let’s take a step back
33. CPC and CPA inflation
34. Unusually high value bookings led to volatility
35. And this was a recurring paern
36. But why?
37. The client would apply premiums
38. Misleading Google’s bid algorithm
39. Google reacted by bidding higher
40. RAT PROBLEM SPARROW PROBLEM OUR PROBLEM PERVERSE INCENTIVES
41. Why not add bid caps?
42. Capping CPCs would lose us incremental revenue
44. We wanted to test limiting Google’s sensitivity
45. Without killing these bookings from occurring
46. And using OCI we could artificially limit the value
47. So we capped the value distribution
48. Capping 95th Percentile 9% of conversions would be impacted
49. Editing the values of 14% of our revenue data
50. The results?
51. ROAS CPC CPA
52. ROAS CPC CPA -9% -13% +26%
53. https://we-discover.com/case-study/outlier-value-capping-adjusting-google-ads-conversion-value/
54. This strategy may not work for your business
55. Always hypothesise why
56. Try to identify potential perverse incentives
57. What you choose to tell and don’t tell Google is fundamental
58. Advertiser Google 1PD COGS Budget CRO Feeds PMax Broad Bidding DDA RSAs
59. https://we-discover.com/blog/monitoring-your-cpc-distribution/
60. The best talk title in the world! Name Surname COMPANY NAME @inderpaul_rai @inderpaul-rai Speakerdeck.com/inders