An iOS14 Thought Experiment - An alternate take on attribution
Cory Dobbin
Helping Brands Diversify Ad Spend With Profitable Omni-Channel Ads | $500M+ in ad spend managed for hundreds of eCommerce brands. Founder @ OTHERSIDE.
Big. Scary. iOS14. I know, you're sick to death of hearing about, and probably thinking "this is all too much info, just tell me WHAT TO DO ABOUT IT!".
Well, here's a roundabout approach to getting (slightly) more "accurate" data in FB after iOS14 (a step by step guide)
1. Set up a post purchase survey on your website. If you use Shopify, consider using ProveIt Surveys. It's a simple, non-intrusive post purchase survey app, plus you save $30 off your subscription with code CORY30. It really is good, I'm not just saying that.
2. Ask 'Where did you hear about us?' and have your marketing channels listed as options to select:
- Facebook/Instagram Ad
- Social Media/Influencer
- Google Search
- Friend/Family
- Snapchat Ad
- etc
3. Run the survey until you get as many results as you can to get some semblance of statistical significance. This will look different for everyone, but I recommend a minimum of 100 responses.
4. Determine what percentage of total submissions came from Facebook/Instagram Ads. For example, if you run the survey to 832 people, and 643 said they come from FB/IG, then you know 77.3% of people come from FB/IG.
5. Take your total number of orders during the duration of your post-purchase survey, and multiply it by that number, in this example, 0.773. You now have a reasonable estimation of how many orders came from Facebook total, outside of just those who answered the survey.
6. Compare that to the number of purchases Facebook reported during that same time frame. If, for example, the above gave you a number of 500, but Facebook only reported 250, you know that there is a 2x disparity (500/250).
7. Take that disparity number, and create custom conversions in your dashboard. Do this for both Purchases, and Purchase Conversion Value.
8. Example: Purchase Conversion Value * 2.
9. Use these custom metrics to report in your dashboard next to the default metrics.
Voila. You now have a way of reporting (slightly) more "accurate" numbers in your dashboard. Repeat the process periodically.
Thank you for coming to my TedTalk.
Global Marketing Access @ Merck KGaA | Marketing & Communications Expert | Brand Strategist | Digital Media | SEO | Content Marketing | Product Marketing | Masters in Expanded Media @ Hochschule Darmstadt.
4 年Well written article