Wouldn't it be nice to have the FB pixel on Amazon?...
Well, we can't. So we'll do what we can.
Match rates are usually around 75% (ranging from 60-85% in U.S. [15-65% outside of U.S., usually in the lower half]).
If you had 100 actual sales, only 75 might show up through Zontracker.
If you had 10 sales, will 7-8 show up? In theory, yes, sometimes that happens. However, anywhere between 0-10 might show up. This is a sample size problem. Accuracy is more consistent as you have more data.
COUPON CODES/AFFILIATE LINKS
If somebody clicks it or uses the code, you'll get 100% accurate. The big problem is that it's hard to navigate the FB world scaling dozens or hundreds of ads with each one requiring a unique code/link to know exactly which ad is leading to a sale or not. Also many brands have spillover sales and view sales where a discount is never claimed or a link is never clicked... Zontracker can pick up some of these sales and track them.
It's popular for Zontracker users to take the data - in some cases it's better than no data and it's better overall than anything else we can do as Amazon sellers - and calculate a more true RoAS for the effectiveness of our marketing. If you have 75% match rates and accuracy, having $75 in revenue could also mean that you actually had $100 in revenue if assuming unmatched users behaved similarly to matched users. Be careful with this estimate but many tool users do this to help with their numbers (such as monthly review).
If you had equal ad spend across two ads and one had 14 sales and the other had 15 sales, or another example 2 versus 3 sales, that is most likely NOT significantly different based on statistics. If you see 20 vs. 30 sales, or even 1 vs. 5 sales, etc., this is where it gets interesting and where we see many FB advertisers have big shifts in their business.
You can still use the data to make relative comparison which ads are performing best. 100% accuracy or 50% accuracy in this case becomes less relevant, because our assumption is likely valid that all ads may be experiencing the same bias. Some users cancel Zontracker because it's not accurate enough for them. Most advanced FB advertisers have reiterated how that's a huge lost opportunity because this last SUMMARY section emphasizes that relative comparison is one of the best things we can do as marketers. Getting profitable FB ads for Amazon is about turning off the bad ads, shifting budget into remaining ads, and repeating that until you run out of bad ads ultimately being left with the better and best ads.