Anyone who uses online channels for lead generation, customer acquisition, or sales knows that attribution is difficult, if not impossible for marketers.  The 2021 privacy updates on IOS and the recent Google updates make this issue even more difficult.  It is imperative that marketers understand what channels drive sales and leads.  Geo-targeted testing can help.

The Problem

A $12M per year eCommerce site could not scale Facebook and Instagram spend for their Direct to Consumer (DTC) website.  It wasn't clear what was driving their sales.  Geo-Targeted or City Pair testing is common in the traditional media world, but it is not often used in the online world.  When thinking about measuring the effectiveness of radio, TV or newspaper ads, Geo-Targeted testing is really the only way this can be done at scale.  Given the challenges with online channel attribution, we used this approach to test online marketing spend.

The Test

For this test, 5 metros were identified and Facebook spend was increased by 5X in those areas.  It went from about 7% of total spend to about 35%.  The test was run for 8 weeks.  Our hypothesis was that the increased spend would drive incremental sales on Shopify, Amazon or both.  Sales were measured based on shipping ZIP code in both Shopify and Amazon.  Spoiler alert: there was no increase in sales.

The results

Shockingly, there was no significant increase in sales as a result of the increased spend.  I'm sure many of you have heard that increasing FB/IG spend drives increased sales on both a DTC site as well as on Amazon.  This is WRONG, it did neither.  

There was a small change from about 10.9% of sales to about 12.2% of sales. However, this 10% increase does not pay for the 500% increase in spend.

Amazon sales in the test areas during this time saw similar results. There was no lift caused by the increased FB spend in the test geography.

Conclusions

It is possible that the spend levels on FB/IG when the test began, about $12K per week, were sufficient to saturate the market. Creative may have been overused in the test. While this is possible, it probably wouldn't be responsible for the complete lack of response. Given the magnitude of the spend some sort of response was expected.

Where to go from here

In the new world of increased privacy on IOS and Google, we all have to live with the fact that FB/IG spend is not indefinitely scaleable.  It is also possible that other emerging channels, like TikTok, can work well for customer acquisition. When in doubt, test it out!

Emily Slatkavitz is a freelance writer at Kettle Hole Partners. In her free time, she enjoys hiking with her dog and trying new recipes.