The “Voodoo Magic” Way We Tracked $3,282,590.79 In PROFIT For 5 Brick & Mortar Retail Stores in 2018, From Only $344,391.36 In Facebook Ads

“I love wasting money on ineffective advertising!” said no one, ever.

But, if that’s the case – then why are so many retail businesses still spending so much money advertising with traditional media (radio, tv, print), with very few ways of attributing sales back to specific ads?

Riddle me this – can you tell me the ROI from each retail transaction you made last year, and trace that back to a specific TV commercial, or radio ad, or print ad – and do it at scale for each specific transaction?

Here’s an example: take a look at this ad below. What if I told you that I spent $2,345.63 last year showing this ad to targeted audiences on Facebook, and can track $86,377.35 in take-home PROFIT (not revenue) from people who saw this ad, and then walked in and bought a car within 28 days after seeing the ad?

That would be CRAZY, right? I mean… there’s no call to action, and there’s not even an offer! How is that even possible?

Well… take a look below 🙂

Let’s take it one step further… what if I was also able to tell you that I tested 1,127 different ad sets last year on this one client alone, and could tell you exactly which ads resulted in a sale at their dealership, and which ads were more profitable than others? And that we used this strategy to spend less on what wasn’t working, and more on what worked – and made the client $1,217,000.76 in profit off of $108,993.38 in adspend on Facebook?

(We both know what’s coming next…)

Now – how are we able to track sales that bought in a retail store, back to the specific ad itself? We use a free Facebook tool called “Facebook Offline Events.” Here’s a quick primer on how this works:

1.) When we launch ads for a client, we also turn on Offline Event tracking for all ads on that client’s ad account. This tells Facebook to keep track of every person who saw or clicked on this specific ad.

2.) Once the month is over, we download the sales report for this client (usually from their CRM), and upload the spreadsheet of all sales for that store last month, directly into Facebook Offline Events. (This can also be done automatically using the API, and/or using other tools like Zapier or Leadsbridge).

3.) Facebook will use the data in the sales report (first name, last name, email address, phone number, date of birth, gender, zip code) to match up with its list of people who saw or clicked on that ad during the prior month. Since Facebook has A TON of data on us, they can usually match most sales with a Facebook profile of some sort.

4.) Within 30 minutes to an hour, Facebook will start matching which of the people in the sales report also saw or clicked on an ad, and then will start populating data for the specific ads that drove sales, and how much profit the dealer made per ad. Offline Events will only attribute the ad that was the “last touch” – so if someone saw 20 different ads during that time frame, they will only “give credit” to the very last ad that person saw/clicked, and not all ads that were seen/clicked.

With that said, there are some limitations to Offline Events that you must be aware of:

1.) Offline Events will only track paid attribution – meaning it will only track sales to someone you paid to serve an ad to. So, if you run an ad to someone, and that person tagged a friend who eventually bought instead of the person you paid to show the ad to, you will not see these sales within the final report. This also means that if you’re selling an expensive item that needs to be financed, if the person you served the ad to didn’t have enough credit, and the sale was put under their significant other’s name and credit, you also won’t see those sales matched up. As such, Offline Events actually shows us more of the floor, and not the ceiling.

2.) Offline Events are not ideal for tracking true sales attribution because the last touch they show may have not been the actual last touch – the actual last touch may have been on Google Adwords, or perhaps they saw or clicked on the ad after they had already decided to make a purchase decision with that company anyways, so the ad itself didn’t necessarily drive purchase consideration for that person. However, there is a nice tool that helps us get closer to actual sales attribution using another free tool called the “Facebook Attribution Tool.” More on this tool in a future article.

3.) Offline Events only go back 28 days for the last touch – and don’t cover multi-touch. In many buying scenarios where the average sales cycle is longer than a month (like beds, homes, cars, etc), Offline Events can’t track all touches within the entire buying cycle – just the very last touch.

So, with these limitations, why is offline events still so valuable? One word: Trends.

Last year, for just this one client alone that we used offline events for, we ran 1,127 different ads.

That’s a lot of ads!

Without having a way to measure actual retail sales (instead of leads), we would have no idea if we were sending a bunch of leads to the client that never bought, and we wouldn’t know if there was a correlation between certain types of ads, and their ability to generate profit for the client.

So, after a few months of running ads and measuring the offline event data, we started to notice that certain ads would have multiple sales attached to them, while other ads that spent the same amount had no sales whatsoever. This gave us valuable feedback – if we spent an equal amount per ad on a few dozen ads, and certain ads showed up with multiple sales attached to them (aka “outliers”), then there was something about these outliers than made people want to buy – and therefore, we should shift our strategy based on what the data was telling us.

We’d analyze these outliers, and start to notice that there were unique elements of those ads that allowed us to capitalize quickly by “doubling down” on the winning strategy by shifting ad dollars into what was working, and eliminating wasted ad spend over time.

We’d start to see months go from…

April 2018:

To May 2018:

To June 2019:

And From July through December 2018:

As you can see, the profit and ROAS (Return On Ad Spend) improved month over month and developed predictability and consistency even at scale.

Offline Events gave us crucial information that allowed us to take the focus off of cost per lead, and cost per purchase, and allowed us to focus on running ads that drove the most profit – even if it meant selling less units (but making more).

There’s obviously a lot of context and nuance to this process that makes it extremely challenging to fit this entire strategy in just one article – so one key takeaway from this is to do your own research on Facebook Offline events to see if this strategy is right for you.

Originally posted by me at: The “Voodoo Magic” Way We Tracked $3,282,590.79 In PROFIT For 5 Brick & Mortar Retail Stores in 2018, From Only $344,391.36 In Facebook Ads