Media Math for Advertising: How to calculate ROI for a digital ad campaign
Updated: Jul 24, 2019
A client said to me recently, “I’m going to spend about $3,000 on some ads and see what happens.”
I said, “How about I tell you what’s going to happen and you can save the $3,000?”
Digital media is so measurable, and it follows knowable patterns so closely, that we can use what I call “media math” to predict the performance of a digital marketing campaign with pretty high precision. That means that old style of advertising experiments- “run some ads and see what happens”- are not necessary anymore. But new types of advertising experiments are.
Digital media campaigns- media math
Let’s look at the media math for a digital marketing campaign using digital display ads, those ads you see on websites and apps. Digital display ads cost, on average, $2.80 per thousand impressions and have an average click through rate of .35% *.
That means that buying 100,000 advertising impressions will cost $280 and deliver 350 clicks for an average cost-per-click of $.80. In other words, this campaign will cost you $.80 to get one visit to your website. If you’re selling things on your website, you then need to apply the conversion percentage and your average sale price- the rate at which website visitors buy and how much they spend on average to calculate your return on investment (ROI).
So, here’s the return-on-investment calculation using conversion and sale rates that I made up for the purpose of illustration:
Media campaign cost: $280
Website visitors: 350
Conversion rate: 5%
Number of sales: 18
Average sale: $25
Total sales: $450
Beautiful! In this scenario, you spent $280 and you made $450.
Making predictions or running tests
The point of this example is to show that these outcomes are knowable in advance. The media costs are known averages, and you can look at historical data to see your conversion rates and average sales. Similar types of cost and performance averages are available for other digital media ranging from Facebook to Google AdWords. That’s why calculations like these are easy to do before any media money is spent. So, the old-style advertising tests aren’t necessary anymore, but new types are.
If this is my e-commerce business and I see the results above, I immediately want to see if I can deliver better results. And that’s when I start to run some media spending tests using these variables:
Location: What if I target my ads to other markets?
Audience: What if I tweak the targeting profile?
Messaging: What if I change the ad creative or offers?
Medium: Maybe it’s time to try a different digital ad platform- Facebook? Pinterest?
So, I start running a series of tests where I design campaigns using different combinations of the variables above to see if I can improve my return on investment and I measure the real time results carefully. Using these tests, and the media math that goes with them, I can gradually dial-in better and better performance.
But, you may be asking, what about brand campaigns where conversion data is foggier and there’s no average sales price online? How do you use media math to measure the performance of those campaigns?
Stay tuned for my next post.
This post was written by John Walker, Founder and Principal at J. Walker Marketing. Contact John directly to discuss your marketing challenges. John@JWalkerMktg.com.