How to Estimate the Cost of A Facebook Ad Campaign
There is a lot to like about Facebook advertising, but one of the biggest challenges still remaining is figuring out how to make it work for you. More importantly still, how much will it cost to build a successful Facebook ads campaign?
Here, we will discuss some numerical rules of thumb that I have used to estimate the potential cost of an exploratory campaign.
Determining Success Metrics
You might think this is easy. Lots of entrepreneurs and potential clients coming to me want website conversions at a sustainable price. Usually it is some sort of order.
However, sometimes they are interested in something a little higher up in the funnel. For example they might be looking for email conversions or app installs. This is more common for B2B-type businesses when you are looking for marketing leads to feed to a salesforce.
When that is the case, you want to be fairly sure of the relationship between the ultimate conversion (like the order) and the “secondary” conversion higher up in the funnel. It is not uncommon to have it such that there is a Facebook campaign collecting secondary conversions at a very high pace only to find none of them falling through to the ultimate conversion.
Collecting Data Points
If your predecessors have run a Facebook campaign before then this is a good place to go and start collecting that data for use.
Usually I calculate website conversion rate on Google Analytics by marking a conversion as a Google Analytics “Event”, tagging it into a Goal, and then viewing the goal in GA’s “Source / Medium” screen in Acquisition // All Traffic submenu.
If the prior Facebook Ads efforts were not as thoroughly documented then you can still get a lot of useful information.
If the prior Facebook Ads campaigns sent traffic that was not properly tagged with the right UTMs then I have found that you can still get a semi-decent idea. Untagged Facebook ad traffic tends to show up on Google Analytics as “facebook.com / referral” (desktop) or “m.facebook.com / referral” (mobile). Unless you have a lot of traffic coming from Facebook (unusual if you are not a publisher) you can get a decent sense of how Facebook Ads traffic worked.
Estimating Traffic Conversion Rates
If your predecessors have never run Facebook before then you can get a decent measure of how it will perform at its most ideal by looking at other website traffic sources.
Paid website conversion is never going to be as good as what you can get from organic or referral traffic. Generally I can estimate how paid traffic would perform by taking about 50–75% of the conversion rate I get from organic traffic.
There are caveats of course — especially when the desired goal is a free secondary conversion (like if you want them to download something) and in that case I would say the relationship does not hold — but it is a decent rule of thumb for ecommerce.
Estimation Traffic Costs
You would think that this would be super easy. Just take the Facebook-reported CPC, right? Wrong. The actual number is a bit more fuzzy.
The actual Facebook metric you should be looking at for a traffic source is Link Click. Facebook buries this deep but it strips out other clicks on a Facebook ad (like a Like or a Reaction).
There’s one more thing to add on too. With certain types of Ads I notice a drop-off from the Link Click to an actual Google Analytics recorded traffic visit. For Desktop it is pretty small, about 10–20% (meaning that 80–90% of our Link Clicks end up becoming recorded traffic) but on Mobile it is more like 30–50%. You should keep that in mind and apply that “tax” to your Link Click CPC.
You can estimate the cost of acquisition by taking the cost per link click and dividing it by your conversion rate.
The estimated individual cost of conversion can then be parlayed into the full cost of a info-seeking run. If you are running a small campaign I usually stop when once the campaign gets to 25 conversions. To account for random volatility I layer on half of the desired number of conversions. So for 25 that is about 25 + 12.5 = 37. To make the calculation easier, I round that down to 35.
So if we have an estimated $50 cost per acquisition then we budget about $50 x 35 for $1,750.
Like I said, this is a general rule of thumb. Some quick bit of mental math that I do in my head that I use when setting expectations for experiments or client calls. There are a lot of possible conditions that can change and drastically hurt conversion. We can talk more about what those might be in a future post.
Originally published at AdsJunket.