One of the most common questions clients of my digital marketing agency ask is “How much with Facebook ads cost us?”
The obvious answer is “As much as you want to spend.” After all, if you only have a $500 ad budget, you only need to spend $500. Facebook will stop serving ads when your budget is exhausted.
But that doesn’t really answer the question.
What clients are really asking is: “What will I get for my money?” In other words — will that $500 get me 100 visits to my web site? Just 50 visits? Will it get me 200 Page Likes? And, more importantly, is there a particular segment (age, gender, interests, etc.) which would be cheaper for us to target?
You will stretch your advertising dollar much further if you can advertise to a cheaper segment. But how do you know what that segment will be?
First, let’s talk about Facebook’s pricing model.
How Facebook Ad Pricing Works
Wouldn’t life be simpler if ads were just a set price, like newspaper ads or radio ads. You want to be in the News Feed? That’ll be $1 per ad. But it doesn’t work that way, and that’s a good thing.
Facebook is trying to make the user experience pleasant, above the needs of advertisers. After all, nobody wants a News Feed full of ads, so they limit the number of ads in the feed. To get one of those spot, you have to bid to get there. While you can set a manual bid, I’ve found better results to just let Facebook price it.
But how much does it cost? That depends entirely on who you want to reach — and how many other advertisers are trying to reach the same people. Remember, unlike a newspaper, there is not one single News Feed on Facebook for all users. Everyone gets their own unique News Feed, made up of content from their friends and Pages they like.
You Pay for the Person, Not the Ad
Imagine that you sell high-end razors for men. And let’s say your perfect target audience is men in their 40s in the United States. At first blush, it makes sense to target those men in ads.
But you need to think about how many other advertisers are also trying to reach American men in their 40s. That’s probably a lot of competition. In order to be seen by your target group, your bid needs to be higher than people selling golf, wine, viagra ((you can’t actually advertise viagra on Facebook)), sports cars, and other products sold by companies with big advertising budgets.
The more companies are vying for those eyeballs, the higher the per-ad cost is going to be.
Now imagine that you’ve learned that, for some reason, razors are used a lot by men in their 60s in Africa. There are much fewer advertisers trying to reach that group, so your price for each ad will be dramatically less.
If only you could find out what smaller groups are cheapest to reach, you could position your campaign around those people, benefitting from the much lower cost, and leave the high-fliers to the companies with big wallets.
Well, you can. But not many people know how.
How to Identify the Cheapest Segments
There is, sadly, no database of people-and-prices. Again, this is because the price is dynamically based on how many other advertisers are trying to reach each group. You might start a campaign aimed at African 60-year-old men and get ridiculously cheap rates, but if Nike, BMW, and Apple all decide they want to target that group too, your price jumps immediately.
So the way you find your cheapest segments is to test all the segments with a small portion of your ad budget and measure it.
Most people do this by simple split-testing — creating an ad group targeting men in their 60s, then duplicating that ad group and creating new ads for men in their 50s, then repeating that process for each group they want to reach. This is slow and painstaking work.
Let’s say you want to find out what genders and age ranges will be cheapest. You’d need to create 12 different separate ad groups (male teens, male 20s, male 30s, male 40s, male 50s, male 60s, female teens, female 20s, female 30s, female 40s, female 50s, and female 60s). And the reporting will be a manual process. This could be a 3-4 hour process just setting the ads up. Then, after your test campaign has run, you’ll download an Excel dump of all your data and be left to fend for yourself to do the analysis.
How to Do All That in Five Minutes
At my Vancouver social media agency, we routinely do multivariate testing for our clients all the time. In fact, most of our campaigns test up to 96 ad variations at the same time. We monitor each combination and pause the groups which are costing more.
The secret is a great tool called AdEspresso. For $49 per month, you can run unlimited ads and unlimited tests. (Use code TODMAFFIN to get 15% off their rates.)
Once your test is off and running, AdEspresso will show you the biggest factors affecting your ad prices.
In the case of this real campaign I’m currently running, the ages are the biggest factors. People in their 30s are costing me much less to reach people in their 40s. Notice on the right, there’s a little pause button that will let me — in one click — pause all groups that are aimed at the older decade. This would take you forever in Facebook. Here, it’s one click.
Here is a great tutorial on using AdEspresso:
Launching Your Test
Once you’ve set up your tests, start them running! You can test combinations of gender, age, interests, countries, headlines, body copy, and more — but remember that the more factors you test the higher a test budget you’ll need to get meaningful results from each group. For that reason, I usually only test two or three factors. A budget of $250 should be sufficient to give you a good idea.
At the end, you’ll be able to use AdEspresso to easily chart which group performed better, based on whatever criteria you decide “better” is (cost per click, click-through rate, cost per conversion, etc.)
Here is a real-world test I’m currently running. I’ve asked the platform to graph age and gender against the cost per click.
As you can see, people in their 30s are currently the cheapest group to advertise to. People in their 40s are costing me about 2.5x what people in their 30s are. And I definitely want to stay away from women seniors, as they’re pricey too.
But what if you already know (or think you know) what your demographic target is.
Let’s use another real-world example for a client of our digital marketing agency (a major sporting event), where I already know the target I want — young women. In this case, I want to test what headline they respond to best. I tested two headlines — one based on emotion, and the other based on a traditional call-to-action:
- It’s Time. Make This YOUR Year.
- Click Here to Register.
What I wanted to measure was the cost for conversion — the actual price I’m paying to get a real, paying participant — against gender.
In this test, men responding to the emotional headline are costing me twice what the other headlines are costing me. I would pause this group immediately.
As you can see, this kind of split-testing is critical to your Facebook ad campaign.
If you want to try AdEspresso for yourself, go to http://AdEspresso.com and use the code TODMAFFIN to get 15% off.
(Note: I’m running a split-test to promote THIS POST and I’ll update the results in a couple of days. Be sure to return to see them!)