Metrics: The Language of Facebook Ads Success
Any time I hear conversations about the miraculous power of Facebook and possible the “tricks” of the trade, they tend to revolve around the sexier aspects.
Things like Facebook’s ability to target users in almost-scary fashion (something something Cambridge Analytica) or characteristics that make a good creative usually take center stage.
Rarely do I hear any talk of metrics in these conversations. And for good reason. Sure, if you’re a true geek, metrics may be exciting. But most people would rather watch an egg boil while doing jumping jacks on lego blocks than master an understanding of metrics.
So how does one make metrics sexy? By showing you what sort of results you can get with a metrics-focused mind.
Here are the results from a few accounts I’ve worked on–all with a focus on metrics first and foremost.
Fashion – a high-end luxury $1000 handbag.
$39,496.96 spent, $116,779 in sales, 2.96x return on ad spend.
Food/Drink – Fitness oriented meals delivered weekly.
$20,723.87 spent, $82,101.64 in sales, 3.96x return on ad spend.
Consumer electronics/Toys – a computer kit for kids (16 days).
$167,580.94 spend, $616,904.36 in sales, 3.68x return on ad spend.
Supplements (Sprayable Sleep) – a sleep melatonin spray.
$224,856.39 spend, $479,158.58 in sales, 2.13x return on ad spend.
Politics – A US congressional campaign.
$24,317.57 spend, $52,994.14 in donations, 2.18x return on ad spend.
Technology/Hardware – a wearable device (cold audiences only).
$18,726.21 spend, $40,126.74 in sales, 2.14x return on ad spend.
You cannot consistently churn out results like these without seeing the Facebook ads universe through a metrics lens. To do so would be like studying economics without the language of math or losing weight without the language of calories, macronutrients, or even carbs/net carbs by those in the keto religi…errr…doing keto.
Metrics are the key to unlocking the full potential of Facebook ads.
Developing an understanding of them allows you to methodically create successful campaigns even through a myriad of changing factors.
In almost all of the client examples that I shared, I took over an ad account that was originally losing money. Usually a company/product founder launches Facebook ads on their own, and like most campaigns (including the ones that I run), they’re unprofitable and give up.
What they didn’t realize is that they were actually on the right track. The difference between the person above and someone who runs ads successfully is that the latter realizes that getting initial results are part of a process and utilizes metrics to iterate until profitability.
An Introduction to E-Commerce Metrics
Whenever I look at a poor-performing Facebook ad account, I can take one look at their key metrics and have a decent understanding of:
- Why their account isn’t profitable.
- What they need to do to get there.
To understand how this is possible, we need to understand the concept of a “conversion funnel.” This might be basic stuff for some of you in this group, but I suggest reading it anyway as a refresher.
The Conversion Funnel
A “conversion” is any sort of action that you want a user to take. For most e-commerce companies, the ultimate conversion is a “purchase,” but it may also be other desired actions too such as getting someone to:
- Submit their email or sign up for a newsletter
- Download an app and/or sign up for an account
- Finish an application process (e.g. applying for a loan)
- Getting someone to subscribe to a recurring purchase
In order to model the success of getting to a conversion, marketers have created a concept known as a conversion funnel. You might hear this referred to as a “marketing funnel” or just a “funnel.” A conversion funnel maps out important events that necessarily occur leading up to a conversion are outlined in the order they occur (from top to bottom or left to right) along with the amount (or proportion) of people who make it to each step.
For example for a person to make a purchase on an e-commerce site, they need to perform the following actions:
- Start on the homepage (or some other landing page)
- View a product
- Add that product to their cart
- Click the checkout button
- Add their payment information
- Click purchase
As we said earlier, anything can be defined as a conversion, which means that a marketing funnel can encompass any type of event leading up to that conversion.
Let’s take a look at a real life example–an obesity clinic whose conversion goal is to get people to book an appointment with a clinician. The steps leading up to booking an appointment are:
- Viewing their information page
- Clicking on a CTA (“call to action,” i.e. some sort of button or text that brings them to the next step) to go to an obesity calculator
- Filling out their age
- Filling out their height/weight
- Submitting the calculator
- Signing up for a session with a clinician
We can see their actual numbers below. Notice where the highest and lowest dropoff rates occur.
Core E-Commerce Metrics
Let’s get back to e-commerce metrics. Most people do pay some attention to metrics. Unfortunately, they mostly focus on top level ad and traffic metrics–such as Page Views and CTR (click-through rate).
These “vanity metrics” as they’re sometimes called are easy to understand and may allow you to brag about really large numbers, but they’re not always relevant. In fact, sometimes they are in diametrical opposition to the types of things you want to achieve.
For example, wealthier individuals are frequently less inclined to click on ads. If you were selling a designer handbag and prioritized the optimization of a high CTR % over other objectives, you would be attracting customers that won’t purchase your product at all.
Below, we’ll go through all of the metrics that actually matter. It may seem like a lot of information to swallow at first glance, but don’t worry. We’ll be leaning on them repeatedly and they will eventually become second nature.
These are metrics that involve any sort of purchase. Whenever I analyze an account or campaign’s performance, these are the first metrics I look that, because they give us a high level indication of profitability.
Average Order Value (AOV) = total revenue / total purchases
This is the average amount that a purchaser spends. You’ll sometimes head AOV referred to as “basket size” or “average basket size.”
Customer Acquisition Cost (CAC) = total purchases / total ad spend
This is the average cost that it takes to acquire a customer.
Lifetime Value (LTV) = total sales (all time) / total customers
LTV is important if you have a product that customers purchase repeatedly. If LTV is relatively high compared to CAC, then it might be worth giving a large coupon or first month free (for subscription products) because you’re willing to eat some of the initial cost, knowing that once they become a customer, you’ll collect a lot more from them over their lifetime (e.g. internet/phone providers).
Cost of Goods Sold (COGS) = total amount spent on your product’s components / total products
This is the average cost of the item that you’re selling. This can also be described as a percent of the dollar amount that went into “producing” the product. For example, if I buy a laptop wholesale for $500 and resell it on Facebook for $2000, then its COGS is 25%.
COGS varies depending on product and domain. For example, a company that sells high protein meals may have a 50% COGS, whereas selling a company that sells software or a political campaign that takes donations has a COGS of 0%.
Return on Ad Spend (RoAS) = total sales / total ad spend
Achieving a high RoAS is usually the end goal for most campaigns. However, whether or not a RoAS is “good” entirely depends on COGS, and sometimes a high RoAS shouldn’t always be the end goal.
For example, let’s say that you’re selling an e-book. Would you rather spend $10,000 and get $40,000 back (4.0x RoAS)? Or would you rather spend $1,000,000 and get $1,500,000 back (1.5x RoAS)?
In fact, you can actually calculate your breakeven RoAS: the RoAS that you’d need to achieve in order to neither make a (gross) profit or a loss.
Breakeven RoAS = 1 / (1-COGS %)
This means that if 75% of your product’s sale value goes into making the product, you’d need to hit a whopping 4x RoAS in order to breakeven on Facebook.
You can read all about the benefits of focusing on CAC vs. RoAS in my previous blog post.
These are metrics related to ads–they’re universal to all platforms (with the exception of some Facebook-platform specific definitions below, such as “link clicks”). Whenever I’m “debugging” an ad account has poor performance, I’ll look to these metrics to see if there are any indicators that ads (or sometimes the audience) may be at fault.
Impressions = the total number of times your ad was shown
This is the total number of times that your ad was shown somewhere on the user’s screen (e.g. a Facebook newsfeed or Instagram story). This doesn’t necessarily mean that the user paid attention to your ad or clicked on it–just that the ad was shown.
Reach = the unique number of people that saw your ads
This is the total number of people who saw your ad at least once.
Frequency = impressions / reach
In other words, this is the average total number of times that your ad was seen by someone. For example, if you had 100,000 impressions with a reach of 50,000 people, your frequency would be 2.0. In general, you want to keep frequency low (we’ll discuss that more later).
Cost per thousand (CPM) = The average cost per 1,000 impressions that you paid
CPM is the “cost” of ads, and it’s determined by the value of a specific audience through ad bidding (more on this when we talk about audiences).
Clicks (link) = the total number of people who clicked the main link on your ad
Whenever I say “clicks,” I am referring specifically to link clicks. There are many available definitions of “clicks” on Facebook. Make sure to focus on link clicks, as the others are essentially vanity metrics. For example, “all clicks” (as opposed to link clicks) also encompasses people who might have clicked to read more comments.
Unique CTR % (link) = clicks / reach
This is the percentage of people who saw your ad that clicked on the link.
Cost per click (CPC) = total ad spend / link clicks
This is the average amount that you paid for someone to click on a link.
Facebook Funnel Metrics
We briefly talked about what a typical e-commerce site’s funnel looks like. One of the brilliant things that Facebook did was to standardize the key events that occur in the purchase process. For example, any time an item is added to a shopping cart, an “AddToCart” event should fire. We’ll talk a little bit more about the importance of these events when we discuss the Facebook pixel next.
Each event, like “AddToCart” has three important metrics that accompany it:
- The total number of times it occurred
- The unique number of times it occurred (i.e. how many people fired that event at least one time)
- The cost per unique number of times it occurred
|Event||When it occurs||Why it’s important|
|ViewContent||Someone views a page that contains a specific product’s details (example).||It’s the first event that the user is interested in learning more about a product after the landing page.|
|AddToCart||Someone adds an item to their cart.||It’s the first event that indicates a purchase intent after the user knows the price.|
|InitiateCheckout||Someone clicks on the “checkout” button and/or starts the checkout purchase.||It’s the first event that shows a serious intent to purchase.|
|Purchase||A product is purchased.||Obvious reasons.|
So why is this important?
Because each key metric corresponds to different part of the user’s path to purchase.
For example, you know that users who added an item to their cart (AddToCart), for example demonstrated an intent to purchase while already being aware of an item’s price. You know that users who initiate checkout were satisfied with what they were getting (before shipping and taxes) to close the deal.
When you combine this information with the metrics we discussed above, such as AOV and CAC, then you can start to understand the narrative around any campaign.
For example, here’s a healthy food that ran Facebook ads for a paltry 0.32 RoAS.
What can we immediately understand from this?
Well, CAC is $108.58, which might be fine if average basket sizes were super high, but in this case they’re a mere $35.75 ($929.45/26). Conversion from AddToCart -> InitiateCheckout -> Purchase looks healthy, however, as there’s usually a 50% drop between each step (very generally speaking).
Let’s keep going up the funnel a bit and see if we can find any other points that call our attention.
We see that the biggest jump in cost/conversion is from ViewContent to AddToCart ($8.30/user to $85.55/user).
If prospective customers are interested in more information about an item but not adding it to their cart, then there are some possible assumptions we could make:
- The item is too expensive.
- The user’s interest, which I like to call “buying temperature” as a shorthand, wasn’t high enough–there wasn’t enough value demonstrated through the USP (unique selling proposition), social proof (look at all the people buying this product), scarcity (only x units available, so act now), etc. This probably deserves a longer post later.
We know that the AOV is $35.75, which doesn’t seem like a price that would scare anyone off by sheer magnitude of the number. This fact, and a relatively low conversion of all clicks to ViewContent (20%), highly suggest an ineffective landing page. Improve the landing page’s ability to raise buying temperature, and you’ll be able to convert at a profitable RoAS.