How To Achieve ALL Your Paid Traffic KPIs: Part 1

By Joey Alter / August 12, 2016

For most of us, running paid traffic on Facebook has been a series of trial and error. Maybe you are a small business owner, and years ago you heard that Facebook traffic was the next big thing. Maybe you worked at an old school paid search agency and your boss told you to get some FB campaigns up as an added service offering.

img2-2-02

Up until recently, if you wanted help with your facebook traffic, you had to hire an agency, or buy a course, or read obscure blogs. Now Facebook has started offering a certification at facebook.com/blueprint. But a lot of you probably do not have the time to go through the many modules offered there and glean your key insights… Don’t worry! After running facebook traffic for over 50 businesses in the past three years, I have seven best practices to help guide you through the morass that is power editor & ads manager.

Best Practice #1: Tracking Setup With New FB Pixel // Custom Conversion Setup

As you may have heard, Facebook is sunseting the old conversion pixels in early 2017. While you will still be able to use conversion pixels for tracking until February 15th, 2017, Facebook will stop actively maintaining them in October, 2017

If you haven’t already, you should begin to implement the new Facebook pixel which combined the old functions of conversion pixels, & custom audience pixels as well as providing some newer, more advanced features.

If you’re running campaigns already using the old conversion pixel, I recommend placing the new pixel on your key post conversion pages. Leave both the new pixel and the old on for 2-4 weeks. Once you have a good amount of conversion data on the new pixel, remove your old conversion pixel and relaunch your campaigns and set them to optimize on your key conversion event.

There are two ways you can track conversions with the new pixel. Standard events or with custom conversions. Here is an example of the new pixel code:

<!-- Facebook Pixel Code -->

<script>// <![CDATA[ !function(f,b,e,v,n,t,s){if(f.fbq)return;n=f.fbq=function(){n.callMethod? n.callMethod.apply(n,arguments):n.queue.push(arguments)};if(!f._fbq)f._fbq=n; n.push=n;n.loaded=!0;n.version='2.0';n.queue=[];t=b.createElement(e);t.async=!0; t.src=v;s=b.getElementsByTagName(e)[0];s.parentNode.insertBefore(t,s)}(window, document,'script','https://connect.facebook.net/en_US/fbevents.js'); fbq('init', '1234'); fbq('track', "PageView"); // ]]></script>

<noscript>img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1234&ev=PageView&noscript=1" /&</noscript>

<!-- End Facebook Pixel Code -->

The new pixel works by using javascript “events’. If you place the default pixel code in the header of all pages on your site, the only event that will fire is the “PageView” event. To track you key conversions you need to slightly edit your pixel code on all your post conversions thank you pages. Here is a screenshot of all supported conversion events:

paid media paid traffic KPI facebook

As an example, if you are running leadgen on facebook, the code on your optin thank you page would look like this:

<script>

!function(f,b,e,v,n,t,s){if(f.fbq)return;n=f.fbq=function(){n.callMethod?

n.callMethod.apply(n,arguments):n.queue.push(arguments)};if(!f._fbq)f._fbq=n;

n.push=n;n.loaded=!0;n.version=’2.0′;n.queue=[];t=b.createElement(e);t.async=!0;

t.src=v;s=b.getElementsByTagName(e)[0];s.parentNode.insertBefore(t,s)}(window,

document,’script’,’https://connect.facebook.net/en_US/fbevents.js’);

 

fbq(‘init’, ‘1234’);

fbq(‘track’, “PageView”);

fbq(‘track’, ‘Lead’);</script>

<noscript><img height=”1″ width=”1″ style=”display:none”

src=”https://www.facebook.com/tr?id=1234&ev=Lead&noscript=1″

/></noscript>

<!– End Facebook Pixel Code →

As you can see, all I’ve done is add a line of code below the PageView event, which tells facebook to also fire a “Lead” event when a user loads the page. I have also changed the noscript line to fire a lead event. This image code only comes into play if the user has an adblocker installed on their browser or has javascript disabled.

The other option to track conversions is to use Facebook’s new custom conversions. These work by using either PageViews based on url slugs, conversions events, or a combination of the two.

kpi facebook paid traffic

To set up a custom conversion, go to tools >> then custom conversions. You will need to gather your key post conversion urls, and paste them into the box like I’ve shown above. Then you choose the category for the conversion event.

You can also chain together urls and have facebook only count users who visit url1 and url2 as a specific conversion. I recommend using this option if youre funnel has multiple possible buying paths with different values for each.
There is one other use case for implementing custom conversions over standard events. The first is if you run an ecommerce store and you want to run specific facebook campaigns for specific products. You’ll want to create product specific add to cart and purchase custom conversions and let data accure on those two. Then once you have a significant amount of purchases, ideally 2-4 weeks worth, you will switch your campaigns for Product A to optimize on the purchase event for product A.

facebook kpi ads paid traffic

If you are using shopify, this should be relatively easy to set up. Create a new custom conversion, choose event, and then choose the purchase event. Under more parameters, choose content_ids, and choose the product ID of the specific SKU you want to optimize for. Then choose purchase as category and you should be all set.

Facebook Best Practice #2: Thinking With Lookalikes

This next tip is for advertisers who have run facebook campaigns before and/or have sold online before. New advertisers will need to run traffic for a bit before taking full advantage of this strategy.

If you’ve run facebook traffic before, you’ve probably used a lookalike. Most likely, you used a 1% lookalike off of your leads, your buyers, your page likes or your old conversion pixels. You probably added interests to your campaigns as well, and didn’t see as good of results.

Theres a reason for this. When we target an ad to a lookalike on Facebook, we’re saying “Hey, show my ad to this subset of Facebook who are similiar to my core audience/seed audience/ uploaded list”. Those users in the seed audience have taken a desired action before, they’ve bought and we’re asking facebook to show our ad to people who have shown similar characteristics to our users who’ve taken a key action.

Targeting interests or behaviors alone has no such algorithmic benefit that lookalikes provide, when we target interests we are just asking for impressions of users who have been collected in a particular interest based on their posts, page likes, apps downloaded, and online behavior on sites other than Facebook as well.

Before you go and test 25,50 or 100 different specific interests, I’d recommend testing more lookalikes based off of different seed audiences. Think about it like this, you may have one group of 2,000 buyers of your product. But in reality, inside that group of 2,000, you have multiple subgroups of users with different groups of common characteristics.

If you subdivide that group of 2,000 in significant ways, and create new seed audiences and corresponding lookalikes, you will audiences that are partially the same but also different. Here is a cheat sheet of different seed audiences you can use for ecommerce, leadgen, or mobile gaming:

  1. Lead list upload
  2. Buyers list upload
  3. Pixel lookalike from leads
  4. Pixel lookalike from buyers
  5. Lead event lookalike
  6. Purchase event lookalike
  7. Leads from a specific funnel
  8. Buyers from a specific funnel
  9. Mutiple time buyers
  10. High ticket buyers
  11. High retention users
  12. High retention users by device OS (android vs iOS)
  13. Leads/Buyers for mobile vs desktop
  14. Leads past 30 days, 60 days, 90 days
  15. Buyers of a specific in game currency pack, buyers of specific in game currency pack by device
  16. Users who’ve reached level 5, level 10, etc in your game
  17. Users who’ve visited your site more than X times in Y timeframe
  18. Top 1%, top 10% of all app openers
  19. Use utm tags to identify your interests by url slug, make a lookalike from a custom audience from that interest
  20. Male traffic lookalike vs. female traffic lookalike
  21. Lookalike from users who’ve searched specific key phrases
  22. Lookalike from all site visitors
  23. Lookalike from email openers
  24. Lookalike from email clickers
  25. Lookalike from users with a high net promoter score

This is by no means an exhaustive list of potential lookalikes, if there is a significant group of users that you track for your business, and you aren’t reaching the level of scale you want, test it!

Also, make sure to keep the next best practice in mind when using this strategy…

Best Practice #3: Using Exclusions To Decrease CPM & Increase Return

Here’s something most people don’t know about Facebook ads:

If you are running more than four active ads for a specific audience on facebook, you are bidding against yourself and receiving a CPM penalty.

Let’s say you’re running an account, your traffic is a mix of conversion objective campaigns targeting lookalikes from 3 different seeds, and 4 interest clusters as well. In this case, unless you are using custom audience exclusions and/or flex targeting for interests, your are bidding against yourself.

To clarify, each demographic slice, gender slice, placement slice of an audience group is treated as separate. So if you are running a 1% lookalike, targeting males in one adset, females in another, you CAN have 8 total ads running to those two audiences.

Also, if you have an audience bidding for clicks, and bidding for conversions concurrently, those are treated as two separate audiences. When you choose a bid type on facebook, you are asking to be served impressions of users in that target audience who are likely to either convert on a specific action or likely to click an ad or likely to engage with a post, view a video, rsvp for an event, etc…

The easiest way to avoid CPM penalties on your campaigns is to exclude your highest percentage lookalike or lookalikes from ALL of your interest campaigns. So if you have 1% lookalike a and 1% lookalike b and interest clusters 1-4, you are setting a custom audience exclusion for lookalikes a&b for interests clusters 1-4. Running them in separate adsets.

This strategy works with more mass appeal offers/products/services. But if you have a product that is more niche, you can try excluding your interests from your lookalikes.

In the past year, Facebook has introduced flex targeting, or in plain english, the option to exclude certain interests from your target audience. To perform this in power editor, go to the adset level add your target interests and then click “exclude people”. You will see a box like this pop up:

kpi facebook tracking tagging

If you have a more niche product, and have seen success only when layering lookalikes & interests, this is the strategy for you. Target lookalike 1, layer in interest cluster A. Target lookalike 1, layer in cluster B, exclude cluster A. Go through all your interests like this, using progressive exclusions.

Note: DO NOT exclude both ways, do not exclude cluster B from A, because in that case you will lose your core users.

Final note on this best practice, add CPM as a column in your ads manager reporting. If you are trying to diagnose a campaign issue, you should easily be able to see CPM spikes and tie them to recently uploaded ads, and then fix the exclusions.

An easy way to simplify your paid traffic workflow & make decisions based on reliable data

While having a proper pixel setup is necessary to get the most value out of your campaigns, you’ll never able to rely completely on javascript pixels to be 100% accurate. And furthermore, if you rely on facebook reporting to run your campaigns, you’re at the mercy of a buggy system where you have to add/remove & change the order of columns that aren’t intuitively named. Or maybe you can’t load any data from your facebook reports. Or maybe you can’t sort your ads in descending or ascending order of spend because you have too many adsets.

Want to Make it Easier to Run Your Facebook Ad Campaigns?

Sign up for a free 14-day trial of FunnelDash now. You’ll get our customizable dashboards that pull all your clients’ data from their Facebook Ad Accounts and put them into easy-to-understand formats. 

You’ll also get our agency funnels, which will help you run lead generation campaigns for your agency, so you can bring in high-quality leads that you can convert into high-paying clients.

Click on the button below to get started.


>