Transcript
Okey dokey, the clock has struck one o'clock so we're going to get started. Welcome to Cooking Up Successful Meta Ads with Limited Targeting Ingredients. I've got John Paglio here in the room who's going to be taking questions in the Q&A. This will be about a 30-to-40-minute webinar, and then I'll have some time to answer any of those questions at the end, so if you have them, feel free to pop them into the chat throughout, and then we will review some answers at the end.
I’m just going to give people a few more minutes to come in the room and then we'll get going. All right, folks. Let me start by introducing myself. My name is Izzy Seidman. I am a digital marketer at flyte new media. I've been with this Portland-based digital marketing agency for four years now, which is crazy to say. And in that time, I've run Facebook and Instagram ads for more than 20 clients across a lot of different industries, including those in B2C, B2B, and a few in the nonprofit sector. So I definitely have a variety of experiences and understand that there are unique limitations for each of those categories.
This is also my first webinar, so I appreciate everyone being here, and thank you in advance for your patience as I go through this for the first time. You can also connect with me on LinkedIn, I've popped a little QR code here if you want to take a sec to scan that, or you can find me by searching, ‘Izzy Seidman’. I would love to connect, and especially if there are things that I don't get around to covering today, which I'm sure there will be, please feel free to reach out to me and just strike up a conversation and hopefully I can provide some advice that helps you on your journey.
All right, I want to start with a little story. My boyfriend, this guy right here, Julian, is in a local Portland-based band called Matriarch. If you're curious, you can find them on Instagram. They also have some music on Spotify. They're indie rock. I highly recommend. But I promise no more pushing their band after this.
A while ago the band had planned a trip to an Airbnb together to record their first EP. I'm friends with all of the people in the band, so I really wanted to tag along, but I have absolutely no musical talent whatsoever. What I do have to offer, on the other hand, is cooking skills. And for the weekend, I decided I wanted to play the role of private chef, which was a lot of fun. I had the meals planned out, breakfast, lunch, and dinner for Saturday and Sunday. I had brought all of my ingredients. I had done some marinating. I was going to have a lot of fun.
But the Friday night that we were supposed to arrive, we were playing it by ear. Now, one of the band members had brought a carton of eggs from a friend of theirs who had chickens. But otherwise, we were left with not much to work with. Unfortunately, there was no nearby grocery store. However, there was a CVS. So on that first night, we took a trip to the pharmacy, and it was my turn to get a little creative with how to cook dinner for these five starving artists.
One of the first things that I grabbed was a big ‘ol pack of ramen, as you can imagine. But I didn't want to limit it to just this sort of college meal, so I also grabbed some Spam. Now, if anyone in the audience has never had Spam before, I highly recommend it. I know it's a little bit of a controversial one, but take it home, slice it up, fry it up. It's really good. And then to attempt to be healthy, I also grabbed a can of corn and the seaweed snacks that they had in the snack aisle. All things that were shelf stable and easy to find at a place where you wouldn't normally think of getting dinner ingredients. And the end result was pretty damn good.
Now, something you should know about me is I'm terrible at taking photos of my cooking. So this is not actually a picture from that meal that I made, but it's a pretty close representation of what the end result looked like. And the lesson here is just that you don't need a lot to make something really delicious. This works, obviously, in the cooking scenario. But the same thing goes for running effective ads on Facebook and Instagram.
I think it has worried a lot of marketers over the last couple of years how much has changed in the Meta ads platform as far as targeting goes. And it's worried me, too. I want everyone here to feel like they can still create successful campaigns with the tools that are available. So that's what we'll be talking about today.
All right, let's lay some groundwork. The past decade just across the board has been riddled with some really high-profile data breaches, lawsuits, the development of stricter data protection laws. All of this has played not just into Meta ads, but across the internet. And in the advertising world, especially, there have been small changes along the way – not all of which Meta announces – which has been a difficulty for advertisers to understand what has disappeared, what can you no longer use? So I'm going to go through a couple of the major updates that have impacted targeting over the last few years.
In 2019, we start with the introduction of special ad categories. This is Meta limiting targeting options for specific advertisers in the credit, employment, housing, social issues, and politics categories, in order to prevent discrimination in those sensitive areas.
So what that means for a lot of these categories is you can't target by age or gender and the detailed targeting interests available for special ad categories are also limited. This is beneficial in that we wouldn't want to discriminate against job ads, for example, or people looking for housing. So it's rightly so there are limitations on these categories.
In 2021, this is probably the biggest one, we had the Apple iOS 14 update. So at some point in time, you probably have all seen this little notification pop up. Basically, the majority of users chose to opt out of cross app tracking. During this update, we were all asked to choose. And as you might imagine, most people said, ‘no thanks’. And the impact of this really is that Meta can't see what users are doing across the web. So they used to be able to have information on what searches you were making, other websites you were visiting, what your behaviors were across those different areas of the web. And now that tracking is so limited for most of the users, we are just giving Meta less data on the behaviors and interests of the people on their platform. So any of the interest targeting that we now have is also far less accurate than it used to be.
In 2022, Meta began consolidating interest groups. This is something that's still happening today, so we're seeing those changes continue. And Meta doesn't always announce them, like I said earlier, so we just have to pay attention to what disappears. And what this really means is that Meta took some smaller interest targeting groups and lumped them together into larger audiences. For example, you used to be able to target based on an interest in lightsabers, or in an interest in Star Wars. Now, the lightsaber group overlapped significantly with the people who were interested in Star Wars, so what Meta did was they consolidated both of those interests into just a Star Wars interest targeting. This seems like a really small change, but what it has meant is that there are fewer options overall for how to target your users based on their interests, and that those options are a lot broader than they used to be.
So you might have seen a little banner in your Meta ads manager about detailed targeting options disappearing. Oftentimes, these will be removed automatically from any ad campaigns that are using them. And when you go to build a new audience, those options are no longer available.
We're not done yet. So in 2024, we continue with what was called ‘data sharing restrictions’. So Meta began rolling out restrictions to certain businesses on their ability to share custom events and user data from their website or app through the Metapixel. Like special ad categories, this aims to protect sensitive information.
So for example, if you're a business in the health industry and you have medical history from your users or medical data on your website, it's pretty obvious that Meta doesn't want to touch that stuff. Same thing goes for businesses in finance protecting the financial history of users. So it limits the ability to import any custom events and retarget users based on those events through specific website behaviors.
Then at the very end of last year, and this has been ongoing, we saw a shift towards privacy first, or what's called, consent driven marketing. We're all familiar with cookies. Google had a hot minute last year where they teased that they might get rid of cookies in Chrome. That didn't end up happening, but it has certainly been part of the conversation around privacy across the web. And more websites are giving users control over which first and third-party cookies are collected.
So very similar to that iOS update, what this means for Meta is that there's just less data about users and their behaviors being stored and shared across the web, and therefore being provided to Meta in order for them to understand who's using their platform and what their interests are.
And then still continuing into this year, we've had yet another update, that detailed targeting exclusions have now disappeared. Exclusions you used to be able to build an audience where you said, “I want people who are into this, but not into that.” So nowadays, you can no longer build an audience that excludes users based on interest targeting specifically. There are still some exclusions for retargeting measures like your own website data. But for example, I wouldn't be able to build a saved audience that excludes users in, say, the top 5 percent of household incomes. Like most of these other changes, this just means that your targeting is forced to be a little broader than it used to be.
It's true, these changes have made precision targeting significantly more difficult. But Facebook is still the most widely used social media platform, and Instagram is number three. So there's no doubt that some of your ideal customers are using those social platforms. After this major refrigerator clean out, let's talk about what's left.
There are still some powerful targeting tools. So we do have some options for detailed targeting if you know how to be creative with them, which we'll talk about next. Meta also launched an algorithm-first targeting option called ‘advantage plus’. Most brands can still build lookalike audiences. And my personal favorite ingredient, retargeting. These last two options share something crucial in common, they both use first party data. And we will definitely talk about that in a minute.
Okay, to illustrate some of the targeting strategies that we're going to go through, let's say I run a business that delivers artisan bread weekly to subscribers. We'll call it the cross club. We'll start by talking about detailed targeting, which are called, saved audiences. So you can build a saved audience directly in a campaign, or you can head to the audience sections of ad manager and create a new saved audience. While the options are more limited now than ever, the philosophy around interest targeting remains the same.
To successfully target by interest and demographics, you need to know your customer really well. And this starts with real customer interviews. I know that sounds a little intimidating, but even if it's just a couple of people that, personally who happened to have used your product or service these, this step is really important, not just for your Meta ads, but for all of your marketing. So talk to them about what their user journey looked like. What were their pain points? Why did they end up choosing your brand? What other brands might they have considered during their choice process?
And if you need ideas for questions to ask, you can look at the categories in META's targeting options. Like let's say, education, relationship status, what sort of publications and entertainment do they like, what are their job titles? All of this can be good fodder for how you can directly use that interview data to format your interest-based targeting.
With fewer targeting options and broader categories, I also recommend researching something I like to call ‘affinity interests’ for your customer. For example, in the Crust Club, I may not be able to target users who are directly interested in artisan bread deliveries. But I can target people who, let's say, also shop at Whole Foods, which sells products very similar to those that I'm going to be sending in my subscription boxes, or those who watch the Great British Bake-Off people who are interested in baking but maybe aren't getting into the kitchen to make it themselves.
Or people who subscribe to similar businesses like HelloFresh, for example. I would encourage you to think about which products, services, businesses, or other interests overlap with those of your ideal audience. Because chances are, if you can't find a direct match for an interest that you think your audience member has, you can probably find an affinity interest to that.
And as you build, pay attention to the estimated audience size. This comes with a nice feature if you're doing it directly in the campaign builder. It will tell you if your audience is narrow, defined, or broad, with defined being that golden middle ground. So too big and your audience may end up being poor quality, as in the people that you're reaching don't necessarily have the right intent to take the action you're looking for. And if you get too small, what can happen is you can quickly deplete the audience and you end up annoying them with repeated ads about breadboxes or whatever else you're advertising.
My boyfriend had asked me when I was reviewing my slides, if there was a magic number I could provide for audience size. And unfortunately, there's not. It relies on a lot of macro factors, like whether or not you serve in a small geographic area if you're a local business versus, let's say, an e-commerce brand who can ship products around the whole country. Just pay attention to this and you'll learn based on some testing and any market research you've done, how big of an audience you should be shooting for.
All right. Advantage Plus. Advantage Plus is, I like to think of it as the instant yeast of ad targeting. So it doesn't have the same depth of flavor as Sourdough, aka our interest-based targeting. But it can get you from a bag of flour to a finished loaf of bread in just one afternoon.
To give you a quick overview of what it is, Advantage Plus is an algorithm-first audience that uses in platform data and artificial intelligence to identify users that are most likely to complete your desired action. So when you're building a campaign and you sublet, you select an objective, let's say traffic or awareness. Meta is going to curate an audience using Advantage Plus technology that they think are going to take that specific action.
Let's talk about some of the pros and cons of Advantage Plus. So, pretty easy to guess. One of the pros is that it's super simple to set up. Nowadays when you create a new campaign, Advantage Plus is automatically turned on so you don't have to do anything to enable this feature and it requires reduced maintenance. With interest-based targeting, you might be continually adjusting, changing some of those interests, increasing or decreasing the size of your audience.
With Advantage, you don't really do any of that. It runs itself. So for advertisers who don't have a whole lot of time on their hand or are first getting into ads and they just want to have a good starting point, Advantage can be a really good benefit for that. On the flip side, what that means is that you have limited control over what that audience looks like. You can make some suggestions to Meta, but they're basically going to go out and find the people they think are best responding, not necessarily based on any interest targeting demographics or anything like that. Advantage Plus also offers typically a larger reach.
So we go back to thinking about that audience size. Meta is going to think of more ideas, probably, than any human brain could think of, as far as potential audience members go. So your reach tends to be larger when you're using Advantage Plus in a campaign. However, we don't get any demographic data output. So you don't get to know who Meta is reaching when you use Advantage Plus, aside from some data on gender, age, and what device type they end up using.
That is not going to come back to you with information about the interests of the people who are responding in Advantage Plus, so it doesn't really help you in the long run for understanding your customers and how you can better tailor your strategy, not just in Meta ads, but in your other targeting.
I've found personally that Advantage Plus works well for e-commerce brands. When you are able to connect your catalog, your product feed, directly to Meta's e commerce they charge I think it's a 3 percent fee for every purchase made through their platform. So it makes a lot of sense to me that this algorithm-driven targeting option works well for ads that also financially benefit Meta.
So if you are an e-commerce brand and you have the ability to connect your catalog, I would recommend using Advantage Plus for any shopping ads like those. If you're not running shopping ads, however, it might mean that the quality of your leads is lower. So because the point of Advantage Plus is to go after the numbers they want to create that desired action, Meta is not going to be spending as much time worrying about what happens after that action is taken.
So maybe you do get a higher volume of traffic to your website, but those users are spending two, maybe three seconds on the website, exiting and never returning again. Meta is still going to count that as a win for them because they accomplished your objective, but the quality of that traffic is going to be really low. So definitely pay attention to user behavior and evaluate the quality of any leads that you're generating with Advantage Plus campaigns before you decide if this is the right strategy.
And one last con for Advantage Plus is, in my personal experience, I've found that these campaigns can really quickly accumulate frequency, which is just how many times has an ad been shown repeatedly to the same customer, which obviously can get annoying after a couple times. So I found maybe in the first week or two, Advantage Plus audience campaigns can seem really good, and then they hit this plateau where results aren't increasing, but frequency is. So you can deplete your audience in that sense.
I had one person in the sign-up forum ask how to turn Advantage Plus audiences off. And I will tell you they don't exactly make it easy. Meta really wants you to be using this tool. In the setup during your add set phase, you would scroll down to where you see Advantage Plus automatically selected. You click this switch to original audience options, and then a pop-up window is going to come up and they will prompt you again to decide, no, actually I don't want to use Advantage Plus. And every time you make a new campaign or a new ad set, you'll have to go through that process again. So hopefully that helps.
This is all to say, Advantage Plus targeting is just one of many new AI driven features in Meta ads. And while I personally don't love all of them, I do think AI still has a place at the table when it comes to your ad strategy. I prefer external AI tools like ChatGPT to assist with ad targeting. It offers the ability to create audience personas, conduct mock customer interviews, analyze market data, and lots of other things that you can use with the LLM of your choice.
So if you are interested in integrating AI to your strategy in Meta ads and in other places in your marketing, I definitely recommend that you attend our next webinar by Flyte’s President, Rich Brooks, on Practical AI Marketing. We're hosting it on March 18th, same time on a Tuesday, and we will send invites via email after this webinar is over.
Okay, as I mentioned earlier, retargeting, aka custom audiences, are some of the most reliable audiences in today's targeting landscape. But to make them, you'll need to collect first party data.
So there are three main sources of first party data that you can use in Meta ads. The first is in platform data. So this is collected directly by Facebook or Instagram. It includes things like your followers, people who have engaged with pages or posts, users who have played, let's say, 10-seconds of a video or the entirety of a video. Any action taken on the platform can be retargeted. Your customer list is also a great source of first party data. So a CSV file with things like name, email, phone number, zip code. Any of that information you've collected through email signups or leads can be uploaded as a CSV to your Meta ads manager, and then used to retarget or to create a lookalike audience based on the users in that group.
And then the last, and my favorite, is website and app data. You can send data of users who have visited, let's say, specific pages on your website, or to take in specific actions. So for example, on the Crust Club, if I had users who were looking at my weekly bread menu, I could target just the users who had visited that page with a new set of ads about signing up for our 30-day free trial in order to target using website or app data, for example.
Let's talk about catering to what stage of the user journey they are in. Let's say my Crust Club website has been getting a lot of traffic recently, but my conversion rate is really low, and I want to run some ads to capture potential customers who have explored my brand but not yet made a purchase. I'm going to start by creating a custom audience and selecting website as my source. Then I can get really specific about which website users I want to include in this audience. I can choose my retention window.
So for example, I have 30 days set here. This really depends on what your user journey looks like. If you have information about how long your typical sales funnel is, that can be really helpful in deciding what your audience retention otherwise it's. Best practice to stick with that 30-day window and then I can target users based on those specific URLs.
So like I mentioned before, maybe I want to put in my bread menu as the specific URL, and I can get even more specific by excluding users who have completed a purchase. So I'm avoiding retargeting people who have already done the action that I want them to do, and therefore those ads would become annoying to them. So we would call this something like an abandoned cart funnel or a warm audience. People who have gotten pretty far down the user journey but haven't yet made that final action.
Okay, the question is, how do you do this? If you're wondering how to get first party website data, it starts with a good measurement strategy. You'll need a little something called, Google Tag Manager, which is a snippet of code added to your website that allows you to decide what types of information you want to collect from your website, and where you then want to send that information to. Through Google Tag Manager, you can install Google Analytics 4 on your website, which collects and stores that website data.
GA4 comes with some automatic events, let's say page views and scrolls, which can be really helpful. But the real fun starts when you begin to create custom events, like view content specific video views percent, scrolls subscriptions, and other file download actions like those. So my weekly bread menu, when they watch a certain portion of let's say a behind the scenes bakery clip, I could retarget users based on that custom event. So any action a user takes or doesn't take on your website can be collected as an event.
Then you can use those events to create your custom audiences and look like audiences. I unfortunately don't have time to walk you through all of the wonders of a good measurement strategy, but if you want to know more, I recommend watching the replay from John Paglio's webinar on The ABCs of Reporting. We also have some great setup resources for things like Google Tag Manager and GA4 on our blog. So feel free to check those out at takeflight.com/blog.
Okay. It's time to put all your ingredients together into one cohesive ad strategy. My best advice is to leverage all of the targeting options across different points in your funnel. For example, if your budget allows, the most successful accounts have campaigns running simultaneously through each stage. Cold interest-based audiences, those saved audiences we talked about earlier, and Advantage Plus audiences can be really great for awareness and reach or engagement campaigns. Warm retargeting audiences, like the one we built about retargeting certain users on specific URLs and lookalike audiences can really help push more users to your website.
And hot custom audiences, or customer lists, are great for bottom of the funnel ads. By working at every stage of the funnel, your audiences rise like a good bread, each one feeding the next. So if you can fill your audience from the top, that will end up increasing the number of people you're targeting in each stage down the line, and therefore increase that bottom line results.
I know it can be limiting when you're thinking about how much money you're able to spend on the platform, so if you are unable to run all campaigns at the same time, just start at the top of the funnel and work your way down.
Okay, thank you guys so much for your attention. Just a little reminder, that the best loaf of bread is made with just flour, water, salt, and the creativity of wild yeast. We have a little time for questions, so if people have any in the chat, John's just going to jump in and let me know what we need to answer.
There aren't any questions at the moment. For everyone, feel free to jump in and ask a couple questions if you'd like. Now's your time.
All right, Kathy's asking, “What's a good starting daily spend?” Okay, so it depends on what your objective is. I know that's a terrible answer. It depends, but I tend to say five dollars a day is a good minimum. You should pay attention to what your cost per click or cost per result is. If your cost per result is higher than your daily budget, that means you're pretty much not going to get any results. So start by gathering a little bit of data. Maybe run a five- day campaign with a higher budget so that you can understand what that cost per result is, and then adjust your budget based on that. Is that helpful, Kathy? Hopefully. All right.
One more. Okay. Anna is asking, “My own campaign to a far lower CPC than any Advantage Plus campaign I've run. That's a good thing, right?” Yes, that is a good thing. Yeah, I know, it's funny because when you're setting up a campaign and you turn off Advantage Plus, it will give you this little warning that's, “most advertisers experience a 33 percent reduction in cost per click when they use Advantage” but Facebook is going to be a little self-serving in that sense. And I've found personally that my interest-based targeting audiences do tend to perform better in the long run than Advantage, so you're on track. Keep going.
All right. Thank you again everyone. We’ll be sending up a follow-up email once the download of this is available. It will include my slides and a transcript as well. So if you need to review anything, that will be ready for you probably by the end of this week, but we'll notify you by email. Keep an eye out for future invitations from us to other elements of flyte school. It's a big push for us this year and we're really excited to bring more educational content to you guys.