- What We Do
- Who We Help
- Case Studies
- About Us
If you've run an ad campaign you know the hard part isn't launching - it's running them profitably over time.
But it's hard to know where to start and what to focus on first. Ever feel that paralysis of analysis? We want to kill that.
In today's video, we'll walk you through a simple system that we use to test and optimize our campaigns over the long-term.
One of the most important things you should learn to master is how to optimize your campaigns over the long-run, how to make them better day by day, and when we look at the data, when we look at what Facebook is telling us, or what AdWords is telling us, it can be overwhelming.
There's that paralysis of analysis, where you don't really test or try anything. So I want to share with you a very simple framework that we use to really ask better questions and to come up with more meaningful tests, so that we can improve our campaigns.
The framework that we're gonna walk you through is specifically tailored here for Facebook, but it's really something that you could use across nearly any type of marketing or advertising channel.
So first things first, let's think about kind of the three different things that we're gonna go after here.
First is deciding what your goal is. So what are you optimizing towards? What's the outcome that you want to have?
Second we want to understand what are the key performance indicators? What are the things that are gonna measure your progress towards that goal?
Third, what are the exact experiments or tests that you're going to run to try to reach that goal? And of course the key performance indicators are the things that you're gonna measure.Those are the ways you're gonna measure the experiments or the tests.
Take the goal first of ad engagement. Anytime you run a campaign, the first thing that we're looking for is are we getting attention? Are we getting engagement from people when they're seeing this ad?
So the way that you'd measure that are things like click through rate, comments, relevance scores if you're on Facebook. How are people interacting with your ad? Are they clicking through? Are they tagging and sharing the ad with friends? Is the message that you're putting in there getting to the right people and is it really resonating? Is it causing people to take an action?
IMAGE ON TESTS
Tests that you can run here are just optimizing that message, so it's the image, it's the headline, it's ad formats. And depending on what kind of budget you have to spend, you can test a lot of these things at once, but we recommend even just testing one thing at a time. Just try one headline versus another, try one image versus another, and start tweaking and optimizing from there.
Okay, let's take the goal of audience refinement. So it's not enough just to get ad engagement, we don't just want to get people to interact with our ads, we want to get the right people to interact and to click through.
So, what we're measuring here are conversions.
Who are the types of people who are coming in? If you're business to business, are you getting people that are the right job title? Are you getting decision makers? Are you getting the people in the door that you really want to be talking to?
And we put here micro conversions as well, because there's other, smaller actions that people can take to maybe identify that they're the right person.
Maybe that's downloading a different eBook, or maybe it's going through and using a free tool that you have. Whatever those micro conversions are or even if it's just clicking through to some other articles that you have on your blog, whatever those smaller micro conversions are can be a point of measurement here for reaching the right people.
And then how do we experiment, how do we optimize?
Well, a lot of the name of the game here is to segment and to get more specific with the people that you're targeting.
So if you have an initial audience that's just based on interests or maybe based on a lookalike list, then you're gonna wanna segment by things like demographics, are you breaking ages up into say 18-24, 25-30.
These different segments of life that people go through where, you know, maybe you're in college, and then you're out of college, and then you're in your first job and then you're getting married.
All of these different ages matter, because you're going through different phases of life. So are you testing in those different demographics, particularly age? Interests, behaviors, guys, the name of the game here is just to take your larger audience and break it up into smaller groups and see how those smaller groups respond to the message that you have.
Another one here is, a goal is optimizing your landing page, making sure that where you're sending people is converting at the highest rate possible. So our KPI here is conversion rate. Are people taking action, are they able to, is there enough trust there? Is there enough of a compelling headline or reason? Is there enough value in the offer that you're giving them? These are all things that affect your conversion rate. And the tests here, and we put a caveat in, is you want to run smart A/B tests.
So gone are the days where you can just be satisfied testing the blue button versus the red button. And how do you test things in a smart way?
Well, it takes some careful analysis of how people are interacting with your landing page. Some favorite tools of ours are like Hotjar, where you can actually watch video sessions of people interacting with your landing page.
Of course, Google Analytics.
But as much as you can gather that behavioral insight, how do people interact with your page and what does that reveal about what you should maybe test.
Is it really difficult to access your form on a mobile device? Okay, let's maybe run some tests that simplify that process for mobile devices.
Are you finding that people who convert also tend to watch a video? Okay, well, let's test auto-playing a video for everybody who lands on your page.
These are just things that you have to uncover by watching people use your site or at least using some sort of analytics to determine how people are interacting with the page.
Okay so here's three examples of goals. These aren't the only goals that you can have, but they do kind of measure through the funnel or through the process of how somebody could interact with your ad.
First there's ad engagement, then there's kinda refining the people that you're reaching with your ads in the first place, and then of course there is addressing the landing page and how we can get that to convert at a higher rate.
I recommend not trying to run all of these tests in parallel. Try to isolate one of these variables one at a time, so that you can really focus and kind of a have a control and have a test, and not be trying to blow through a bunch of tests all at once. It's gonna make it a lot more difficult for you to understand what's really going on here. And what types of changes are moving you closer to your goal.