Merry Christmas my ninjas,
Snapchat Santa is over here with your free knowledge BOMB.
Secure The Bag With Snapchat: Media Buying Guide
Rule #1 Tracking is 🔑
- First you need to set up your tracking. Without tracking you’re literally burning money.The recommended tracking tools we recommend are:
- Google Analytics: (here make sure to create Goals, so that if people visit your thank you page, the conversion is tracked),
- Voluum: – We use Voluum to help is understanding our traffic on a fine tune level.
- Hotjar: (with the Heatmaps and Recording active on your pages), and the
- Snapchat Pixel: This should be obvious but I have to include this for this. For my Padawans.
Confirm that all your tracking tools are working fine. Use Google Tag Manager to place all of your pixel in there, so that all your tracking codes can be handled from one place without confusions.
Its important to note that when you setup your Snapchat pixel, do it with Google Tag Manager and set up a 5 sec delay, so that the bounce visits and people who visited your site by accident won’t be taking into consideration when you build your custom audiences and lookalike audiences later.
Optimize For Conversions
2) Create a Website Conversions campaign. At the beginning, since the Snapchat pixel is new, the ads will be optimizing for Swipes, but once you get 1,000 conversions on the other Events, Snapchat will allow you to optimize other events as well.
3) Create a minimum of 5 ad sets (ideally with 4 ads with different videos on each, unless you already have a proven winning video from previous data) in which you’re split testing different audiences.
- -The recommended format we recommend is the “Snap Ad”, with the User Stories placement. Also (specially if you’re running e-commerce) test the Discover Feed placement as well, which you can choose with the Story Ad format.
- Goofy and funny ads tend to work very well.
Test Test Test Till Paydirt!
You should test interests, but also advance demographics such as financial situation and education level.
- If you’re targeting an English speaking audience, make sure to select the “English” in the Language option.
- It’s important to be testing only 1 variable at the same time. For example, if you’re going to target split test “Credit Card Owners” VS “Back to College High Spenders”, make sure that they have the same age and gender.
- The daily budget of each ad set should be enough so that your ad spend is at least 3 times (ideally 5) the cost per acquisition you can afford for the desired action.
Cut Losers, Scale Winners!
4) After a couple of days of having your ads running (2~3 days), check out the data and pause the worst performing ads.
5) Now we proceed a do another audience test. Here do the split test by doing a breakdown of your winning audiences by age and gender.
- In this stage, if you see that according to your tracking there are specific geographic regions that are not performing well, here it’s time to exclude those regions.
Birds Of A Feather Flock Together
6) With your newfound winning audiences, it’s time to split test the other variables. Variables that are going to make the biggest difference (besides the audiences) are: videos, text inside the video and your landing page. When you do these tests, make sure to have the Share option enable (according to our tests, enabling it will get you a cheaper cost per swipe) and the Preload option off (keeping it off might not improve your conversions and will mess up with your tracking tools).
Buy Your Data!
7) Once your pixel has hit 1,000 Events (which you can see in the Snap Pixel section), it’s time to create your lookalike audiences. There are 3 levels of lookalike audiences you can create: Similarity, Balance, and Reach. Create those 3 lookalike audiences, and they’ll be ready to use in approximately 24 hours.
8) Now it’s time to test the lookalike audiences. Create a new website conversions campaign to test them. At this point, now you should be able to not only optimize for Swipe, but also for the other events as well, such as Page Views, so you should start taking advantage of that feature. Test your lookalike audiences in different ad sets in the new campaign and let them run for a couple of days to see which are the winners.
- If you’re going to optimize for Page View, you’ll have to use a Manual Bid. The manual bid you should have should be between somewhere close the minimum amount you need to break even and the amount you need to get the desired ROAS while it keeps spending. For example, if you know that your conversion rate on your website is 3%, and your payout per conversion is $10, the you might want to test using a bid between $0.30 (since with a 3% conversion rate that would give you a cost per acquisition of $0.30/0.03= $10, which would be the breakeven point) and $0.15 (which would get you a cost per acquisition of $0.15/0.03= $5, which would give you a 200% ROAS).
9) Once you’ve found the winning lookalike audience, something else you can test is creating interest overlays with those lookalike audiences (for example, you can test an audience of Credit Card owners and that also belong to the Balance Lookalike audience), as well as the other variables, such as videos, text in your ads, and your lander.
- When testing your lander, something that can really make a huge difference is your headline. From our experience a good headline can make a difference (in one of our test we went from getting a 1 sec average duration on the site, to over 1 minute on the site, just by changing the headline).
- When split testing changes on your lander, we recommend that test the changes in different ads instead of simply rotating the landers, since Snapchat doesn’t like that you make too many changes to your lander once an ad is approved, so rotating might put your ad account in risk.
Surf The Profit Waves Bruh!
10) When scaling, go wide. Instead of having few ad sets with high budgets, it’s better to have many ad sets with lower budgets, and start surfing, that is, that if you see that an ad set is performing very well, you can increate the daily budget by 50%, and then ad the end of the day put it back to how it originally was at the beginning of the day.
- Always pay close attention to the ad fatigue. If you see that your lifetime Frequency. If people have seen your ad multiple times, the performance of your ad might start dropping, so when that happens it’s time to rotate the creative.
11) After you have been doing manual optimization for a while, you might want to create some rules on tools like RevealBot to automate your process.
- You can create a rule to turn on and off your ads at a certain time.
- A good rule to have is to pause the ads that get a cost per lead 3 times higher than your breakeven point.
- Also, you can put a rule for the surfing, which is about increasing your daily budget by 50% in the middle of the day if you ROAS very good (like higher than 100%), and then at the end of the day reduce the daily budgets to their original amount.