5 Lessons I Learned from my Snapchat Advertising Test

I’ve been watching Snapchat from afar ever since it launched. From afar means I do what I typically do with social media platforms. I quickly build an account and then watch user behavior without adding much to the conversation. It’s a little like that guy in your office who always stands around conversations without ever saying anything. It might be a little creepy, but it’s a great way to find out how people interact. The second I saw simple Snapchat advertising start filtering its way into the Snapchat experience, I had two opposing thoughts. The advertiser in me was thrilled that a new platform might be available. The human being in me was disappointed that myself and marketers like me would soon be responsible for ruining yet another cool communication tool.
It would be much later that the everyday marketer would be able to use the platform for self-serve Snapchat advertising. MUCH later. While the early adopter/big budget brands had been spending a decent amount of money to help Snapchat figure out the kinks in their system for quite some time, it wasn’t until last week that I was able to access and test Snapchat’s answer to Facebook advertising. In about the same time as it took for a dancing hotdog to become the most culturally important piece of meat of a generation, I pulled out my company credit card and ran a test. Read on to see five things I learned that could be useful for the everyday marketer.
SNAPCHAT ADVERTISING CAMPAIGN SPECS
I wanted to test a multi-day Snapchat advertising campaign to see how performance might vary from weekday to weekend. The campaign ran from early Friday afternoon to early Sunday afternoon. We’ll get into this later, but the minimum daily spend sits at $100. I spent $300 total on this test. Snapchat offers four optimization objectives. For the sake of well-suited creative that didn’t offer a landing page destination, this campaign was optimized for long-form video views. The creative featured an episode of a new social web series we were working on. It hadn’t been released yet, but a few episodes had been filmed. It was a good fit for the feel of Snapchat and ran 2:08 in length. You can check it out below.
THE LANGUAGE
Every new platform comes with a new language. How would we know that a 3-second video view is valuable unless Facebook’s terminology told us it was? (Sarcasm.) Here are the definitions of a few items you might come across in the Snapchat advertising platform.
Top Snaps — The creative a user first encounters. There are many variations of Top Snaps you can utilize. This is what encourages a user to swipe up to see more.
Swipe Ups — Think of this like a click. A user sees a Top Snap ad and swipes up if they want to see more.
eCPSU — Your total spend divided by the number of swipe ups delivered. Think of this like a cost per click.
Attachments — These are way cooler than the spreadsheets Josh always emails you. The Attachment is what the user sees after they swipe up on the Top Snap. It could be a long-form video, article, website, or app install.
Average Attachment Screen Time — The average amount of time, in seconds, your attachment was viewed.
Now that you’re a Snapchat advertising pro, let’s get into the good stuff. Check out my list of five things I learned from my advertising experiment. Comment with questions or suggestions!
- IT STILL COSTS TO PLAY THE GAME
You no longer need $400,000 to step onto the court, but it’s still not the most cost-effective platform. Compared to Facebook’s $5 a day minimum for a similar campaign, Snapchat’s $100 a day minimum feels like a cable buy. Is there a way around it? Sure, but you’d have to be ok with your flight lasting less than a day. That’d be some pretty real-time messaging. The minimum bid per video view on my test campaign was $.05. That’s pretty high for a social video view. The final CPV ended up being $.01, but your bidding strategy does seem to be limited by a minimum. Most social advertisers will love that your overall budget does auto adjust to favor the highest performing ads. We’ve been spoiled by Facebook into assuming that’s automatic, so I was pleased to see some optimization was automated. The CPM on my short test was $3.42. All of the monetary numbers look pretty solid. As long as the minimum bids and daily spend numbers come down as they onboard more regular users, the spend barrier shouldn’t prevent too many marketers from participating in Snapchat advertising.
2. IS A VIDEO VIEW REALLY A VIDEO VIEW?!
We live in a News Feed world where convincing someone to watch something for 3 seconds is worth celebrating. I won’t get into a marketer’s favorite bar conversation on whether a user’s 3 second glance actually makes an impact, but it’s important to know what Snapchat considers valuable enough to charge you for. It’s especially interesting on this platform because one ad could actually contain multiple videos; the Top Snap and the Attachment. Snapchat describes a video view as the total number of impressions that meets the qualifying video view criteria of at least two seconds of consecutive watch time or a swipe up action on the Top Snap. Think about that! A video view doesn’t take your full Attachment video into account at all! A video view happens when your Top Snap is viewed for two consecutive seconds OR A SWIPE UP ACTION OCCURS. That means you’re getting charged for people to see and possibly swipe up on your “tease. I’d love to see that change to being charged for views on the long-form content I created an ad specifically for. As with most social video view metrics right now, it seems like Snapchat is padding their numbers with this convenient metric to encourage other advertisers to jump on board. Am I upset about the length of time it takes to be counted as a video view? Not really. Two Snapchat seconds is like 10 on Facebook. Considering the longest Top Snap can only be 10 seconds, a 20% view-through rate isn’t terrible.
3. SWIPE UPS ARE CLICKs
In my test campaign, 1.5% of users swiped up on the Top Snap to see the full video. The closest comparable metric I can find to a swipe up on other channels is clicks. With Snapchat allowing you to take users to your website and app all in the platform, a swipe up truly becomes a click-through. Compared to an average display click through rate of .35% according to Hubspot (we’re still doing that?), 1.5% is pretty solid. Compared to Facebook, Snapchat actually comes up looking good. It’s close, but most Facebook benchmarks have average click through rates hovering around 1.4% according to Word Stream. It’s important to keep in mind that this campaign was optimized to produce video views, not website traffic. I would be interested to see how a traffic campaign would perform with this metric. If you look at the average Google search ad click through rate of 1.91% according to Hubspot, Snapchat has some work to do. It seems pretty self-explanatory if you understand anything about user behavior on the platform, but Snapchat wouldn’t be my first channel to utilize if traffic was my end goal. Although the click through rate isn’t as low as I expected it to be, swipe ups are a big ask and a fairly new behavior that Snapchat has to train its users to use. While the swipe up percentage wasn’t terrible, my attachment completions were. Out of the 1.5% of users who swiped up to check out the full video, only 300 completed it. That’s paying $1 for every video completion, which is pretty unacceptable. Was the video too long for a platform that’s famous for 10 second videos? Probably. Was the targeted audience too broad to be interested in a very specific concept? Definitely. We kept our targeting broad to see who would engage (more on that later). Was our test video so good that people couldn’t finish watching without incurring bodily harm? Probably not. Maybe… but probably not. For future campaigns, I would want to play around with video length, format (ours was a traditional 1920 x 1080 instead of Snapchat’s vertical recommendation), and concept to get more swipers to watch the entire attachment.
4. STILL TOP SNAPS CRUSHED VIDEO TOP SNAPS
The title on this one is pretty self-explanatory. Our campaign had a mix of four different types of ads. While all had the full video as the attachment, we tested multiple different types of Top Snaps to see what might encourage the most swipe ups. The highest performing ads were, by far, the ads containing Top Snap images with slight movement instead of Top Snap videos. The two highest performing ads received almost 10x more impressions, produced more than 16x more swipe ups, almost doubled the swipe up rate, and received more than 11x more attachment completions. It wasn’t even close. You can see a quick example of a still Top Snap and a video Top Snap below.
5. END OF CAMPAIGN REPORTING NEEDS SERIOUS WORK
This was the biggest disappointment of my Snapchat advertising experience. I had no idea who engaged at the end! Snapchat’s version of reporting is a simple download function of the campaign’s performance metrics. The biggest thing Snapchat ads are missing right now is the ability to tell me what parts of my target audience were interested in what I was saying. I don’t know if a younger or older audience spent more time watching my ads. I don’t know if the geographic targeting even worked. I have no idea if my viewers were interested in online shopping or dog walking. While they have targeting tools similar in nature to Facebook and Google, they’re not currently using any of it to give me useful, post-campaign audience data. Facebook tells me which age groups, genders, and geos interacted with my ads on which devices and because of what they’re interested in. Snapchat kept me in the dark on what I was most curious about: What parts of my target audience loved the ads and what parts didn’t? Here’s to hoping they can layer pre-campaign targeting data into post-campaign performance in the future.
It was a great test, and I would absolutely buy Snapchat advertising again. As they scale and convince more marketers to experiment with the platform, I’m hoping minimum pricing decreases and reporting capabilities increase. Do you have questions about my test campaign that I didn’t answer in the article? Send me an email at MattMcAllisterMarketing@gmail.com, and I’ll do my best to answer each one!
