Expanding your reach into new social media channels can be daunting. Not every platform is the correct choice for every brand. What platforms you should focus on depends on your goals and audience. A common mistake in social media advertising is spreading dollars across too many platforms, including the ones that don’t make sense for your company. What’s more, many brands ignore platforms they don’t understand, missing out big revenue opportunities.
In this recorded webinar (with recap below), Two Octobers’ Head of Paid Media Chloe Pascoe outlines a test-and-learn approach, so you can feel confident about where you spend your time and money.
How do you determine the best social media platforms to advertise on for your industry, as well as for your specific company?
The most important thing to remember with a test-and-learn strategy is to set appropriate expectations. Be honest and transparent about what you plan to do, what you expect the results to look like, and how you’ll report back on results.
TikTok is a perfect example. When TikTok came out, I think advertisers as a whole discounted it. We thought: TikTok, that’s just people dancing to videos. But that is bias talking. What research have we done? Is TikTok actually going to work for my company—is it going to give us growth? Even though I’m not on TikTok, should I consider it? That’s why it’s important to test.
Testing allows us to compare the performance of leads from different channels. Sometimes, social ads and paid search ads work against each other. If you find social ads performing really, really well, while search ads aren’t, you might want to re-allocate your media spend. Testing helps us uncover these opportunities. You’ll also want to compare the quality of leads between channels. If social is reaching your audience, but it’s not converting, then that’s a different conversation. Testing sets us up to have that conversation.
Testing can actually be used to expand your campaign coverage over time. Keep an eye out for good results from testing new audiences, new locations, or new content that should become part of your ongoing campaign across multiple platforms.
I recommend using a separate budget for testing, so that you don’t dilute your current efforts, if that’s possible.
Make sure to research and budget for the cost differences between platforms. Each platform has different average costs. For example, LinkedIn is more expensive per click than other platforms, but it’s a great place to be if you’re selling b2b.
Determine and align on the purpose of the test. Are you testing awareness, aiming for website traffic, or trying to grow conversions? There are lots of ways to measure success in social media. When you’re clear and aligned as a team, you’ll be able to make good decisions throughout the test, as well as interpret the final results.
Don’t forget, each platform measures data and reports on it differently. TikTok and Facebook report different metrics, and even the same metric can mean slightly different things. Make sure you understand these differences and set expectations with your team.
When advertising on a social media platform for the first time, focus on seeing if it’s the right platform for you. So if you’re going to TikTok for the first time, find out if anyone wants me to be here—is this my place to be? Select KPIs that reflect that goal. Testing new platforms tends to require loosening the purse strings. Getting top ROI should not be your goal with this kind of text.
Alternatively, when testing expansion, you’re trying to build on your existing success. You’d approach the test differently by testing new content or audiences, and setting KPIs to demonstrate performance improvements over-and-above current results.
Different advertising platforms reach people in different stages of the purchase funnel. With search advertising, we’re reaching people who have already expressed an intent, indicating that they’re already in the purchase funnel. But in social, we can serve to most any user of the platform, reaching audiences that we might not be able to reach in search.
Design the test with an understanding of which stage of the funnel you’re targeting: awareness, consideration/relationship or conversion. Which stage(s) are we focusing on, and what are our goals for each stage? We may set a target for new users reached in the awareness stage and a target for conversions. But it’s important to align on which goals you’re focused on—and which you’re not—when setting testing expectations.
We recommend reporting bi-weekly on tests. Be transparent; report on both positive and negative findings. Not being straightforward about test numbers can create distrust, both for agencies and in-house teams. The test reporting should be focused on saying, “this didn’t work—what’s next?” or “this did work—let’s continue.” Both help us think about how we are going to move forward.
Provide industry and client-specific trends as a baseline for comparison. If the entire industry is down during a period, we’d want to consider the results of our tests differently. If Q2 is generally a poor quarter for your marketing program, you should not be aiming for peak performance, but rather an improvement over typical downturns.
I typically recommend six weeks before evaluating a test, but sometimes teams are reluctant, and two weeks might fit better. I would never recommend less than two. Ensure leadership is on board with your timeline.
Align with leadership about the ultimate metrics that will determine the success or failure of these tests. Leadership may want to see impacts beyond impressions, clicks and conversions, such as average cost of sales (ACoS) and ROI. Ensure you’re set up to capture and report accurately for these metrics too.
Test-and-learn requires daily check-ins. It’s not test-and-forget! Do daily pulse checks, and create weekly reports. With larger tests, you’ll be using real-time data to make some optimizations during that time frame. And any test can develop unexpected problems. Reviewing the test consistently is essential to keeping on track.
We often start with a test about the viability of a platform, and after demonstrating some traction, move into further testing to refine and improve our outcomes. Our approach will change from “can we deliver ads and engage with this audience effectively?”, to “where can we find the best performance within this platform?”
These are a few types of test you can pursue:
If you’re still wondering, “where do I start?” take some inspiration from my favorite tests:
To reiterate an important point: we recommend reporting about bi-weekly on the tests you’re running. Maintain trust by transparently reporting on both positive and negative findings. Tell the story with the data to guide and report on what should happen next. There’s always a lesson to be learned from a test–how will you leverage those learnings into your next test? And consider recommending a pivot in the existing test, rather than ending it all together.
Need a little help with your social media advertising? Reach out, we’d love to help.
ChatGPT traffic in the GA house! Plus new features in GA4 and understanding GTM first-party…
This article details the process of building two BigQuery tables for path analysis, with a…
Preview five great dashboards for SEO reporting and analysis, and find the one that works…