Are you constantly trying to catch up with every new social network that comes out?
Do you find yourself juggling with copy and images to make sure to share at least a certain number of social media posts each day?
Everyone seems to be so focused on quantity that we often forget the impact that small, incremental improvements can have in our social media strategy. When was the last time you tried smart variations to see where they took your engagement and conversion numbers?
Granted, experimentation can sound a bit intimidating. I know because I’ve worked with hundreds of entrepreneurs who feel that “social media tests” are some obscure tactics only data scientists can handle. Nothing could be farther from the truth.
Given the right tools, and having adopted the right mindset, I strongly believe that anyone can manage a lean social media testing program. I’m writing this article to share seven simple tests that you can run today.
But before we begin testing, you need to lay out some metrics to input your data in for analysis.
Setup Your Metrics for Documentation
To make sure your test results are as clear and informative as possible, decide which metrics matter for the changes you are trying to provoke. For example, if your main social media goal is to increase shares, prioritize metrics like shares, retweets, and repins.
That being said, for each of the following test ideas, I’ll be sharing two alternative social media post versions that you can try. When you think about these different versions, here’s a handy format to collect data:
In the “What’s different about it” column, you can specify the variable you’re testing — say, if you’re experimenting with a black background in one graphic (Social Media Post A) and a green one in another (Social Media Post B).
The second column, “How many people saw it”, would be where we consider each post’s reach when we compare them.
It would be unfair to judge how good or bad the variables did if they each had a different chance of reaching your audience, but were measured according to one fixed metric of effectivity.
For example, if one post resulted in 100 shares and was shown to 1,000 followers, is that comparable to getting 100 shares from a post shown to only 500? Nope.
In the third column, “What’s your success metric?”, insert the measurement that makes sense for your social media goals. For this example, we’ll use the amount of shares.
To come up with a rate that fairly shows each variable’s effectiveness, compute for what percentage of the people who saw this post actually shared it.
This is what our table would look like after a sample test:
Now comes the final and most delicate part of this testing process. How are the two figures you came up with so different that we can be certain that changing a post’s background color ( in this case) will drive better results moving forward?
We can make this as simple or as complicated as we want, but I just want to share a quick tool that helps me figure out if the differences I’m spotting in social media graphics are statistically significant and I can confidently call out a winner.
GetDataDriven’s A/B Significance Test was created to test different versions of a website. Ignore the page wording in here and fill the form’s fields with your posts’ reach and share numbers.
See where it says “your A/B test is statistically significant”? That’s what a successful experiment looks like. Ready to test ideas?
01. Shift Your Images’ Focal Point: Objects VS. People
Try sharing the same post with a different image: one where the main center of attention is a person, and one where you are directing the viewer towards an object.
In a study about imagery in web design, a group of scientists found that human images with facial features induced a user to perceive a site as more appealing, having more warmth or social presence.
On the other hand, using an object as the focal point could help emphasize product features that users are actually interested in previewing before they click. Will this app, item, service be worth my time?
That’s when an object-based approach like displaying a mockup or style photo can draw attention to particular actions that your user will be able to perform if or when they make a purchase.
Think about the last time someone mentioned a mobile app and showed a compelling screenshot: did that trigger your interest to click on that brand’s social media image?
This approach might be a better fit when your social media followers are highly involved in your core product’s technical features.
To try this method, follow these steps:
- Share a social media post attaching an image of an inanimate object
- Share the same social media post, attaching an image of a human face instead. Try to make this second post go out at a similar/identical time and day of the week to prevent other circumstances from affecting (and potentially confounding) your test’s results.
- Measure shares, views and comments to see how you audience reacts to each.
02. Play Around With Your Images’ Design: Color and Typography
Select a single design element, whether that’s color or typography, and create alternative versions of your social media image using a different level of each.
For example, you could share the same social post twice where the only change is the use of a specific color (versus working with black & white) in your palette.
While it might seem like a minor change, modifying a graphic’s main color can notoriously impact its effect on viewers. In 2015, Buffer released a comprehensive roundup on the most effective fonts, colors and templates used in high-converting social images.
They quote a 2009 study from Georgia Tech where over 1 million Pinterest graphics were evaluated to determine which colors were related with the most shared pins.
Researchers concluded that red, purple and pink promote sharing, while green, black, blue and yellow all stop people from sharing.
Considering that fonts can also play an instrumental role in how social media graphics are shared, launch two posts where the only difference is the font used to communicate the message.
03. Play With Your Call to Action and Landing Pages
Are your followers feeling compelled to click whichever link you are sending their way? Experiment with different action verbs (or calls-to-action) to see which one is truly catching their attention.
Also try out different landing pages to test what happens once followers actually click through.
Here are some ideas you can try:
- Appeal to humor
Is there any way you can approach the need to check out this product/service with a fun angle? For example: “No more ugly sweaters. This year give them something they’ll actually keep!” — for a Holiday retail campaign.
- Appeal to fear
Would anything negative happen if this follower doesn’t use the product/service? For example: “Don’t let allergies catch you off guard. Check out what X can do for you!”
- Appeal to happiness or fulfillment
Highlight the positive results obtained by using this product or service in your call to action. For example: “90% of users report increased productivity — are you ready to try X?”
04. Try Animated GIFs
Visuals no longer have to be static on social media. Facebook, Twitter, Google+, Tumblr and Pinterest now all support animated GIFs, leaving ample space for experimentation. Follow these simple steps to determine how effective GIFs are in comparison to static images:
- Go ahead and create a simple animation with a quote or data point that might interest your audience.
- Use the same data point or quote to create a static image. Alternatively, grab a still from the animation in step 1.
- Take note of how sharing a static image compares to that animated GIF’s performance in terms of clicks, comments and shares.
05. Experiment With Copy and Tone
Sometimes we forget that human conversations vary in tone, length and intensity. As reflections of natural speaking, social media posts can also portray that diversity. Try rephrasing your initial posts in at least 5 different ways:
- State your message as a question
For example: “Did you know that there are at least 10 free social media tests you can run today?”
- State your message as an exclamation
For example: “Social media tests can be simple and fun! Try these → “
- State your message in imperative mode
For example: “Run these simple social media tests and watch your numbers soar:”
- State your message in first person
For example: “I’ll give these social media tests to try ASAP → “
- State your message in third person
For example: “Here are 10 social media tests that you should try ASAP → “
Similarly, emoji provide an entire spectrum of emotional expressions that you can use to make your message sound more human. See what I mean?
06. Target Different Audiences
Sometimes your message can’t be described as effective or ineffective in absolute terms. Whether a specific type of copy or image appeals to your followers or not may well depend on what that target audience is like. In consumer behavior, we segment different human groups according to two different types of variables: psychographic and demographic. While psychographic variables include lifestyle, attitudes and preferences, demographic ones refer to more “hard line” traits like age, gender or location.
Both Facebook and Twitter allow you to segment the paid ad campaigns that you decide to run. On Twitter, you can target users by psychographic variables like interests and demographic variables like location and gender.
Similarly, Facebook allows you to create a special audience with demographic characteristics like a certain age, and psychographic characteristics like their behaviors.
Examples of social media tests with psychographic segmentation:
- Launching one message for Mac users and one for Windows users
- Launching one image for design enthusiasts and one for business professionals
Examples of social media tests with demographic segmentation:
- Sharing one message with young men 18-30 and one with older men 31-50.
- Launching one image for women and one for men.
- Testing one version of your (English) copy with Canadians and one with Australians.
07. Try New Time Zones
With social media scheduling tools like Hootsuite and Buffer it has become increasingly easier to send out messages using multiple platforms at off-peak times.
Have you ever tried appealing to a time zone that is radically different from your own? If you are in UTC-5 (New York), it would be interesting to try out a few shares in UTC+1 (Central European Time).
The team behind CoSchedule, an editorial calendar tool, compiled results from 10 different studies on which times/days are best for sharing social content. Check out the article to find effective time slots for Facebook, Twitter, Pinterest, Linkedin, Google+, and Instagram.
If you’re having a hard time coming up with original content for those additional time slots, try Edgar — a social scheduling tool that literally recycles your posts so that they get maximum exposure.
08. Optimize Shareables for Each Platform
Are you doing enough to engage users in different social platforms visually? Here are some examples of image sizes that you may want to try in Canva:
- Pinterest: Create long, pinnable images
- Facebook: Create rectangular images following Open Graph recommended specs.
- Twitter: Tweetables + Twitter Cards
Different social platforms indicate a specific image size that adjusts to the way they display information. Optimizing for these various sizes will result in a better experience for your existing and prospective followers.
In my experience, optimizing graphics for Pinterest’s long, vertical approach to images can radically improve engagement metrics. Of course this is subject to how important the Pinterest demographic is for your product/service and how much engagement you currently get in that platform.
To launch a lean experiment with image dimensions, try sharing out a square image on Facebook followed by a rectangular graphic to link to the same article. Compare how the two fare in terms of shares, comments and clicks.
09. Use Hashtags
Hashtags are tools to organize content. They help us navigate through the cluttered interfaces that are most social networks these days and find content related to our specific queries.
Since your audience is probably using these tags to discover content, it is in your best interest to find out which the most popular ones are in your space. Once you do, go ahead and tag your posts to make them even more visible for your audience.
Hashtags also lend themselves to experimentation, since different groups of tags will appeal to different crowds. Explore, for example, what happens when you target #startups vs #femalepreneurs. The results might surprise you. Here’s a simple approach you can get started with:
- Select two hashtags that are directed at two different niches
- Create and share a social media post using one of them.
- Create and share a social media post using the second hashtag. Make sure this second post goes out in similar conditions (day, time, season).
- Measure both posts for likes, comments and shares.
10. Experiment With Mosaics
Sometimes readers want more than a simple snapshot that represents the entire piece of content you’re linking to.
Displaying a more elaborate graphic can help give followers a preview of what they will find in the body of your article or landing page by displaying some of the key visuals they’ll be exposed to once they click.
Fortunately, Canva makes building these mosaic graphics incredibly easy. To launch a quick test, share the same piece of content (article, ebook, etc) using a single image and a mosaic of several images compiled into one.
Create your own mosaics and measure how users respond to the visual in terms of clicks, likes, shares and comments.