Before you read this blog post, I have a request. Close your eyes and think about a recent instance when you used data to inform a decision.

Having a hard time pointing to a specific use case? I’m not surprised. In fact, 60% of social media practitioners say measuring return on investment for their efforts is still a top three challenge for their campaign planning and strategies, according to Marketing Profs.

Data matters. It’s really easy to use data effectively — and even easier to use it ineffectively. How do you ensure you have an effective data strategy? Talk to your team. Not just the person sitting next to you, but those teams across the office who have different goals and KPIs for the same campaign. The social team could be determining success by likes and comments, the e-commerce team could be looking for revenue and the PR team could be looking for impressions and form fills. If everyone is on the same page, you can design a campaign that will provide data around everybody’s key KPIs.

How can we dig deeper into data to understand what works for our brands? What opportunities are we missing when we don’t pay attention to data? I answered these questions and more during a recent webinar in coordination with DMN, a leading provider of digital and data-driven marketing analysis, research and content. Enjoy!

Secret #1: Benchmark against yourself

As a Success Manager at Curalate, I get asked at least twice a week: “Are my Instagram interactions higher than average compared to other companies in my industry?” or “What is the average revenue for Like2Buy?” (Like2Buy is Curalate’s shoppable Instagram product.)

I understand why brands ask these questions, but benchmarking around one specific data point makes no sense. What seems like apples-to-apples is almost always apples-to-oranges. Brand Y is asking for average numbers for Like2Buy but spends no money on advertising. Brand X’s Like2Buy numbers are higher but they spend $100,000 in monthly ad units driving to their Like2Buy page.

At the end of the day, only you know your brand’s spend and details around what is and isn’t being included in every campaign, so benchmarking against yourself can be even more illuminating.

Secret #2: Correlation doesn’t imply causation

Our sporadic social media behaviors lead us to believe that correlation doesn’t imply causation. There’s really no hard proof to show that just because someone liked your image, they ended up visiting your site, purchasing your product or visiting your location.

Take Snapchat for example. You can provide CTAs encouraging users to screenshot an image to “save” the look. When examining the data, the number of screenshots equals causation because the action requested is the action being measured.

More CTAs in more places will help build your effective data strategy. When your visual content is shoppable or actionable wherever it can be found, you can track the performance of your distributed content and learn which images and products are driving action.

Secret #3: Spot-on ad targeting is not a guessing game

You should be using lookalike audiences for Facebook ads. If you aren’t, you’re just guessing. For those of you that are less familiar with the Facebook ads manager, it allows you to determine who you want to target with ads. You can select things like location, sex, age etc.

I’m working with clients who are gaining audience insight by using Curalate-powered lookalike audiences gaining audience insight and creative content to power smarter and more effective campaigns and ads. Client campaigns are seeing results as high as: 81% increase in CTR, 40% decrease in CPM and 40% decrease in CPC.

lookalike audience

Secret #4: Data is qualitative too

Are you able to tell me on your most recent post, what percentage of comments included a positive sentiment? What about comments that were positive but used negative terminology like: “I’m so disappointed this is sold out!” This is all data — it’s just qualitative not quantitative.

Do you send rights requests for images? Take a look at this amazing interaction with HGTV. Not only did this fan give permission to regram, but she tagged a friend who also had an overwhelmingly positive interaction with HGTV. What percentage of rights request comments are affirmatively returned? This is all data!


Secret #5: When you find something that works, you throw it a parade

I’m not suggesting that you put the same piece of content in every place all the time, but if you know a piece of visual content is killing it — celebrate it! Omnichannel visual content is actually better streamlined — so examine the numbers, figure out what’s resonated best and extend their lifespan by repurposing them.

And don’t forget about causation and correlation. If you have an image that’s driving a ton of clicks for Like2Buy, feature it in an email newsletter. If you have an image getting a ton of positive brand sentiment, put it on your blog! On your website! Celebrate it for getting smiles and likes and leading to further discovery.

This creates recognition around content that has been proven to resonate. When users see familiar content that they’ve interacted with or responded well to in the past, they’re excited about finding it and interacting with it in more places.

Where do I begin?

Here’s some next step ideas:

  • Talk to different teams prior to running a campaign and make sure all different types of data is being measured in the middle and the end of the campaign.
  • Keep a log! Don’t have access to Curalate? Use a spreadsheet to start keeping track of interactions, followers, sentiment, add metadata tags etc. You can organize data manually, it’s just way more work.
  • A/B test. Always. Try a different CTA, try to repurpose an image, swap out a stock photo with an editorial style image, and keep track of what’s performing.