Using metrics to inform social media and display ad design

With a wealth of data available at your finger tips, it’s never been more important to leverage that information to shape your output.

Using data from multivariate or A/B testing enables marketers to test different creative solutions for a successful campaign. It might be a different headline, lead image or perhaps differing versions of a landing page which have alternate layouts. And this data can then be analysed and fed back into creating the next campaign.

Given the constraints of display and social in terms of pixel size and volume of content which can be housed, there are limitations (particularly when you compare to a website and how much scope there is for data driven design) but nonetheless small tweaks and changes can still be powerful.

For example, consider two CTAs: Shop Now and Buy Now. Shop Now is more inviting and may resonate better with a product launch, whereas Buy Now has more immediacy and might work better for seasonal sales. And a combination of metrics and A/B testing will enable you to understand which CTA performs better. Or if you are able to update your creative once it has gone live - for example via dynamic ads - you might first run with Shop now, and then move to Buy now for the second phase of the campaign.

In terms of Display ads, the key performance metric is the ‘clickthrough rate’ which tells you how many people have clicked on your ad based on the number of ‘impressions’ (another metric) which simply means how many times your ad is served.

Depending on which ad platform you are using there are further metrics you can capture. For example, DoubleClick from Google can tell you how much of a video in an ad has been viewed, and whether a user has interacted with an ad such has hitting a play button.

There are many elements of the ad you could consider A/B testing to find out what performs best. For example, if your campaign is for a single product such as a phone, what shot best works for the phone? Is it the phone held in a hand, a phone showing someone interacting with it or the phone on its own against a plain background. Do potential customers engage better with a lifestyle shot or a product shot?

When it comes to social there are many more formats across multiple platforms for both organic and paid but similar metrics apply. For example impressions and clickthrough rates are available on Facebook and Instagram which can also tell you engagement rates broken down by likes, comments and so on. Consequently the data is more granular and of course you can interact directly with your customers.

All of this data can then inform design. Coming back to the phone launch example, if you are running an Instagram story is it better to open with a video, product or lifestyle still? It is commonly accepted that video performs better, but do you know what sort of video content your customers prefer? For example, if you are a food retailer, do your customers prefer seeing video of food preparation or of people eating? Data will not only inform design but enable a better ROI.

When it comes to landing pages, the metrics will come via the above sources (assuming you ran a campaign associated with your landing page), in addition to your web analytics. Bounce rate, time on page, clicks - these are all available metrics and if your marketing team is running tracking pixels, then there should be a wealth of data to use and feed into your campaign designs.

The flip side to all of this is that you don’t necessarily want data to be the sole driver of design (unless you are running retargeting ads which might pull in ecom photography), but this is where having a form of design system can work allowing for differing creative flex. For example, if you are running display ads for a sales push, they might be simple punchy designs because ultimately customers want to know the price / discount and product it applies to, but if you are running product launch ads they might be less data driven to allow more creativity and tie in with the TV, print or DOOH campaigns.

There are many ways to slice the data, but ultimately having a solid understanding of what is resonating with your customers will then enable you to feed that knowledge into your designs.