18.12.2024

A/B testing – what works in retail marketing?

In all marketing, an interesting question is which message resonates most effectively. The answer is found through testing, but in traditional marketing, this takes time, and comparing results is difficult. Digital signage makes A/B testing easy. Read this A/B testing guide to find out how you can develop your in-store marketing for the better.

As is customary in a guide, let’s start with the absolute basics. In marketing, A/B testing means presenting two different versions of an advertisement and comparing the results they produce. When it comes to advertisements shown on store info displays, A/B testing helps understand what kind of message and content produces the most desired results. Whether the goal is to bring more visitors into the store or to increase sales of promotional products, measuring results is straightforward when clear data is available. By cross-referencing visitor numbers and sales data with the specific messages displayed, you no longer have to guess which content drives results and which falls flat.

Read also: How do I convert from the shopping aisle? Note these three things!

What should you consider in A/B testing?

The variable element

In a single A/B test, only one variable is compared at a time. If there are multiple differences between the two versions being compared, you won’t be able to tell which of them was ultimately the decisive factor, and you won’t be able to apply the insight to future advertisements. For example, two different versions might each highlight a different feature of the product, or one version might give more prominent placement to the selling price while the other emphasises the discount percentage.

Test duration

The test must run for long enough for the results to be statistically significant. For instance, two different ads can be rotated on alternating days over a given period, and sales on those days can then be compared.

A consistent environment

For the test results to be comparable, care must naturally be taken to ensure that the versions being tested are shown under similar conditions. For example, both versions should have an equal number of peak trading days. It may be impossible to create completely identical conditions, but if the difference in the numbers is large enough, there will be no ambiguity about which version won.

Metrics and analytics

Use clear metrics to determine which version performs better. When the subject of the advertising is a product for sale, the simplest and most relevant metric is sales. If the test conditions are as consistent as possible, the answer lies in the ‘units sold’ data. Other easily measurable things include, for example, the number of visitors who reached a campaign page via a QR code in the advertisement, or another specific action that could only have been prompted through the advertisement itself.

What should you consider in A/B testing?

The variable element

In a single A/B test, only one variable is compared at a time. If there are multiple differences between the two versions being compared, you won’t be able to tell which of them was ultimately the decisive factor, and you won’t be able to apply the insight to future advertisements. For example, two different versions might each highlight a different feature of the product, or one version might give more prominent placement to the selling price while the other emphasises the discount percentage.

Test duration

The test must run for long enough for the results to be statistically significant. For instance, two different ads can be rotated on alternating days over a given period, and sales on those days can then be compared.

A consistent environment

For the test results to be comparable, care must naturally be taken to ensure that the versions being tested are shown under similar conditions. For example, both versions should have an equal number of peak trading days. It may be impossible to create completely identical conditions, but if the difference in the numbers is large enough, there will be no ambiguity about which version won.

Metrics and analytics

Use clear metrics to determine which version performs better. When the subject of the advertising is a product for sale, the simplest and most relevant metric is sales. If the test conditions are as consistent as possible, the answer lies in the ‘units sold’ data. Other easily measurable things include, for example, the number of visitors who reached a campaign page via a QR code in the advertisement, or another specific action that could only have been prompted through the advertisement itself.

What can you test with A/B testing on advertising displays?

Content and message. In in-store advertising shown on info displays, the variable element is typically the headline, text, colour, image, or video in the advertisement. Price, discount percentage, a gift, or a similar benefit to the customer can also be tested, making it possible to find out whether a larger offered benefit increases sales sufficiently to justify the cost.

Time slot and frequency. Instead of testing the content itself, you can test how different times of day for showing the advertisement, or how often and for how long the advertisement is shown at a time, affect results when the content itself remains the same.

Targeting. You can also test how different target groups respond to advertising, for example, by clearly directing the message of an advertisement at customer segments belonging to different demographics.

One advertisement might speak to young people and another to their parents, or one might be aimed at men and another at women. Through testing, you may uncover hidden target groups that were not initially considered the primary buyers of the product.

Display placement. If two or more displays are available, or if the position of an individual display can be changed, it is possible to test where advertising content performs best.

Read also: How do I do good retail marketing?

What are the benefits of A/B testing for digital signage system users?

Better ROI (Return on Investment)

A/B testing lets you optimise content and save money, because only the most effective content remains in use.

 

Targeted marketing

By testing different targeted content, you can find the messages that best resonate with exactly the customer group you want to reach.

Continuous improvement

A/B testing is not a one-time process – it enables ongoing development and content optimisation.

Testing is a continuous process – don’t rest on your laurels

A/B testing is what is known as an iterative process, meaning it doesn’t end once a winner has been identified from two versions. The results can be used for continuous optimisation: the most effective elements carry forward, and in subsequent tests, a different variable is changed.

A Digital Signage system provides excellent tools for A/B testing, as different content can be easily scheduled, and testing becomes almost automatic. Even a single store benefits from the results, and in chain retail, the benefits are even more transformative thanks to the larger sample size!

Explore our complete solution