Testing Advertising Methods

[Mertie Oldow]

After collecting test market feedback, the data is analyzed and data-backed insights are utilized to improve the product, service, or marketing strategy. To collect, aggregate and analyze the data, reporting tools are often used as these are able to automate data processes and showcase results in easy to understand visual reports.

Note that MVT is a more advanced marketing testing method and is, therefore, most suitable for high traffic sites. Compared to A/B tests where traffic is usually split in two, in multivariate testing, it can be split in 3, 4, 5, and more groups.

So what does this translate to for eCommerce marketers? Well, your websites and apps are your real estate, so focus on testing those! See how easy it is for users to find a product/offer they want, how they interact with ancillary products/offers on your store, how they navigate to the checkout, how they fill out the shipping form, etc. Usability testing can also work wonders for testing loyalty programs.

There are several usability testing methods at play. Most commonly usability tests are split into: moderated vs. unmoderated; remote vs in-person; and explorative vs. assessment vs. comparative. The combinations go from there as you can have a moderated in-person assessment test, an unmoderated remote comparative test, and so on.

Again, the objective is to measure user experience: content testing is not to determine whether users like your content, it is to determine if they can read and understand it. A good example would be the user manual for assembling a piece of furniture: you can test if the text is legible, written accurately and whether it provides the information necessary to use the product.

Incrementality testing falls under the A/B testing umbrella. However, whereas A/B testing is used mostly on creative elements and measured by click-through rate (CTR), incrementality testing focused on sales uplift and is measured by conversion rate (CVR).

Nevertheless, conversion rate optimization (CRO) is just one side of the coin. If you really want to see significant outcomes of user testing, you have to employ these practices across multiple user touchpoints. In other words, you must test the entire user journey, from initial marketing channels, all the way through to how potential customers navigate your website.

For that purpose, marketers exercise customer journey mapping. This allows them to not only develop empathy for the customer, but also understand consumer behavior patterns that influence conversions, loyalty, and a number of other business metrics that are directly connected to revenue.

Regardless of the type of market test you are carrying out, the most important thing is that you are conducting tests, collecting copious market research, and thus, gaining access to all of the valuable insights and data that can better inform your strategy.

For the best results, it is clear that the holistic, 360 degree, all-encompassing, customer journey map reigns king. With high-impact benefits such as higher conversions, loyalty and revenue income, the marketers that can include this potent weapon in their arsenal will ultimately reap the best rewards and edge ever further in front of the competition by better understanding their target market.

This article is an introduction to advertising testing. It includes several different methods of advertising testing, the types of advertising that can be tested, key metrics to measures it against, and some tips for successful and impactful advertising testing.

Advertising testing is a crucial step in creative development. It is a way of predicting the likely success of an advert or campaign, often before it is live (although not always), by reviewing whether it will meet objectives among the target audience.

Dependent on the method, it can also provide valuable feedback on anything that may need to be tweaked or improved to ensure the best impact. Additionally, if the ad or campaign performs well, it provides much needed confidence in the creative direction.

With this method you recruit a number or research participants to take part in a survey. Participants are shown a creative (or number of creatives) and then asked a series of follow up questions to understand how they respond.

This method splits out your sample into matched groups (cells), and shows the participant in each group one creative. You then compare the results between the groups to see which creative performs better on key metrics.

This is a fair way of asking about each advert, but requires a larger sample to reach a robust base size for each group. Also, you will need to ensure that the groups are matched on any key attributes, to ensure that any differences between the groups are because of the creative and not because of personal differences (i.e. different genders or age groups may respond differently to an advert).

This method presents an advert to a research participant and asks them follow up questions. When they have finished responding to one advert, they are they shown another. The adverts should be randomised in order, so as not to bias one advert over another.

This method shows two or more adverts simultaneously and asks respondents to answer follow up questions with them side by side. This creates a trade off scenario, where research participants will select one advert as more x than the other(s).

Again, this is a more efficient use of sample, making it cheaper. However but if adverts are similar, the results are not likely to vary much, making it challenging to find a winning creative. It can also feel more laborious to take part in, dependent on the number of adverts being tested. Finally, it may not draw out specifics or nuances, as participants will be less likely to give all creatives full attention.

Two (or more) adverts are launched at the same time, to a smaller, matched audience, and you monitor which has more engagement. Digital advertising is particularly well suited to this approach, as you can track CTR or conversion rates.

Biometrics are used to understand physiological reactions to stimulus. There are range of different approaches (you can learn more about them here), but the most common of these are eye tracking and facial coding from expressions. These can be really useful tools for advertising testing, as they move beyond what people say they think, and get to the heart of subconscious feeling and behaviour.

Eye tracking is useful for all visual stimulus. It monitors where people look and for how long, to understand attention. To learn more about this approach, read our introduction What is Eye Tracking for Market & User Research? We also have a great webinar from Lumen that explores How to Win in the Attention Economy, which you can watch on demand here.

This approach is great at seeing what elements of a creative are standing out and grabbing attention, and can be benchmarked against other adverts in the same media or category. However, it is best when coupled with a degree of questioning to understand what has explicitly resonated, or to see what brand or messaging was recalled after exposure to the advert.

You can test messaging, copywriting, images, videos and audio. You can test an initial campaign concept, mocked up ads, right through to finished adverts. You can also test a full range of media types; TV, press, digital, radio, OOH, social, even door drops.

Just remember that if the ad is unfinished, you need to inform the research participants so that they are answering based on what they predict the finished result will look like, rather than telling you it looks amateur.

Before you start, make sure you have it clear what you are trying to achieve with the campaign. Are you trying to build brand awareness, are you trying to drive a call to action? Making sure these are as clear as possible means you are able to pinpoint whether an ad delivers on these metrics.

This may seem an obvious point, and it aligns with clear objectives. If you need to understand what a specific segment think about your ad, make sure you have enough of them in your sample to be able to get a read. Also make sure that you ask the key questions to define them (such as segmentation questions). If you have multiple target audiences, make sure the research design allows for all of those groups to be analysed, which may require boosting sample or setting quotas.

It may be that the ad you test is perfect and ready to go, but it is very likely that the research will unearth some small (or large) tweaks that it is advisable to make, so try to give yourself some buffer time between receiving results and fully launching the campaign to implement any necessary changes to ensure maximum success.

Also, consider at what point to test. Do you need input at concept stage to drive development, or is it more useful to you when you are closer to a finished advert for final sign off and tweaks. In some instances you may want several rounds. Try to plan where you can.

Ad testing is the practice of evaluating advertising effectiveness, to see if your ads are reaching their intended audience and achieving their intended goals. Ad tests can be done through various methods such as market research surveys, A/B testing, focus groups and data analysis.

You can test different aspects of an ad. You can zoom in on the copy to find the snappiest words, test media placement, images and colors to see what stands out or reflects your brand best, or simply test to see if it moves your target audience the way you want it to.

Additionally or alternatively, you can use a focus group to dive deeper into your survey results if needed. This can be particularly beneficial if your survey is inconclusive, or if you want to gather more knowledge on responses.

If you want to keep the research going, you can create different versions of an ad and run them alongside each other, to find out which one is the perfect ad. If the survey results were clear, however, you can just go ahead and launch the ad or campaign that came out as the winner.

c80f0f1006