Measuring the performance of your radio campaigns is unfortunately not as straightforward as TV, but it is possible. Radio, unlike TV, is not a direct response channel – your customers aren’t likely to download your app or visit your site directly after they hear your ad. This makes it nearly impossible to accurately attribute visits to individual spot airings, which in turn limits your options for optimisation by station and timings.

But there are still many ways to get some significant and informative performance metrics together. For instance, you can still measure KPIs like traffic, unique visitors, and conversions, but it won’t be as fine-grained. There are several strategies for measuring radio performance, though some are decidedly better than others. Here are some we recommend, and some we don’t think work so well…

Maybe this is too far along (or this article way too long) for you. Check out our guide on whether radio is the right channel for you. If that checks out, then we also have a guide on executing a kickass campaign.

Our favourite:

bayesian methods are great for measuring radio performance

Bayesian methods

Bayesian statistical models are currently the most refined method for measuring the effectiveness of a radio campaign. The model tries to predict how a KPI (like website traffic) would look without a campaign running, and then compares it to actual traffic data. It can be as complex as you need it, including information like seasonality (e.g. weekday vs. weekend), underlying long-term trends (e.g. organic user growth), and the effects of other marketing campaigns. In the end, you will get a clear picture of when your user behavior has changed, whether that is a result of your radio activities, and how large the effects are. 

The major benefit of this method is the relatively high level of accuracy. Also, you don’t need much data to run an analysis – daily visits generally suffice (although the more data you can include in your model, the more accurate it will be). This is excellent for radio, where you have a delayed response of a few hours or days after your customer sees a spot until they actually visit your website or download your app. 

The drawback of this method is, it’s very labour- and knowledge-intensive. You need both the understanding of statistics to run the analysis, and deep domain knowledge about your industry and the market to identify the external factors (also known as priors) to include in your model. This may be a topic to outsource to an agency or other marketing service provider or even… Us! We have the method, can estimate your campaign’s effectiveness for you and help you in interpreting the results.

Middle of the pack: 

Baseline uplift

To get a general idea of whether your radio campaign is leading to new visits, you can start by using the baseline-uplift method. The basic approach is to measure a baseline of traffic, for example in a similar city where you are not running a campaign, and compare it to the traffic in the area where the campaign is running. 

This gives you a very rough picture of whether radio is effective at driving traffic, but does not account at all for effects happening outside of your campaign – like long-term user base growth, seasonal effects and other marketing efforts. This often leads to over attribution of traffic to the campaign you’re measuring with this method. 

User poll

A simple but effective way to find out if radio is working for you is to include a poll in your site or app to ask users the question – “how did you hear about us”? 

The insights you can glean from this depend very much on where in your user acquisition funnel you place this questionnaire. The earlier you place it, the more you’ll find out basic information about which channels are driving traffic. The later you place it, the more you’ll discover which channels are generating high-quality users – for example if you ask the question close to checkout or purchase. 

Quick tip: You can run small A/B tests to see how deep in the funnel you can place this kind of poll, and how many questions you can include, without damaging your conversion rates. 

Not recommended:

Voucher codes are not recommended for measuring radio performance

Voucher Codes

It may be tempting to use a discount deal with a special code to track users like you would with an out-of-home or print campaign, but remember you are already trying to get customers to remember your brand, value proposition and CTA all within a 20-second ad. Since they likely won’t respond directly to your ad, the chances they’ll remember a voucher code on top are slim. 

Test, learn, scale: How to take your next campaign up a notch

As with any good growth marketing strategy, the first campaign is only the beginning. The goal is to generate enough insights to make qualified decisions on where to invest your marketing budgets. Is radio the right channel for you? Can you scale it? How can you improve for the next campaign? The answer to all those questions could be: yes!

Making sense of the data 

First, you’ll need to look at the data you’ve collected and have an eye out for your CPA (cost per acquisition). Let’s assume you’ve run your first test campaign on radio with all other channels remaining, and/or you’ve used Bayesian methods to control for other campaigns and external factors. In the best case, you’ll see your conversions increasing and overall CPA remaining the same. This means radio is the right channel for you, and you can scale up your investment.

This is only the most ideal case, however – in all other cases, you’ll need to analyse the data further to figure out whether radio is not working at all, whether it might work in combination with other media, or whether there are other flaws in your user funnel. This is where having multiple measurement methods in place, or even investing in additional market research and insights, can be crucial. 

Iteration and following campaigns

If you’re one of the few where radio works as a stand-alone channel – congratulations! You can scale up your investment. But it’s rare that the results are this clear, and radio generally works best as either a regional media, or in combination with other marketing channels that support different parts of your user acquisition funnel. 

Cross-media campaigns, in particular adding out-of-home (OOH) to the mix, can strongly magnify the effects of a radio campaign with a similarly regional focus. It’s an opportunity to increase your touchpoints with your target customer on a channel they interact with differently, and which supports brand recall and response. 

At a certain point, it’s worth considering whether adding TV could be the next logical step – especially if you want to roll out a marketing campaign nationally. TV has more flexibility for storytelling and presenting your CTA, and you can clearly attribute direct response to individual spot airings. Here, you’ll be able to optimise in a more granular way and attribute conversions directly. 

In short, radio can be a conversion driver and has a strong place as a tool for regional promotion and testing. We love radio!

We know the airwaves like the back of our hands. Get in touch at hello@dcmn.com if you’re intrigued by this classic, but vital medium.