When you get served up an ad online, it has been delivered based on several different data points. Here, smrtr CMO Paul Argus [pictured] sheds some light on targeting and retargeting and explains how data can be used to find new customers.
When it comes to digital advertising, the terms ‘targeting’ and ‘retargeting’ are often used interchangeably. However, there are subtle differences between these two approaches which could make a big difference in the overall effectiveness of your campaign.
The basic difference comes down to the audience each approach is aiming to reach. When targeting, marketers are looking for ‘cold’ audiences (meaning an audience that has not interacted with your brand recently), while retargeting is aimed at warmer audiences that might have visited your website recently.
The logic behind retargeting is relatively straightforward. If a potential customer has interacted with your brand they are more likely to make a purchase. This can even apply to ‘hot’ audiences, where customers have shown heavy engagement, such as adding an item to the cart before exiting. These customers are deemed the most likely to convert and will therefore be the most receptive to marketing campaigns.
The downside here is that as customers are quite late in the buying journey by the time they are retargeted. In some cases they may have already made a purchase. This, combined with the fact retargeting focuses heavily on the use of third-party cookies – which are set to be discontinued in the coming months – means this technique is set to become less relevant for marketers in the future.
Targeting
Targeting, meanwhile, is less straightforward of an exact science. Given customer signals are not as clear, marketers must draw on other factors to ensure they are messaging relevant audiences. When done well, however, targeting serves as an effective tool for finding new customers and making marketing spend more effective.
For a long time, this style of online targeting has been built around digital identifiers, particularly third-party cookies, which track how someone navigates the web. While this has been shown to be effective, there are also limitations. Cookies only track devices, meaning the ‘offline’ (and real) world is left largely unaccounted for.
But by using data – whether it be first, second or third party – marketers can start to map real-world interactions and target potential customers based on their behaviours and characteristics. A great example of how data can be used to target new audiences is when promoting an upcoming movie. Firstly, transaction data can be used to model individuals more likely to visit the cinema again or other similar films that might be out at that time.
Additionally, online signals can also be added. If the movie is starring Chris Hemsworth, for example, marketers can target audiences that have shown online interest with Hemsworth, such as following him on Instagram. By combining this online and offline data, an accurate audience can be created and the marketing spend can be allocated mainly at individuals who are likely to buy a ticket.
Similarly, psychographic signals – which relate to characteristics like personality, hobbies, values, beliefs and interests – can be combined with online data to target interested customers. Patagonia has demonstrated the power of this in recent years, through realising its outdoorsy customers are highly environmentally conscious. This has meant the brand can start to appeal to people who demonstrate an active interest in protecting the environment.
These signals can be collated via an Identity Graph. Identity Graphs are databases which show customer profiles and all known identifiers attached to these customers in a privacy-compliant manner. The smrtr Identity Graph uses a wide number of opted-in, privacy-compliant datasets to connect different forms of Personal Identity. Importantly, as we enter a ‘privacy first’ based web, online marketing will become less about targeting a single individual (device) and take a more aggregated segment based (people) approach. While this does limit the accuracy of the targeting given assumptions are made about the people being grouped together, the advancement of analytic techniques and data availability mean this is minor. Additionally, predictive segment based targeting opens up targeting at scale and also reaches people before they make a purchase decision – for so long the achilles heel of re-targeting.
At smrtr, we believe the true picture appears when multiple data points are connected together. We combine online and offline signals to help our partners target new customers by taking a longitudinal (rather than rear view mirror) approach. We are prepared for a world without cookies and use the aggregated data of over 16 million Australians to help build audience segments that work.
