The ultimate aim of digital ad campaigns for any online brand is to serve each unique user with products that they will want to buy or ads that effectively raise the brand’s awareness. The better targeted the ads, the more likely the user is to interact with them.
When setting up ad campaigns, many companies have relied on targeting tools such as their pre-existing assumptions about their customers in order to create target groups for advertising messages. Of course, businesses know an awful lot about who buys their products, and these targeting tools and data are useful. However, making strong assumptions can also lead to bias, which can eliminate valuable potential customers from the campaign.
In this article you will learn about:
- The importance of budget optimization
- Good practices for effective budget planning
- The future of advertising in regards to budget planning and optimization
Table of Contents:
- Contextual targeting removes the need to segment campaigns in advance
- Understanding bias
- Understanding buyers
- Understanding by us
- Read the Auto Industry Digital Advertising Guide 2021 to find out more
Contextual targeting removes the need to segment campaigns in advance
We’re going to explain how contextual targeting as a targeting tool removes the need for campaign managers to make assumptions about target groups. We’ll start by looking at a few vehicles that might typically be marketed to one group of buyers.
Female buyers account for around 60% of Fiat 500’s sales. Indeed, insurance data suggests the figure could be as much as 73%.
A 2020 study of car insurance quotes crowned the Vauxhall Adam as the most female-biased car in the UK.
Women are more likely to buy electric cars than men. The Renault Zoe is a compact and popular hatchback EV.
*Information and images taken from buyacar.co.uk
Let’s look at two assumptions. Firstly, a user views a Fiat 500 online. The first conclusion we might draw is that the user is a woman. However, if sales figures show that 60% of buyers of
that specific model are women, that means we would probably be wrong 40% of the time. That’s not much better than flipping a coin.
Secondly, coming from the other side; if we know that the buyer is a woman, we might reasonably assume that she is interested in a compact hatchback, perhaps an electric vehicle. However, female car buyers are also keen on SUVs, sedans, and even small utility trucks like a Toyota Tacoma or Chevrolet Silverado—all of which are very different from a compact hatchback. Again, if we were taking a guess on audience targeting, we have a very large chance of making a false assumption.
In the data-driven age, there really is no need to be making false assumptions when it comes to audience targeting…especially when it is so easy to have an open mind about who may be interested in our products.
Of course, car manufacturers should create advertising that appeals to core buyer groups. It makes sense to use an aesthetic that appeals to female buyers when creating ads for the Vauxhall Adam, Fiat 500, or Renault Zoe. However, we no longer need to pre-program marketing campaign algorithms when targeting clients in all circumstances.
It can be useful to include demographic, geographic, or even business-specific data in brand awareness campaigns, but the most effective solution when running consideration and performance campaigns is to get rid of profiling altogether and look at what users do online – not who they are.
If a 50-year-old man and a 22-year-old woman are both looking at a Harley Davidson bike online and display the exact same browsing behavior, then a good algorithm will probably rate them as equally likely to make a purchase, while a human will still hold some bias and assume that the older male is the stronger prospect.
Understanding by us
There are two key elements to our technology that allow us a better understanding of buyers and their level of purchase intent
Proprietary technology from RTB House which uses first-party data for understanding buyers. With no guidelines on specific demographics to target, ContextAI learns from behavior in the browser and displays ads based on predictions, not biases.
It’s important to note that we don’t use any demographic data in our targeting, which makes for less assumptions today, but also puts us at a significant advantage as we approach the cookieless future. We are already using behavioral targeting, not relying on data harvesting and putting users into boxes.
You can find out more about how we are ready for the cookieless future in this quick video about user groups and targeting without third-party data.
Deep Learning is a huge step up from Machine Learning, which is itself a more advanced form of Artificial Intelligence and allows us to utilize cutting-edge tools such as contextual targeting. For every 80 calculations made by an ML engine, DL will make 200,000.
As the first DSP to implement 100% Deep Learning-powered technology, we run campaigns which build an audience in real-time. Deep Learning systems learn and adjust as time passes; they find the patterns in customer behavior and serve them with the right ad content through methods such as contextual targeting.
For vehicle manufacturers, this means that profiling is taken out of these automated marketing activities. You can focus, for example, on female buyers when designing vehicles and building your overall marketing strategy. However, when it comes to programmatic ad campaigns, you only need to set your KPIs and then let the Deep Learning algorithm take the wheel.
Read the Auto Industry Digital Advertising Guide 2021 to find out more
We are experts in advertising and audience targeting for the auto industry. If you want to read more about our thoughts, check out the Auto Industry Digital Advertising Guide from 2021 and see what the future holds.
Good to know: How AI Helps in More Efficient Video Advertising