Art generators such as DALL-E2 and the text generators like ChatGPT, captured our imagination while also raising some hackles around the potential of deep fakes, plagiarism and other legal conundrums.
While the nuances of the ethics of AI generated content and art will be debated for years to come, the more commonly used branch of this tech, machine learning, has been embraced by businesses to excellent effect.
In digital marketing, a field already dependent on data, machine learning–powered software now allows brands to holistically manage their organic and paid search.
While paid search advertising accounts for a sizable portion of digital marketing spend, marketing leaders have always struggled to quantify the relationship between paid and organic search and their effectiveness.
New technology now allows brands to use machine learning to assess many thousands of internal and external factors that impact paid and organic search, determining the value of each as well as their correlation.
By identifying their true organic reach CMOs can make informed decisions about their paid search. This would include burning questions like: whether to bid for keywords that are already receiving high organic coverage; to what extent paid strategies are cannibalising organic strategies; whether to pay for brand and brand generic terms; as well as informed positions based on data from current organic coverage.
Programmatically managing paid and organic search based on analysis of both channels, along with the power of self-learning, has seen tremendous results. In fact clients using it are, on average, seeing a 25% revenue uplift and a 20% boost in efficiency.
Using algorithms to better target customers is another place machine learning is flexing its marketing muscles.
Dutch low-cost airline, Transavia Airlines, has tapped into the power of algorithms and machine learning, making use of live weather feeds to change up its content to be more audience appropriate. For instance, if the Netherlands is experiencing a cold spell, they will profile warmer destinations for getaways. They have also used snowfall updates to profile great skiing opportunities. On the flip-side, when the local weather is great and the whole of the Netherlands is revelling in the sun and not considering alternate destinations, paid media spend is programmatically scaled back to reduce wastage.
The airline understands that no one buys a ticket to sit on a plane for nine hours. They know that what they are really selling is the lure of the destination. So, by gearing their content to sell the idea of the holiday, they are appealing to the real motivator for customers. More than this, the digital team has also integrated real time seat availability into their pricing – all of which is dynamically fed into their banner creative. This machine learning engine and more appropriate content approach has increased bookings by 45% by utilising automation to create a more connected user experience.
While machine learning services have been eagerly adopted by European clients, South African brands hoping to make use of the technology will need to address the elephant in the room when it comes to skills to support data-driven solutions.
Remote working has seen many international players shopping for skills in local waters and, as a result, already scarce data specialist skills now command salaries between 20 and 30% higher than 2021, and 2023 will see the trend continuing. If local brands really want to tap into the power of machine learning (and even AI), in coming years, they will need to start building a talent pool now.
More importantly, it makes excellent sense for brands to find partners which understand the powers of technologies like machine learning and which can work with their clients to assess, understand and leverage their internal and external data opportunities. The brands who stand to win big in 2023 are those which are willing to try new things, have an agile mindset and who have a test-and-learn approach hard baked into their approach to all data opportunities.