On the surface, the way we manage and strategise advertising is unrecognisable from the earliest forms of mainstream marketing. It is run - and greatly enhanced by - technology and is enriched by the availability and application of extensive data and analytics. But in reality, the challenges remain exactly the same as they did 150 years ago.
In the 19th century industrialist and marketing pioneer John Wanamaker delivered the now infamous quote: "Half the money I spend on advertising is wasted; the trouble is I don't know which half." Today, 100 years after his death, much of the technology we use in marketing is still trying to answer that question effectively and consistently.
In 2021 it's all about audience segmentation, analytics and insights, different channels and format, brand safety optimisation, or a business' own bespoke technology. Like us, many agencies are working across both national and global clients, addressing a multitude of different challenges and strategies for both on and offline businesses.
So how do agencies with such a broad breadth of offering ensure all the individual elements of advertising technology are working seamlessly to make the marketing they deliver better and not simply more complex?
Firstly, we must be aware of the growth of data silos within different technology providers. The likes of Google, Amazon, Apple and Facebook operate not only their own extensive data but they also own the media and the measurement. With these providers also setting their own pricing, and decisioning media buys based on algorithms, we are very close to the old 'black box' media network days. While it's not always possible to break them down completely, what we increasingly need to do is use data science to understand what's going on within the silos rather than blindly taking the metrics they release as a single point of truth. If you can't measure the results… don't believe them!
Another thing to be aware of is the rapid rise in automation. It's an incredible concept but certainly warrants a little caution from those of us who are operating it. We're all very keen to get computers doing jobs quicker and more efficiently than we can them ourselves. And while I'm not predicting a Terminator-style AI takeover any time soon, I do believe there is a risk we are unwittingly seeking perfection without understanding the tools and how they operate.
AI and automation is programmed to complete a job above all else, and so will take on its own standards of what 'perfection' looks like. It will do anything and everything to achieve the goal it has been set. From our side, we must be very specific about the goal we want the technology to achieve - the one that's most valuable to us - and make sure we set any parameters that also need to be accounted for. Viewability was an important lesson to the industry on this. If we don't tell algorithms that a Post View conversion is only valuable if the ad was seen in the first place, we end up with low viewability and wasted budgets, driven by algorithms that believe they are getting closer and closer to success. This means we need to be quite sure of our measurement, and I would suggest 'attribution' of success should be increasingly balanced with a drive to 'incremental'.
Leaving automation to one side, we still have some areas of human focus for us marketers.
The human touch
The first is in the creation of content. Having that original, impactful content to feed into your marketing machine is imperative to success. Content is not data, it is not real time, it is not technology, it is something nuanced and creative and where the buck stops in our quest for marketing success.
If we want great results from our marketing, we need to ensure we are feeding the right ingredients into it at the beginning. Though with an exciting rise in technology within this space, I expect there to be exciting changes to how we deliver and measure success within content.
The second is in the identification of anomalies. When it comes to data and measurement, never underestimate the power of the anomaly, as it is often the most actionable data point. As we round up data across publishers, channels, campaigns and territories, we have a tendency to flatten out the data into averages. This is useful for a snapshot of what is happening but causes real issues when it comes to understanding why. Knowing a certain publisher is outperforming on a Friday afternoon may lead us to learn something that can influence a keyword list in another territory. Data is great but explore it and ask questions. While machines can often optimise very effectively, if we lose the learnings then we can't invent and transfer that knowledge to other strategies and business areas, which an algorithm may not be able to do.
Cut through the hype
Finally, no overview of the challenges of martech would be complete without a nod to our old friend, the cookie. We have become a little used to 'buying audiences', without understanding exactly what that audience really is. As third party cookies disappear, there is going to be less data available out there, so now is the time to understand our own audience past a set of cookies. Advertisers also need to take a user-centric view of what their brand experience is like for their individual users, with a view to a fair exchange of data for experience.
Above all else, the single, overarching takeaway must be that advertisers and their agency partners are focused on ensuring all this brilliant technology, its providers and their partners are working in a way that achieves something that is beneficial for their advertising, and ultimately business. The value of any technology is only insofar as it can be applied to your needs or challenges, and correct use of technology is reliant on you understanding those needs. This will help cut through the hype, and make technology partners work for you, ultimately benefiting both parties.
Copyright 2021. Sponsored post by Charlie Ashe, Head of Client Data and Strategy, tmwi