You’ll have all heard about how Artificial Intelligence and Machine Learning has been affecting the industry over recent years. They aren’t just buzzwords, fads or trends. AI and Machine Learning is a revolution. At Google Marketing Live 2019 we saw Google talk about three revolutions within Search; Mobile, Machine Learning and Privacy. We’ve had the Mobile Revolution, and we are well and truly in the Machine Learning and Privacy revolution. However, generally within the industry you could argue that we are still seeing a lot of scepticism towards its adoption.
Here, we’ll explore some of the reasons for this as well as understand how humans still play a vital part in AI and Machine Learning. More importantly, we’ll hopefully convince you that in order to drive successful campaign performance you need to be embracing AI and Machine Learning, not rejecting it.
Why are some advertisers sceptical of AI/Machine Learning in Marketing?
It’s seen as a black box.
Google, Bing, Facebook, Amazon etc. keep many of the elements of it’s algorithms under wraps when it comes to AI and Machine Learning. At the end of the day, they are businesses competing; looking to drive the best experiences for their users and customers. If they made everything open and available to everyone, there would be no differentiation.
But for advertisers and businesses, there has been a feeling that the introduction of more machine learning has delivered less clarity and transparency. No one really knows the exact formula to how certain tools are making the decisions. What exact signals are being used to decide which creative should appear? Do weather conditions feed into these decisions? Does my location also factor into these?
These un-answered questions and have evolved into panic around “power grabbing” and feelings of having less control over marketing budgets.
It’s going to replace me.
This feeling of “less control” can sometimes lead to marketers being worried about the longevity and sustainability of their work. “Will by job exist in 5 years’ time?” is a fear that a lot of specialists will have. I even had one marketer telling me that they were no longer looking forward to hearing the latest innovations and updates from Google Marketing Live in fear that a new product will swoop in and replace them.
The key here is to better understand your position in this revolution, and how you should evolve.
It’s better at Bid Management, Measurement, Prediction. Some don’t like this.
AI and Machine Learning can now bid differently within every single auction in Paid Search, which is much faster than marketers can react. By understanding thousands of signals at any one time, Google can make better decisions about the user. Based on our restrictions we apply; it can then bid differently against that user dependent on how likely they are to convert, in real-time.
AI and Machine Learning have created unrivalled opportunity to better bid through more effective measurement/attribution and prediction techniques, which would be very difficult for humans to do by themselves. This technology is much better at finding patterns in big data.
Because the logic behind identifying these patterns are hidden within the ‘black box’ through a mixture of algorithms and signals that we may not have visibility of, it again can cause fear and scepticism.
Why should you be less sceptical of AI/Machine Learning in Marketing?
We built (and build) & taught (and teach) AI and Machine Learning
Many think of AI and Machine Learning in Marketing like some sort of robot from another planet; programmed to execute without logic. This is not the case. We aren’t going to get the results we need until we teach the system what’s right and wrong. We still need to tell the system what our targets are, how much we need to spend, on which demographic etc.
Another example is Google’s Responsive Search Ads. This is a tool we can use within Google Ads to find the best combination of Ad Copy for the user based on the data it has access to. Yes, this tool is great for Ad Copy testing; but we still give Google the assets. Google only finds the right combination for the user, we still provide the creative elements. To do this effectively, we must be creative with our messaging.
We are more creative with Ads and can resonate better with customers than Machines
The creative is what the user will make their decision on whether to click through to the website or not. In order to get a user to click through to the website, we must create attractive, engaging and relevant creatives. To do this, humans still must give engine’s these assets.
Machine Learning still isn’t great with linguistic nuances or a play on words, and so we still need to provide uniqueness through different approaches to try to appeal to customers.
Audiences Targeting can help us to be more strategic with our approaches
An area of search that has grown considerably in recent years is the depth of Audience Types that we now have access to, with many more to come this year. By being able to better segment our users both before and after they have visited our website, this allows us to be more creative with our messaging, landing pages, strategy and reporting.
By better understanding your potential customers, we can build better ways of spending marketing budget.
Humans are inquisitive, always looking for the next innovation
In 2017, Google changed the definition of Exact Match to include broader search queries. This was received with anger across much of the search industry as it felt like a needless exercise other than to give advertisers less control over which specific search queries we could target. The industry fought back and built scripts in order to help counteract this update. iProspect was well prepared for the updates through our best practice approach too.
These are examples of how humans can still work around some of the negatives of AI and Machine Learning, and I don’t think this will change. We will continue to build things to answer bespoke challenges. There will always be challenges for us to tackle.
Your business/client’s business is unique with its own nuances and challenges that AI might not understand
I think the best approach is to be technology agnostic; always start with the client’s challenge and then look to find the right tools in order to remedy this. Our challenge then is to build up the knowledge of these tools and work out ways to integrate these into client’s businesses and platforms for bidding and reporting.
AI and Machine Learning is only as good as the information that it ingests to make decisions. We need to be embracing its powers, but in order to do this we need to make sure that our client’s goals are measured effectively and fed into the machine. This is where humans come in.
AI and Machine Learning currently is great at managing one goal, but quite often a client may have multiple. For example, we want to achieve the greatest number of Sales at a certain Cost-Per-Acquisition, however we need to also deliver a certain amount of Revenue alongside this as a secondary KPI. Of course, success is better defined with one metric in mind; but let’s face it, it’s not always the case. The AI and Machine Learning will be able to achieve sales at a CPA, but it may ignore other factors to achieve this.
We still set the goals we need to achieve. We still set the targets that the client says we can’t over-shoot. It’s our role as advertisers to work with the client to make the best recommendations when there is no data, and to set expectations when the data tells us that the targets are not achievable. Managing these expectations is still a human job.
What does the future hold?
There is a need to be more transparent
There is need to find the balance between giving advertisers enough information that they feel confident enough to use the tools, and not sharing too much that it means that everyone will be able to “crack the code” which could create bad user experiences. We’ve seen how users are getting greater controls over privacy settings in Google, but maybe it’s time for advertisers to also get more transparency in return.
You are going to have to evolve, or become extinct
Embrace AI and Machine Learning when it makes sense. It’s better at doing a lot of day-to-day tasks that are repetitive and time consuming. By introducing this into your campaign management, it will not only drive better business performance but it will also free up your time to do the things that humans are better at; being personable, integrated, understanding, creative and strategic.
AI and Machine learning is one of the biggest revolutions to hit the industry, and I believe for the better. Humans will always have a part to play in this, you could argue that it’s bigger than ever. People + Machine Learning is what is going to deliver clients the best results. Those that stand still and continue to reject its benefits will be left behind. If you understand your position in the revolution and how it should evolve, you will find that you can drive better business performance and ultimately be more fulfilled.
- Responsive Search Ads available to all advertisers in Google Ads - March 2, 2020
- Google launches Continuous Audience Sharing for manager accounts - February 29, 2020
- Google Shopping Ads now available on Gmail - February 23, 2020