Artificial Intelligence and Other Adventures — in Marketing
“Can AI take care of building my creative? Can it set up targeting? Can it read the content of my ad to predict what kind of audience I should go after?
Come to think of it, now that we can look for weather information, quick tips on cooking, traffic updates and movie insights by just turning to Alexa, how quickly can we create an Alexa skill and build some voice ads?”
Before we attempt to answer the above questions, there’s a fundamental one that needs to be looked into:
Is it really true that new age technologies like Artificial Intelligence are set to revolutionize the corporate world in general and marketing specifically?
In broader terms — from the perspective of making our work more accurate, supplying the right kind of data, speeding up processes that are otherwise slow, and uncovering insights we might otherwise overlook — AI has indeed, overall, helped to bring in more sophistication and ease in our everyday jobs.
But, there still is a lot of hype around the usability, future and realistic expectations around how and where AI can be implemented.
No surprises there, that the search trend for the past 5 years shows that curiosity around ‘AI in Marketing’ is on a slow, but sure incline:
But there’s a more interesting study by HubSpot, which showcases the Hype vs. Reality on Artificial Intelligence in general. Ranked at 7th most difficult to understand as a concept, 20% of internet users also consider it as being one of the most overrated technologies out there.
Further, a March 2017 survey by WBR Digital and Persado found that 76% of US and UK retail marketers said there was confusion or lack of clarity about what AI marketing could be used for, while 59% said there was trepidation and lack of trust in the technology.
Organizations in general, face this issue where finding the right ways of implementation, lack of clarity + skills and overall business case identification are huge factors preventing them from using the potential of AI to the fullest.
Interestingly, in a 2016 survey by e-marketer, a significant number of marketers were hesitant to adopt AI, because they fear their own jobs may be at risk. This has seen a substantial shift in a positive direction where in a 2017, Capgemini survey 83% of AI implementers said that the technology had created new job roles within organizations, which goes to say AI is more likely to ‘Augment Human Output’ than put people out of work. That’s good news!
However, as marketers what can we do to avoid pitfalls while implementing the technology? And is AI all that it claims to be?
Let’s address the first part :
1. Understanding that AI is More than a Shiny, New Toy:
A lot of time is spent on testing, trying and piloting new AI based technologies which have limited to no business impact. It is important to define the objective and then evaluate technology as a means to achieve the outcome more effectively and efficiently. Circling back to the question in the beginning, we need to move away from purely asking questions like “how can we build a voice ad for Alexa” to “how can I generate more traffic?” or “how can I drive more sales?” And subsequently evaluate if Alexa is indeed the right way to do so.
2. The Perils of Vendor Hype:
Take the time to understand the technology, the promised results and most importantly the fine-print. Now, don’t get me wrong, a lot has been achieved by AI companies, but in an effort to sell and stand-out the final results and claims are mostly idealistic or “in progress”. I’ve personally seen examples of companies touting to “dynamically build and optimize creatives with AI”. Under close inspection all the vendor did was test and optimize close 80+ versions of humanly prebuilt creatives. Be sure of what to expect and invest in evaluating the loopholes before parting with your money. With a lot of companies jumping on the AI bandwagon, it is crucial to be discerning in your choice.
3. Being Tech-Literate:
There’s a lot of ambiguity around what’s AI, and what is just data analysis; what a machine is capable of and how much is human led. Having the right in-house knowledge of what’s in the market and what to expect from AI needs to be coupled with a closer understanding of how the vendor’s technology works. It’s sad, but true that a lot of companies are misled into investing in technology they don’t fully understand.
4. Bringing Human and Data Angles in Tandem:
Data, data, and more data — get ready to work closely with various stakeholders when it comes to data management. It is critical to understand that the better your data the more effective your AI tech implementation which subsequently will lead to a lot more data generation. Storing, managing, keeping a track of and most importantly driving results from all this data will require massive collaboration and buy-in from various stakeholders. Not to mention a possible rethinking of data management systems.
5. The Smart vs. Creepy Conundrum:
As a brand and as a company, have a point of view on AI. Evaluate its implications and the specific use-cases for your customers. One slightly experimental implementation can cause your brand to go from “smart, caring and relevant” for your consumers to “creepy, scary and unethical”. Don’t be afraid to experiment, truly evaluate your stance and then pick your battles.
And finally, is AI all that it claims to be? In the true spirit of corporate diplomacy, the answer is No and Yes.
If you dream of one day effectively outsourcing creative thinking and human intervention — Needless to say that’s far from reality because AI is designed to speed up systems for humans and not quite do away with human thinking, creativity and innovation altogether.
And, yes because AI implementation scenarios are vast and it will pleasantly surprise you with its efficiency, nuance, and ability to pick up details that humans would otherwise find difficult to process.
So, keep exploring and happy adventures!
