Confessions of a Buzzword Junkie

By Maria Flores Portillo, Persado VP & GM, UK

I only realised this recently, but I have a small confession to make: I am a buzzword junkie. Looking back at my career, there is one common thread: every time there is a new-new thing, my brain polarises towards it and tries to make sense of whether it’s just hype or something that is solving a real need. Often, I've followed them full time.

First, it was mobile. In the pre-iPhone era, every next year was the “year of mobile” until Steve Jobs actually changed the world as we knew it. Around the same time, programmatic advertising was growing and soon created the holy grail of marketing: target the right person at the right time by leveraging those harmless cookies we were happy to leave behind (who would have thought about GDPR?). Then came online video and now, we have reached the age of Artificial Intelligence. Once a subject of Sci-Fi movies and futuristic fantasy, the solution is now a current reality and the buzziest word in marketing.

And that’s the good thing about following shiny little objects — eventually, you get to touch them, unmount them, look at their ins and outs, meet real experts and get invited to all sorts of events that try to demystify the hype. As a trend-lover, I’ve been in the trenches of the AI revolution. These are my insights from the inside.

Is AI All Hype?

I’ve spent my entire career in tech and people have never focused so much on the underlying technology itself. On a recent meeting with a prospective client, a marketer actually asked me if we used neural networks. Seriously? I cannot recall a time when someone asked me if I was using time division multiplexing when I was selling mobile indoor networks.

The truth is, the greater focus should be on the way the technology is applied. But the underlying tech itself is not as esoteric as it sounds. There is no Terminator-like future in sight. To date, AI is primarily a way to process data in a very, very smart manner.

My answer to the ever-popular question, “Is AI all hype?”: Stop asking that question.

Get over the hype accusation. AI is here, and it’s going to significantly transform our experiences and workplaces. Google it.

The No-Brainer Side of AI — and the Interesting One

Now, let’s discuss how marketers can apply it. Since AI always evokes all sorts of images of androids, I tend to use the human brain as an example of the two main advantages of AI:

Our left brain*. It takes care of all of the logical pieces of our thinking. It calculates, associates ideas in a structured way and strives for efficiency. That’s where AI has started. It’s a very powerful tool to bridge that gap between data, insight and action. For marketers out there, it resolves the pain of rule-based marketing and should prevent all of those highly paid data scientists from spending their time scrubbing data. This is the part that will unavoidably transform organisations, as it replaces human work and creates huge efficiencies. It is good news, even if change is always a bit scary. We just need to make sure we skill humans up fast enough to deal with the new reality: one where algorithms make some things faster and better than us. New positions such as training algorithms or QA’ing outputs will emerge soon.

Our right brain. Where the magic really happens: creativity. This is where we come up with original ideas. And this is the interesting part of AI: Because of the way that algorithms evolve in their learning, it is possible for a computer to come up with things or conclusions that we couldn’t come up with, even when it comes to music and art. It may be possible for a machine to detect our emotions or our brainwaves to adjust how they interact with us or serve us in a personalised manner. It is still early days, but this is why I got to the field: the ability of machines to make better things.

How to Evaluate AI and Drill Down a Vendor

What does your organization need to adopt AI? To start, it needs a purpose, and commitment.

1. Identify the problem to resolve. It could be supply chain efficiency, marketing automation or customer service call times.

2. Identify how much solving that problem is worth. This will help get the right people excited and willing to implement change. Typically the main bottleneck.

Once you go into evaluation mode, seek to understand the following:

1. First-party data requirements: Look at the state of your data. Chances are, you may need some work there, especially if it’s stored in several databases and formats that don’t sync. Sometimes, your AI vendor can take care of that, but you may need to get to some compromises.

2. Where else the vendor is sourcing data from: AI cannot work well without Big Data. Check the quality of any third parties.

3. The integration requirements with your existing systems and third parties: The AI company’s approach (algorithms). If the vendor does not explain this well, it’s a red flag.

4. Whether AI outputs are QA’d (they need to be): Ask hard questions on maintenance and human supervision

5. In case of emergency: Is there an abort button if needed? What are the consequences of shutting it down?

AI is a special type of shiny little object, for it has the power to transform our civilisation in ways we have not started to comprehend. Things will not change overnight, but brands and individuals need to open their eyes and look for ways to adapt and leverage AI. If the Internet changed our lives through unprecedented access to information and material things, Artificial Intelligence has the power to redefine (and expand) human potential.