Getting started with AI: Initial steps for marketers

In a few years’ time, AI will be so efficient in providing information about consumers that very little marketing will be based on guesswork or intuition.

That’s the prediction of analysts gauging the capabilities of machine learning and AI combined with the vast amount of data becoming available through the IoT. And most marketers agree: A recent Demandbase study found 80 percent of B2B marketing managers at firms with more than 250 employees believe AI will revolutionize marketing by 2020.

That’s not to say the transition to the AI world will necessarily be easy, short and cheap. But marketers’ ability to start optimizing its power could make or break the success of their firms moving forward.

Here are a few steps to consider if you’re in the first stages of planning to optimize AI:

1. Familiarize yourself with the general capabilities of AI across industries. PC Mag offers links to a number of online tutorials here.
2. Build MVP version first. “Go for the quickest wins first to prove value early on and win buy-in across the board”, advises Andy Betz on Martechtoday.com
3. Think about the primary problem you’re trying to solve or facet you’re trying to improve. The last thing you want is a complex new system that has all the bells and whistles but ultimately doesn’t improve the crux of your marketing plans. “Having so many different tools at our disposal can present challenges and unnecessary complexity,” notes Betz.
4. When in doubt, choose the solution(s) most likely to improve your customer journey, since your ability to make that journey seamless and highly customized is going to be crucial in order to compete.
5. Analyze who in your organization will be accountable for development & implementation, and how they will go about that. Do you have the talent available, or would you be better off hiring a consultant who can get you started? What is the probable time frame for learning how to optimize the tool and teaching it to your relevant staffers? Two to three months may be a good interval for a pilot project, according to Luke Tang in the PCMag article.
6. Be aware of how you will quantify the added value of the tools you choose so you can present results to the rest of your company and use that info to inform further buying decisions.
7. Once it's launched, continually evaluate the performance of your AI tool, and take another direction if it isn’t significantly impacting results.

"2018 will be the year another seismic force begins to take hold and transform marketing and advertising," writes Reggie Bradford in Forbes. "AI is coming in every field and in every business, like it or not. Its effects in marketing, if understood and harnessed correctly, will be game changing."