Artificial intelligence is set to transform traditional marketing, including the way wine is promoted, writes agriculture journalist Glenneis Kriel.

The rise of smart machines promises to revolutionise the marketing industry, allowing advertising content to become much more targeted and individualised. Instead of shooting in the dark and hoping for the best, artificial intelligence (AI) allows companies to adapt material to suite specific browser preferences.

Machine learning uses algorithms to analyse thousands of data points on each specific person, ranging from their age, gender and ability to afford specific products to previous online and offline purchases, hobbies and colour preferences.

Paul Stemmet, founder of digital company Yap that uses AI to improve advertising and marketing outcomes, says that he is not aware of the technology being used in wine marketing in South Africa, but it is already being adopted in the media, banking and motor industries.

In the wine industry, implementation would allow the industry to match browsers with their preferred colour of wine, wine variety or blend and to adapt promoted content according to browser preferences. “When visiting a company’s website, machine learning would allow the site owner to proactively filter information so the browser is only exposed to the content they’re interested in,” Paul says.

As more data becomes available about a browser’s preferences, machine learning would allow site owners to recommend “similar products”. Instead of sending SMSes or emails to inform everybody on their news list of sales, companies would be able to only target those who are likely to buy the product using information customised precisely to their interest.

Advertising or the website itself may become more refined to match a browser’s lifestyle and other preferences. If for example the browser is outdoorsy, the wine could be promoted in the context of a picnic, or if the browser is more into fine dining a fancy restaurant setting could be used. Even the colour scheme of the website or advertisement could be adapted to accommodate specific colour preferences.

Incorporating machine learning into a website might sound like a daunting and expensive exercise because of all the fancy-sounding tech involved, but it’s not. “A great deal of work has been done over the past few years to simplify data science,” Paul says. “Wineries merely have to ask their web developer to add this capability to their content management systems, and you’re good to go.”

The only drawback is that individualised marketing requires more content than blanket advertising, where a one-shoe-fits-all approach is used. The higher success rate would however more than make up for this cost, Paul says.

An added advantage is that the technology would allow wine companies to get rid of those annoying pages where browsers have to enter their birthdays to access the page. Instead their age would’ve already been identified in the few seconds it takes them to open a page, even if various people are using the same computer.

Offsite marketing campaigns

For offsite marketing – when you want to place an ad on another company or person’s website or social media pages – machine learning could be used to refine the target audience and calculate the chances of specific people in this group reacting to the individualised content. So instead of paying a fortune for a traditional blanket advertisement that might not be of interest to the majority of browsers, the individualised ad is only pitched at people who are likely to buy the product.

“AI makes advertising smarter,” Paul says. “So instead of a browser being bothered by an ad for a premium quality wine they can’t afford, they’re now only confronted with wine in their price class. Anad could also be programmed to appear at a specific time of day, when the person has been found the most likely to react positively to marketing.”

Besides people being harassed by fewer ads, AI would turn into a kind of push notification to ensure people are reminded of what they need or want before they actually start looking for it. The algorithm may over time learn how long it takes a customer to drink wine, so that wine owners would be able to send them promotional information just before they run out in also to notify owners when customers fail to make “regular” purchases.

To make use of this type of advertising, wine estates should connect with a digital demand agency such as Dentsu Aegis Network, Ogilvy and Ole! Media Group. They use various platforms, including Google Ads, Rocket Fuel, Media Math and AppNexus, to match browsers with the most appropriate advertising content. Some of these platforms use transparent pricing models, while others charge a certain percentage commission on sales.