Yes, the Watson technology is still maturing, but I am convinced that within five years the Watson platform will learn faster and make better predictions with each new field it understands. That’s because, as Karasick told me, “If you train a system like Watson on domain A and domain B, then it knows how to make the equivalence between terminologies in different domains.” That means as Watson solves problems in chemistry; it can generate probable solutions in Physics and Metallurgy too.
Imagine how this might be applied to marketing. By using Watson as a service, a business could train Watson to understand its customers, then use predictive models to recognize new products or services that their customers will buy.
Here’s how Watson can revolutionize marketing
Predict new trends and shifting tastes
Watson is a voracious consumer of data, and it doesn’t forget anything. You can feed it data from credit cards, sales databases, social networks, location data, web pages and it can compile and categorize that information to make high probability predictions.
And most shockingly, Watson is well ahead of its competitors in sentiment analysis. According to Karasick, Watson can recognize irony and sarcasm - and properly apprehend the intended meaning. That means Watson can quickly analyze large sample sizes to determine whether a movie trailer, product offering or clothing line are going to work with consumers.
Analyze social conversations – generate leads
Most social listening solutions on the market today do an adequate job of giving the marketer signals and reports about their industry, competitors, partners and current customers. But it’s up to the marketer to analyze the information and take action.
As Watson has demonstrated in other domains, it can foreseeably predict what information is most important and make recommendations on how to act on it. For example, if it finds a cluster of people discussing problems that the marketer’s solution solves, Watson can automatically notify the sales team or take action on its own to educate the prospective customers.
Determine whether a new innovation will sell or not
Because Watson can learn from one domain of knowledge and make high probability predictions in another, it’s reasonable to assume that if a company wanted to understand whether a new innovation will sell or not, Watson could analyze a company’s current market and customer base to provide success probabilities.
We’re a long way off from a Watson with the taste of a Steve Jobs, but if it has enough understanding of the situation, it can produce insights that can give companies a clearer picture of the opportunities and threats.
Computer calculated and automated growth hacking
If you’re a marketer and not familiar with growth hacking, please study up fast. Growth hackers focus on innovative A/B testing techniques to maximize conversions on emails, websites, social media, online content or just about any digital media available to them. It’s a low cost but more effective alternative to traditional media.
I can see how Watson could proactively and intelligently test, measure and optimize digital content, ads, website pages even a company’s product to efficiently maximize customer growth. Andy Johns of Greylock, formerly a growth hacker for
This clearly is the digital march of progress. Watson has the potential to eliminate ineffective marketing, improve good marketing to great marketing, and to predict how to better spend marketing dollars in the future.
Put it all together and you’ve revolutionized marketing.
