AI Marketing + Go: How Google’s AI “AlphaGo” Beat A South Korean World Champion
The game of Go has been making headlines in the last few months, but today marks an especially historic event, especially for AI marketing fans: Google’s DeepMind computer program AlphaGo has won the final game of Go against South Korean world champion Lee Sedol, the last round in an epic 5-game tournament between artificial intelligence and human skill and strategy.
The final match, which by all accounts was a close game, is the most incredible feat for Artificial Intelligence (AI) and machine learning since the days of IBM’s Deep Blue. The reason AlphaGo’s wins carry so much more weight today is in the difference of scale. The ancient game of Go was one presumed too difficult and dependent on intuition for computers to play at a world-class level, with infinite more variations than chess.
With some seasoned Go players in our own company and as definite AI Marketing enthusiasts, we at Retention Science were especially excited to witness this achievement in deep machine learning. It also got us thinking about the parallels between Go and AI marketing — and there were more than you might think:
1) Marketers, like Go players, need to be flexible (and quick)
Go, unlike chess, is not a rules-heavy game. The number of possible Go games and winning scenarios, as mentioned, is virtually infinite. The best players (and AI’s) are ultimately the ones that embrace the game’s fluidity and put in place flexible strategies that can quickly adapt to a constantly evolving game. Similarly, today’s marketers need to embrace wide reaching, cross-channel campaigns while remaining able to adapt to constantly evolving consumer behaviors in order to lift their conversion rates.
2) One size will not fit all for Go or marketing strategies
Because there are so many possible outcomes in Go, imagining a one-size-fits-all step-by-step plan to win is a waste of time as the likelihood of failure is far greater than the likelihood of success. In the world of e-commerce marketing, the same can be said of rigid rules-based drip campaigns. Building an 18-stage email welcome series might seem like a surefire tactic to ensure higher conversion, but consider the myriad of possible paths that your users can take across channels — it’s unlikely that those 18 stages will speak to those customers as intended. An AI Marketing platform (like Cortex by ReSci) can take into account not just multiple stages, but stages that defined by individual customer behaviors and data, not from anecdotal or situational guesses.
3) Technology is the ultimate game-changer
Today’s match cemented the idea that AI can help computers make strategic decisions not just on par with, but in some cases better, than humans. The game of Go is complicated, but with the right technology, those complicated decisions can lead to great results. In the same way, AI Marketing technology can help you tap into the highest potential of your customer base.
The technology behind AlphaGo is Google’s DeepMind; in the same vein, there is now powerful AI marketing technology behind new tools that offer similar possibilities. This AI marketing technology is capable to adapting to your customer’s behavior, just as AlphaGo adapted and learned from previous games to defeat Lee Sedol, for better results over time.
Schedule a demo by clicking the button at top right- we can’t wait to show you how AI Marketing is changing the game for other online retail businesses.
Next week, we’ll go further into how AI marketing can bring your game to grandmaster level.