Jeff Ma on Blackjack and Succeeding in Marketing with Data and Analytics
When Jeff Ma speaks, there’s always a passion and excitement in his voice. Whether he’s talking about his days on the MIT Blackjack Team (a team that used card counting and analytics to beat casinos at blackjack), hanging out with celebrities during the filming of 21, or discussing the value of data and analytics, the enthusiasm Ma has is captivating. During his keynote at Symposium Scottsdale 2017, Ma shared insights on how to become a better marketer through data and analytics to affiliate publishers and advertisers, who seemingly hung on his every word While attendees at the panel were probably hoping to hear some secrets at winning millions at blackjack, Ma provided more important insights instead: Ma discussed the value of using analytics and data to understand and succeed in marketing efforts.
Ma starts the keynote by introducing the audience to a key concept: embracing failure. Ma’s point of embracing failure is simple – we cannot succeed without the risk of failure, and we can’t grow from failure unless we look at what went wrong and why. This is a theme that Ma revisited towards the end of the keynote, but for
Ma, being the expert in blackjack that he is, explains that by becoming data driven you’re trying to “reduce the edge the casino has” against you. However, as marketers, we’re not typically going up against casinos – instead, our edge comes from the challenges getting customers to click, convert, and/or purchase. Using data, according to Ma, is the way to reduce the edge and challenges we face in affiliate marketing.
From a marketing perspective, what can be achieved through the use of data? There are a lot of answers to this question, whether you’re a publisher or an advertiser. According to eMarketer, some of these insights for affiliate marketers include understanding demand/interest from customers, growing volume of resources, and becoming more customer center. Publishers can also benefit from using data to make insights on how to improve their content, track their visitors, and see what’s working and what isn’t. The linked report from TechTarget also mentions some additional benefits advertisers gain through data, such as brand development and better campaign tracking.
And on the list can go. In short, properly leveraged data can give you countless insights. What’s more, more marketers are recognizing the value of data and, according to another article in eMarketer, 2017 is expected to have more marketers investing in data-driven campaigns to achieve multichannel attribution, real-time optimization, and cross-channel and cross-device insights.
The benefits are all there, and the question now becomes “how?” How do we start becoming data-driven and make the most of our data to make informed, strategic marketing decisions?
The answer is two-fold. On the one hand, you have the tools and resources already available through your network (Rakuten Marketing, for example, offers features like reporting, attribution through Cadence, and tracking to create a holistic view for both publishers and advertisers), but Ma focuses on a different answer. Ma identifies that to become data driven, we must overcome challenges and inclinations that plague all of us.
Emotional, Omission, and Cognitive: 3 Biases
Ma’s keynote is significant in highlighting three biases that we face when making data-driven decisions: emotional, omission, and cognitive. Each of these can inhibit the ability to effectively make the right choices with, or even trust, your data. Let’s look at each one a bit closer:
Emotional bias: Ma explains that in blackjack there are a lot of basic, math-driven rules that every player should abide by in order to be successful. The problem is, according to Ma, most don’t because they have an emotional investment that causes them to make decisions they shouldn’t be making.
Omission bias: Omission bias is more easily described as “inaction” – you know there’s a right choice, and you know the data is telling you to make that choice, but because that choice is associated with risk, you decide inaction is better than taking action and having to deal with possible failure. Ma explains this through several examples in blackjack where you know you should hit and the math shows that you’ll likely win if you do, but because you don’t want to take the risk of busting out, you decide to stay.
Cognitive bias: Cognitive bias is the classic case of making a choice because other people expect you to make that choice. Ma explains this in terms of blackjack where you might want to hit on a hand you have or split your hand because, statistically, the odds will favor you if you do, but you decide not to do that because the other players at the table might judge you poorly for doing so.
These three biases, according to Ma, interfere with our ability to rely on and trust our data, causing us to make poor decisions even though we have the information we need to make the right one. For example, in the world of content publishing you might focus too much on a product you love to an audience that doesn’t care or seem interested (emotional); decide only to focus on a small portion of your audience that does care about the product even though a much larger portion of your audience has made it clear they’re not interested (omission); or only talk about products that other people say you should be talking about even though you’re not interested in them but you don’t want to be an outlier (cognitive).
These biases come in many forms, and sometimes they’re not as black and white as the example above. Ma notes that things like making gut decisions and loss aversion are two symptoms of these biases, and they can be seen in both publishers and advertisers. Your data might tell you the right product to promote to the right audience, but you may decide to go against your data because “your instincts know better.” Studies such as the one written about in the Harvard Business Review show just how these fallacies come about, but long story short: your gut isn’t better than your data.
There’s also the concept of loss aversion, which you can typically see take place in omission bias. According to Ma, we as humans love to win as much as we hate to lose. In fact, we may hate losing more – so much so that we want to avoid failure more than we want to achieve success. Loss aversion, explains Ma, occurs when we’ve already succeeded a little (won a hand in blackjack) and we have the potential to win much more – but we could also lose what we’ve won, so we decide it’s better to not take the risk and walk away, even if the chances of succeeding were still high.
So, becoming data-driven has a lot of challenges, but how do we get past these challenges?
Risk vs Reward
Overcoming biases is accomplished in a few ways, but one critical aspect to identifying – and overcoming – these biases is “risk vs reward”. This is a common concept, and the basic idea is that you have ‘x’ amount of risk you need to take in order to yield a reward, ‘y’. If 'x' is greater than 'y' we’re likely to shy away from taking the
Ma shares a personal story to highlight this point. Ma talks about his mother being in the hospital with an illness that was terminal. They could do nothing and wait for the illness to take his mother, or they could take a risky procedure that could be fatal or it could save her life. Ma weighed the risks (losing his mother) to the reward (his mother surviving) and decided to move forward with the procedure. The procedure was a success and his mother survived, despite the fatal risks.
The lesson Ma is illustrating highlights going in-depth to understand the risks and rewards of a decision. In marketing, you can face a range of low or high risks and rewards. However, understanding the risks and the rewards through data analysis can make sure that the decision you make is an informed one, and you're not leaving things to chance (or gambling!). It all comes down to understanding what your data is telling you, and understanding what that means to you and your brand or publishing model.
5 Steps to Succeeding and Making Better Decisions with Analytics
Ma shifts his focus to how marketers can succeed with data and analytics. First, he outlines the process itself – first, you collect the data, then you analyze the data, and then you implement the changes. While it’s easier said than done, this is the basic steps for what you need to do with data.
More importantly, however, Ma highlights the five ways to make better decisions utilizing your analytics. These are critical because, while the steps to getting the insights you need are fairly straightforward, putting that data to work can be challenging or even intimidating.
The five steps are as follows:
Don’t favor inaction over action: You can’t be afraid to do something – especially if your data is saying you should change something up. Ma gives an example of Facebook's growth from 30,000 users to 500 million. If the people at Facebook didn't take risks or try different things, they wouldn't have had the growth they did.
Focus on the big picture (long term perspective): Ma describes looking long term with a blackjack example – the player needs to be able to see beyond one hand, and think about the game on a larger scope. You may lose a hand, you may win a hand, but if you’re not thinking in the long term you’ll struggle to find success.
Process over results: Ultimately, you should be looking to understand what you did right or wrong in your marketing campaigns and efforts. Did you succeed because of the offer? Because of the partnerships? Because of the placements? Being able to evaluate what works and what doesn’t is, in Ma’s opinion, more important than the results themselves, because you can replicate your processes that work and avoid the ones that don’t. Trying to repeat results without understanding the results is difficult, and doesn’t offer nearly the same support (especially for long-term goals!).
Stick with it!: If at