Tournament Time: Analytics is the answer

predictions, analytics and decisioning
Tools like predictive analytics, machine learning, and real-time decisioning work together to help you attack almost any problem more intelligently – whether that’s winning your office tournament bracket, or saving the day for one at-risk customer.

It’s the month of Madness. At exactly this time every year in the U.S., the college basketball world loses its collective bananas – with every fan holding their breath, hoping and praying that their school (their school!) will get into the tournament and bring home a championship. Luckily, here at Pega we’re more than just tech nerds – we’re sports geeks as well – so we’ve provided a week-by-week tournament breakdown to help you master your bracket.

Week 1: Basketball, tournaments, customers: In data we trust

Week 2: Analysts, algorithms, data scientists: It’s all about the math

Week 3: Upsets happen: Change is constant, and adaptation is everything

Week 4: Fans and customers: Relationships matter, don’t foul it up

Week 1: Basketball, tournaments, customers: In data we trust

We’re quickly approaching the Sunday before the tournament opens, and it’s crunch time. Time to run the numbers. There are 351 teams competing for 68 spaces - the competition is beyond tight. There have been five months of college games for the selection committee to evaluate – that’s five months of matchups, jump shots, dunks, free-throws, 3-pointers, personal fouls, and parquet floors. Five whole months of double-dribbles, alley-oops, and Sports Center highlights. Five months of bad calls. Five months of coaches throwing chairs; five months of heartbreaks and hoopla... every bounce generates information.

When the committee decides which teams get in, they look at everything – every single byte of data matters. Even the smallest thing makes a difference.

The same thing can be said for your customer relationships. In basketball, those 351 teams average 30 games per season. That’s about 10,000 games played in total, every year… which isn’t really a big number. By contrast, large enterprises can have millions of customers. For example, a large bank may serve 50-100 million customers across its lines of business. Those individuals generate massive amounts of information: Card swipes. ATM withdrawals. Deposits. Transfers. Mobile banking. Website browsing. Email opens. Online ads. Mortgage calculators. The list goes on, and on, and on.

All of that activity can add up to a more than 20 or 30 billion customer interactions every year – and that’s just for ONE company. That information must be leveraged to understand those customers and ensure those customer relationships thrive.

Predictions are tough, no matter how much information you have at your fingertips. Basketball? Tournaments? Customer Engagement? It all comes down to the data you collect. “In Data We Trust.”

Pega’s own CMO, Tom Libretto, and CTO, Don Schuerman, join me, Matt Nolan, in our weekly segment Pega Sports Talk to discuss all things basketball. Check it out in the video below!

Week 2: Analysts, algorithms, data scientists: It’s all about the math

The results are in. Did your team make it? My Purdue Boilermakers did, in case you were wondering…

So now that we know who’s in and who’s out, we can get down to business predicting who’s going all the way. Keep in mind that this is a billion dollar stage, and the speculation is intense.

A lot rides on these predictions – from your everyday-eddy office brackets to big money in Las Vegas… but how does a person even begin analyzing all the data that gets collected, and use that to make intelligent choices? Essentially, it’s time to put the statisticians (or data scientists) on the job. They’re the only folks with the skills required – and they’ll use predictive analytics and various forms of AI to determine who’s most likely to advance in each round.

This can be hard to do with only 10,000+ games worth of data to sort through. That seems like a lot, but compared to large banks, telcos, or insurance companies – who are likely collecting customer data from millions (maybe even billions) of channel interactions every year – that’s not a lot of data to work with.

Regardless of whether we’re talking about 10 thousand observations, or 10 billion – the tools we’ll use are pretty standard. They include technology such as:

  • Propensity Modeling, with Machine Learning – to predict what’s likely to happen.
  • Complex Event Processing (CEP) – to bring in real-time data, as it happens.
  • Natural Language Processing (NLP) – to convert text information (like conversations, or analyst reports) into usable data that provides real intelligence.
  • Performance Simulation – to test our predictions, and see how well they perform.
  • Customer Decision Management (CDM) Engines – to convert all of the data, analytics, and projections into actual decisions, and distribute them out into the universe.

Arming yourself with these tools will help you be data-driven, precise, and balanced. They’ll work together to help you attack almost any problem more intelligently – whether that’s winning your office tournament bracket, selling a million more units of product, or saving the day for one individual at-risk customer.

Check out the week 2 edition of Pega Sports Talk, as Tom, Don and I discuss who we think will go all the way in this maddening tournament.

Week 3: Upsets happen: Change is constant, and adaptation is everything

Underdogs and Cinderella stories are the highlights of every tournament, and just one crazy upset will send everyone rushing back to their bracket to evaluate the impact. Absolutely nothing is certain when it comes to March college basketball... but did you know that there has never been a perfect bracket recorded, officially? With 68 teams playing in a single-elimination tournament, the chances are actually just 1 in 9.2 quintillion (18 zeros) of predicting every game with 100 percent accuracy.

Some things are just hard to predict- no matter how good the science, or your tools.

But we make some things harder than they have to be. For example, when a company engages with a customer, the average response rate for a campaign is less than 1 percent. That means that their predictions are wrong 99 times out of 100. Yes, modern consumers have lots of options, and can use dozens of channels to engage in – but a 1 percent response is dismal, no matter what the circumstances.

Organizations have to do better if they want to stay engaged with their customers. If a company isn’t consistently talking to customers about what matters to them – about relevant things – then their messages will mostly be ignored. That kind of situation isn’t good for anybody:

  • The customer gets over-exposed, dodges all marketing, and can’t take advantages of offers.
  • The company’s customer relationships suffer, making it harder to achieve financial goals.
  • The entire market suffers with additional regulations to protect consumers from “spam.”

The problem isn’t apathy – every company out there is working desperately to engage their customers more efficiently and effectively. They know it’s important because smaller, more nimble digital-native competitors are nipping at their heels and constantly eating into their market share – they can’t afford to sacrifice their customer experience.

Customers interact with you every minute of every day someplace – you have to learn from it and be ready to adapt instantly. Are they happy? Sad? Angry? In-Market? Out? Do they need service? You need to know, and get ahead of it. You’ll never earn their trust if they’re ignoring everything you say.

On Pega Sports Talk Week 3, Tom, Don, and I discuss tourney upsets and the amped-up level of excitement for this week’s gameplay.

Week 4: Fans and customers: Relationships matter, don’t foul it up

Yeah - the tournament is finally over. For better or worse, we know who won, who lost, and which teams exceed all of our expectations. But as the excitement of the biggest college sporting event of the year fades away, the unofficial second season is just about to start.

A college team’s performance in March can play a huge role in its future. For example, Loyola University Chicago took advantage of their surprise success by opening a pop-up merchandise store and discussing ways to drive increased applications and enrollment. A great tournament showing boosts recruiting, enrollment, ticket prices, and alumni donations for years to come – and a school that fails to capitalize on that momentum is missing a major opportunity. The relationship a school has with its students, athletes, and supporters requires constant tending – and the higher their expectations, the more difficult it is to satisfy them year-after-year. That isn’t a bad thing necessarily, but the university has to stay on top of it – and make sure it’s investing in itself – attracting the right kind of talent and support to sustain a peak level of performance year-after-year.

Companies need to approach their customer relationships in the same way. To paraphrase the immortal Maya Angelou:

“They won’t remember what you said. They won’t remember what you did. But they will always remember how you made them feel.”

In the customer’s mind, you’re only as good as your last interaction. If it was negative (awkward, irrelevant, or distracting to them), then it will become that much harder to engage effectively the next time. But if you invest in those individuals appropriately – by leveraging the data, analytics, and technology that’s required to truly know them as people – then the emotional connections you build with those customers will pay you back in spades.

In the final Pega Sports Talk segment, Tom, Don and I wrap up with our final thoughts on the tournament. Thanks for watching!


LEARN MORE:


ABOUT THE AUTHOR: Matthew Nolan, marketing director at Pegasystems, is a college basketball superfan. When not rooting for the Purdue Boilermakers, he’s enabling the vision and go-to-market strategy for Pega’s marketing and advertising technology portfolio. Before joining Pega, Matt was General Manager of the National Data Cooperative at Target Analytics, and served as Director of Modeling, Analytics & Data Services for Blackbaud, Inc. He is a regular keynote speaker who shares his professional insights on more than 17 years of marketing technology experience.