You Are Not a Number
The new science of customer success
NEW! the official blog of Cerebri AI
Welcome to the Future
Not only is artificial intelligence the hottest topic in technology today, but also it is probably the least understood (and, to a certain extent, even feared.) Given the rapidly evolving AI landscape and the potential for fundamental, systemic change, we see an opportunity to help demystify as well as evangelize the transformative power of AI in the enterprise. Welcome to the official blog of Cerebri AI. Please fasten your seatbelts …we are ready for takeoff!
July 26, 2018
“Yes, And…” – How Improv Improves Machine Learning
by Andrew Kraemer
People often have mixed feelings when they hear the words “Improv Comedy.” Some know the pleasure of hearing comedic masters create hilarious worlds from the most mundane of suggestions, while others cringe as they recall trying to support a friend by attending their — incredibly uncomfortable and decidedly unfunny — improv class graduation performance.
Being good at improv isn’t something one is born with. Governed by a set of rules (irony alert!), improv is a practiced skill only mastered through repetition. These laws provide a framework to inspire creation. Improv isn’t just making jokes; it’s a philosophy. And as it turns out, creation and philosophy can be quite funny.
As taught in smelly improv classes all over the world, one of the core tenets of improv philosophy is the “Yes, And…” rule. This concept guides each character in a scene to constantly accept the dialogue of others as an integral part of their shared fictional work — hence the ‘Yes’ — and always add something to that world — hence the ‘And’.
By applying the “Yes, And…” approach, improvisors quickly scale up ideas — without fear and with a sense of play — to create something entirely new. The underlying assertion is that ensembles produce better results than individuals acting alone.
This is true of improv, but it is also true of data science. At Cerebri AI we are in the prediction business. Whenever we tackle a new problem, need to create new models, or find a new product solution, we assemble our talented crew of experts from a wide variety of fields and establish a “Yes, And…” environment for the entire group. Using this distinctly collaborative approach, we created Cerebri Values™, the first universal measure of customer success. We also had quite a few laughs along the way.
Andrew Kraemer is the tallest data scientist in the Austin office of Cerebri AI. Despite his best efforts, his parents still don’t know what he does. Hopefully they will read this post. Hi Mom!
July 23, 2018
The Rise and Fall and Rise of Big Data
Not so long ago, “big data” was the hottest buzzword in technology and every VP at a major enterprise jonesing to add an “S” to their title was racing to harness the mystical powers of massive yet supposedly manageable data sets. Armed with big data, no problem would prove insurmountable, all forecasts would be met, and every status update would be green.
But lately that luster appears to have faded.
After a sharp rise starting in 2012, a Google Trends query reveals “big data” peaked as a search term in October 2014, and since then interest dropped gradually but consistently and currently stands at less than half its all-time high.
So what happened? Did big data go away? Nope. in fact we now live in a world of even bigger data, of unfathomably large amounts of 1’s and 0’s where so much new digital information is being generated daily — over 2.5 quintillion bytes of Instagram selfies and cat videos, but also boatloads of e-commerce transactions, online product research, and comments/reviews — that according to one estimate from IDC and Dell EMC, the total amount of data in the world is more than doubling every two years, rivaling Gordon Moore’s famous law about increases in computing power.
Clearly there is no shortage of data. In fact, one of the fastest-growing sectors of the $3.5 trillion IT industry — cloud computing — is devoted to wrangling these staggering quantities of bits and bytes in a manner that doesn’t explode the opex line of a major enterprise’s P&L.
Big data has become a fact of life. What’s now driving the conversation are the means to finally achieve those KPI endgames, as the ability of highly sophisticated analytics techniques has, until recently, proved incapable of exploring the depths of these new Great Lakes of corporate data.
Let’s return to Google Trends for another perspective on the “big data needs killer tools” line of thinking:
Although not quite a tipping point in the Malcolm Gladwell sense, the moment, in November 2016 — less than two years ago — when machine learning surpassed big data in Google searches nevertheless marks a critically important paradigm shift, as the conversation about the fastest path to a corporate competitive advantage with data pivoted from inputs to outputs.
Before then, none of this was possible at scale. At Cerebri AI we move the border of what’s possible every day using advanced analytics, mapping previously uncharted big data territory.
Our Cerebri Values™ software, powered by machine learning, lifts customer success for major enterprises by turning big data into killer data.
July 19, 2018
Life is a Journey, Not a Destination. Customer Success is Both.
If you ask the Internet, noted American author Ralph Waldo Emerson is usually attributed as the source of the well-worn maxim that “life is a journey, not a destination.”
One problem: Emerson never said that.
While the provenance of that proverb remains murky, here are some notable authentic Emerson quotes along similar lines:
“Do not go where the path may lead, go instead where there is no path and leave a trail.”
“What lies behind us and what lies before us are tiny matters compared to what lies within us.”
“To be yourself in a world that is constantly trying to make you something else is the greatest accomplishment.”
Emerson, who lived during the nineteenth century, was a big believer in the uniqueness of each person, known as individualism. Roughly two centuries later, our exceptional Cerebri AI team of data scientists and software developers is generating empirical proof that Emerson was right.
In the new science of customer success — enabled by advanced analytics powered by machine learning — we have learned that traditional top-down approaches to customer engagement lack the capability to understand and respond to an individual’s unique customer journey.
Authentic customer journeys are like signatures or DNA or fingerprints — no two are alike. As the title of this blog proclaims, you are not a number. In fact, you are the sum total of each step in your journey, and each step is different. For major enterprises in today’s hyper-connected and ultra-competitive environment, the importance of this principle cannot be overstated.
July 16, 2018
Raiders of the Lost Data
Briefly returning to the same hot dog stand from last week, after the Buddhist monk places his order he pays with a $20 bill. The vendor hands him his hot dog then turns away to help the next customer. The monk says, “Wait, where’s my change?”
The vendor smiles and replies, “Change must come from within.”
This sage advice also applies to customer success. As noted in the previous “Power of One” post, focusing on each individual customer’s unique journey is the key to a deeper understanding of their experience over time, and the foundation for sustaining a mutually beneficial and valuable relationship in the future.
For a major enterprise with millions of customers, the first step to changing your customer success strategy must also come from within, because the multitude of data points that comprise each individual customer journey — including sales, marketing and service — have in all likelihood been captured and recorded somewhere. And once it’s cleaned up and sorted into comprehensive individual customer journeys at scale, your existing data becomes a powerful source of competitive advantage. That’s the good news.
Now for the not-so-good news: in most XL-size enterprises, this data doesn’t talk to each other and instead sits disconnected in various corporate silos, like the fabled Ark of the Covenant in the final scene of the classic Steven Spielberg film Raiders of the Lost Ark, boxed up and forgotten.
At Cerebri AI, like Indiana Jones, we specialize in tracking down hidden treasure troves of customer data. This change comes from within, and it fuels the engine powering Cerebri Values™, our newly launched SaaS product that uses machine learning to lift customer success. Now normally we don’t encounter deadly booby traps or get dropped into pits full of snakes as part of the data discovery process, but this brave new world of artificial intelligence is always an adventure!
July 16, 2018
Time Series, Journeys, Machine Learning, and Risk Management
Alain Briançon, Ph.D., Vice President – Data Science, Cerebri AI
This point-of-view paper highlights selected papers and studies that applied machine learning to time series (event series) for the purpose of risk management in financial institutions. What varies between the studies selected are the data stream/time series and processing used. This paper is, by no means, an analysis of the state of art in research, rather it is a series of touchpoints that exposes:
the increasing comfort in using these techniques in the financial industry.
the relative importance of events with various attributes.
highlights of key techniques relevant to the Cerebri Value System.
Cerebri AI developed the Cerebri Value System, which is powered by machine learning and employs a radical new method of lifting customer success. Cerebri Values quantifies each customer’s commitment to a brand or product and also dynamically predicts “Next Best Actions” at scale, which helps large companies focus on the highest-ROI tactics for accelerating profitable growth. In the case of financial or insurance institutions, risk is an essential element of this design. It is in that context that linkage of these papers to Cerebri AI is highlighted.
Charpignon, Marie-Laure, Enguerrand Horel, and Flora Tixier, “Prediction of consumer credit risk.” Stanford University (2014)
Charpignon et al. used machine learning to predict consumer credit risk. They used four types of models (logistic regression, classification and regression trees, gradient boosting trees, and random forest), applied to a wide range of data points including: age of the borrower, number of dependents in family, monthly income, monthly expenditures, total credit card balance/total credit card limits, and payment statistics. As expected, machine learning was able to predict defaults with good accuracy. The relevant element of this analysis from a design perspective was the overfitting of the random forest algorithms used. This overfitting indicates that there are redundancies in the criteria used to do traditional risk analysis. These criteria are often computed from raw data rather being raw data. Computed data streams are more likely to be correlated with one another and more prone to overfitting. Using time series (aka customer journeys) of raw data points should avoid this overfitting and improve performance. Reinforcing learning systems should be able to exploit raw data properly as well as, potentially, integrate business rules.
Khandani, Amir E., Adlar J. Kim, and Andrew W. Lo. “Consumer credit-risk models via machine-learning algorithms.” Journal of Banking & Finance 34 (2010): 2767-2787
Khandani et al. applied machine learning to events (time series) to perform risk analysis for consumer credit default and delinquency. They used transaction-level, credit-bureau, and account-balance data for individual consumers. Some of the attributes used were the type of expenditures (discretionary vs. non-discretionary, car oriented, cash withdrawals and the like), going beyond the slower-varying traditional scores. Their forecasts were very accurate in predicting events up to one year in advance. It shows the promise of using raw event series modeling attributes of different natures. An important element highlighted was how to manage broad economic variations that impact/baseline consumer credits.
Sousa, Maria Rocha, Joao Gama, Elisio Brandao, “Introducing Time-Changing Economics into Credit Scoring,” FEP working papers n. 513 November 2013 ISSN: 0870-8541
Sousa et al. compared a traditional (e.g., not time-series based) framework with a framework using time-changing factors, including internal default analysis and macroeconomic trends. They concluded that introducing data streams is suitable for dealing with temporal degradation of credit scoring models and avoid specific drifts.
Morison, J.F. “Marrying Credit Scoring and Time-Series Data” published in Risk Management Association journal (May 2010)
Morrison integrated credit scoring information and macro-economic data organized as time series. Rather than using broad macro-economic data lifted from third source and risk aggregation bias, he aggregated the time series related to the behavior of consumers in the same geographic area, subjected thus to the same macro-economic behavior. A combination of macro derived data and aggregate data is probably the sturdiest approach going forward.
Events in consumer journeys do not, not the most past, follow a regular pattern. Payroll payments, automatic payments are obvious exceptions. Gaps in events create variability in the quality of the raw data. This variability can be handled by modeling the variance of the data (noisier data have larger variance). Variations/dispersion in data quality has been understood and modeled through changes in the variance of the data in ARCH (autoregressive conditional heteroscedasticity) models. ARCH models are commonly employed in modeling financial time series that exhibit time-varying volatility and are well suited for macro trends. Cerebri AI uses memory-based designs which is well-suited to process event series with time lags of unknown size and duration between important events from the consumer journeys.
The application of machine learning to event streams to predict risks is a maturing field. Cerebri has incorporated best-in-class techniques and its proprietary metric of customer commitment (the Cerebri Value) to deliver predictions, recommendations, and improve processes
July 13, 2018
Weapons of Mass Disruption
Generally speaking, paradoxes are almost always worth a closer look. To wit: finding new customers today has never been easier, but retaining customers over the long run has never been harder. How did the time-honored rules of customer success become obsolete? Why do top-down marketing models based on demographics look more like dodo birds and dinosaurs with every passing fiscal quarter?
The catalyst prompting this seismic shift in consumer behavior did not exist at the turn of the century, and within two decades has become something many people say they cannot imagine living without. As you probably can guess, the hero of this saga (or villain, depending on your perspective) is the almighty smartphone. Over 1.5 billion smartphones will be sold globally this year, and penetration among all U.S. adults is now at an all-time high of 77%, up from 35% as recently as 2011 (Source: Pew Research Center).
The trend among young adults is even more striking—94% of Americans aged 18-29 own a smartphone, and 39% of that group report they are online “almost constantly.”
This dynamic is the stuff of nightmares for C-level executives in major B2C enterprises and financial services companies: it’s midnight—do you know what your customers are searching for online RIGHT NOW?
The advent of the smartphone era, when coupled with global availability of mobile high-speed Internet access, means no consumer ever has to delay gratification when it comes to their needs and wants.
(Loyalty may not be dead, but apparently it’s taking a very long holiday and did not respond when asked for comment.)
Companies must acknowledge the painful reality that even their most committed, long-term customers are only a couple of swipes away from becoming their competitors’ newest sales leads. Switching costs are vanishing. In this chaotic environment there is only one path to success: accurately anticipating your customers’ needs before they decide to ask Google or Amazon and take preemptive action to close the sale. Snooze and you will undoubtedly lose.
This is what we do at Cerebri AI. We connect the dots and make predictions about what your customers are ready to buy and the best way to make sure they buy it from you. Here’s the really good news: the key ingredient in this new recipe of customer success is the sum total of every digitally recorded interaction between a company and its customers.
At Cerebri AI our software product, called Cerebri Values, extracts previously unknown or unseen patterns of behavior as a basis for predictions that provide a competitive advantage. And we do it using artificial intelligence and machine learning, another technological breakthrough that is radically reshaping the economic landscape. Machine learning predicting human behavior – yet another paradox worth a closer look.
July 12, 2018
The Power of One
A Buddhist monk walks up to a hot dog stand and says, “Make me one with everything.”
What does this classic one-liner have to do with customer success? The answer is found in the fundamental incompatibility of traditional sales and marketing models—top-down funnels—and the best way to build valuable, sustainable customer relationships, which is precisely the opposite: from the ground up, one at a time.
Major enterprises with millions of customers have billions of data points about those customers, but rarely is that data organized into comprehensive and chronological individual journeys. In a typical XL-size company, data sets reside in functional silos that never communicate, so the view of the customer is always dangerously incomplete.
Cerebri AI is an enterprise software company that uses machine learning to power our Cerebri Values product, but we start and finish at the level of the individual customer. We gather data from every corner of an enterprise to build a complete customer journey—one with everything—for millions of individual customers. And no two journeys are alike, as you might expect.
Armed with this ontology, the Cerebri Values approach to customer success is a radical departure from the classic “Segmentation, Targeting, Positioning” marketing model (STP for short).
At Cerebri AI we believe in the power of one, and thanks to advances in artificial intelligence in the last few years we now have the capability to detect patterns of behavior and make predictions about individual customers at scale. When you order Cerebri Values, you get one with everything.
July 11, 2018
Customers and the S-Word
If committed customers are the lifeblood of all successful companies, and cultivating those individual relationships is a critical variable in corporate formulas for sustainable profitable growth, then why do big corporations tend to treat customers like faceless numbers?
The obvious solution — building relationships one at time based on a deep understanding of that individual’s present situation and past experiences with the product/service that anticipates their future needs — may work for a smaller business, but for a major B2C enterprise or financial services institution with millions of customers, that approach usually flops due to the dreaded S-word: it doesn’t scale.
Traditional customer engagement models are top-down, not because the smartest people thought it was the best way, but because there was no other way. Until now.
Cerebri Values™ our newly-launched product, uses trillions of machine-learning calculations to reveal the real people behind the numbers, one customer journey at a time. This is old-school clienteling powered by cutting-edge artificial intelligence. And it scales.