On the IBF On Demand podcast, I recently spoke to a real innovator doing great things in the forecasting space. He comes from a disease forecasting background and many of his predictive analytics techniques are directly applicable to business forecasting. We’re not just talking about what will will happen but why, and identifying changes in consumer behavior before they happen. 

His name is John Cordier and he runs a consulting firm called Epistemix. Backed by the National Institutes of Health and the Bill & Melinda Gates Foundation, Epsitemix models the spread of infectious of diseases, and now applies that expertise to business.

There is much we can learn from him when it comes to maximizing the value of our own data. The following are some highlights of that conversation.

Can you give a breakdown of how you approach epidemiological forecasting? Then we’ll dive how those approaches can be applied to demand forecasting.

We approach forecasting from the bottom up. Those in our space would recognize agent-based modeling as our underlying technique, but all you need to know about how we forecast is that we represent every single person in the entire country and forecast based on their behaviors.

In the infectious diseases space, our technique has found is that within the United States there are 9 “ontological units of epidemicity.” Meaning there are 9 regions in the US with distinct seasonal behaviors. The data tells us why in late spring COVID-19 cases go up in the South but not in the Northeast and then inverts in fall, and so on.

We’re able to generate these seasonal patterns from human behaviors. There’s so many behavioral variables you can build into a model. What we end up doing is calibrating to the one or two most important pieces of data that we prepare the model against. With COVID we disregarded cases because case counts was too noisy for us to have any accurate projections so we use hospitalizations and deaths which are more stable, yet not perfect.

So if I’m understanding this correctly, you’re looking at the seasonalities but you’re mainly looking at the human behavior driving some of those spreads and then instead of just modeling the data, you’re monitoring how the drivers are changing and impacting each other?

Yes. We aim to capture the non-linear relationships. We’re getting these non-linear connections between different actions so we can say which behaviors are driving an epidemic – or indeed the adoption of a product or service – forward. If it’s a disease we can test what interventions will drive those numbers down, or in the case of a consumer product or service, test strategies we think might drive sales up.

This is fascinating because I can see so many use cases for it. I’ve been preaching for years that we need to look externally and develop that ‘outside-in’ type of thinking in predictive analytics for business forecasting. Are we limiting ourselves by only looking at sales data?

If you’re only using the data that exists internally and you’re making your decisions on those assumptions, you’re really saying that the future is going to look like the past. If you’re not looking at how behavior changes or how the environment is changing and how that impacts the drivers of adoption or purchases, you’re going to miss the tipping point of adoption or when you hit market saturation

One example of this is a social media app that came out in 2020 called Clubhouse and everyone was saying it’s going to totally take over Facebook and Instagram. Well, if you looked at the behaviors of adoption you would have seen that the tipping point is going to come pretty early. We forecast that tipping point in early 2021 right after the Consumer Electronics Show. You could have known then that in fact this wasn’t actually going to be the next big thing.

In business predictive analytics we start with data and the different types of standardized models that can be applied to understand the data, but you start by representing the reasons why something’s happening. Why do you start with the ‘why’?

Our goal is not just to predict demand, but to give decision makers the ability to understand how to shape demand given the resources they have available to them. Whether forecasting a health outcome, product sales, distribution of an idea, or the number of votes a given candidate is likely going to win, our synthetic population is the starting point to generate a forecast from the bottom up. 

Our interactive synthetic population includes a representation of every person, household, school, workplace, hospital etc in the country – consider this the ‘clay’ or ‘substrate’ that you’re starting with. Then we enable businesses or subject matter experts to test their assumptions about how their population is changing. These set the baseline assumptions.

You can then generate models that recreate the past data and explore “what-if” questions about how things might look different given changes in behavior or changes in the environment. Once you have a working model you’ve really created a simulation of the most meaningful drivers of behavior and outcomes that you can test demand shaping strategies against.

What is synthetic data?

Synthetic data is a broad description of techniques used to take observed data and create data that downstream increase the confidence of a decision somebody has to make. In our world of synthetic data, you can take a real data set and describe how any behaviors in the population might change. Using that as a baseline forecast, you can test assumptions to better understand the outcome you’ll achieve through different decisions you can make.

Because our synthetic population has both time and geography included, you can generate synthetic data sets complete with geospatial and time series data.

So the output is more of a probabilistic type of forecast?

It’s a stochastic type of forecast, so you’ll get something different every time. Everything is probabilistic. Our users might run a thousand simulations and get a different result every single time. But once you have that data, you have more or less ‘beat down’ the noise to get narrower confidence intervals around your uncertainty bounds of the most likely outcome.

It’s a relationship between people, things, and the environment and how all three are going to change in the future, right?

Yes. We’re able to understand the interactions between people in households, at work, in the community, and how people react to things they come in contact with online – all in combination to see the emergent behavior of entire populations. Malcolm Gladwell and others have written about what makes social epidemics take off – it comes down to those three things: what population, what thing, and the environment they exist in.

Whether it’s a product or a disease, the terms adoption and spread are similar. Imagine a scenario where you and your partner are at the dinner table and you ask “have you heard about the thing in the news today?” Maybe they didn’t, but you’re now the third person that has brought it up.  After hearing it from a few people, they look it up and tell another 4 people. Next thing you know, it’s made national news for the next cycle and millions of people are exposed to the information.

Information, behavior, and diseases – when using a bottom up approach – are emergent phenomena. At Epistemix, we help companies project emergent phenomena into the future.

Do you have any other business related examples?

We have a couple of customers that are launching consumer apps. What they’re trying to understand is the demand for their app, the network effect they’re able to create, and what actions they can take most influence adoption. 

They use the synthetic population to test marketing strategies designed to influence demand. One of the companies we’re working with, Earbuds – a new music sharing app – is trying to understand the most sustainable path to 150,000 monthly active users by June of 2023 by testing influencer driven campaigns, targeting specific online communities, and other marketing campaigns.

I see parallels with retailers modeling population shifts, for example people moving from New York to Texas and Florida.

As it happens we did a project last year with the Remaking Cities Institute out of Carnegie Mellon University and the question they wanted to ask was how remote work acceleration is impacting where consumer retail is going to be. 

We did a study to see where the workforce of Chicago is going to be in 2024. Because we had to root this in an actual problem that businesses wanted to solve, we framed the question as “where do I get my coffee in 2024?” The whole idea was to base coffee shop site selection off of consumer preferences of where they’re working, the projected density of the population, and existing locations. 

We identified locations within Chicago where the supply will be too high and other locations where there are going to be too few coffee shops. Given the information, generated from the synthetic population as a starting point, coffee shops and developers can understand the future population density and drop in three or four coffee shops in an area to capture new demand.

The cool thing about working with the synthetic population is that you can add information as you learn more about the problem you’re trying to solve. For example,  take the question of “what will happen to coastal cities in the US over the next 50 years given the increase in extreme weather events?” The data doesn’t yet exist, but you need to forecast what might happen if… Using the Epistemix platform you can test how the different “what ifs…” influence people’s behavior and test what decisions you can make to shape or be positioned to take advantage of the changes to come.

My Thoughts On These Cutting Edge Approaches

This was a great, high-level conversation. You can watch the full conversation here. You may not be ready for all of it but we can start leveraging some of the key elements. The key is to start thinking ahead of where the consumers may be in the future and how they’ll be behaving. Are you ready to not only look at a forecast but to create an idea of what your market will look like in the future? If we can do that, sales and marketing can act on changing consumer behavior before it happens.

This is a mindset change for a lot of us. This is Predictive Analytics, using different types of modelling – getting into probabilistic and forward-looking projections, and trying to understand the ‘why’.

This level of analytics is very much still emerging. Understanding and adopting these techniques requires continual learning, changing your thinking, and learning new skills and information and bringing them into your forecasting process. If you’re prepared to challenge yourself, you may find that you advance both yourself and your company.

Click to order your copy now.

To add advanced analytics models to your bag of tricks, get your hands on Eric Wilson’s new book Predictive Analytics For Business Forecasting. It is a must-have for the demand planner, forecaster or data scientist looking to employ advanced analytics for improved forecast accuracy and business insight. Get your copy.

I recently had a fascinating and enlightening conversation with one of the leading figures in predictive analytics and business forecasting, Dr. Barry Keating, Professor of Business Economics & Predictive Analytics at the University of Notre Dame.

He is really driving the field forward with his research into advanced analytics and applying that cutting-edge insight to solve real-world forecasting challenges for companies. So I took the opportunity to get his thoughts on how predictive analytics differs to what we’ve been doing so far with time series modeling, and what advanced analytics means for our field. Here are his responses.

What’s the Difference Between Times Series Analysis & Predictive Analytics?

In time series forecasting, the forecaster aims to find patterns like trend, seasonality, and cyclicality, and makes a decision to use an algorithm to look for these specific patterns. If the patterns are in the data, the model will find them and project them into the future.

But we as a discipline realized at some point that there were a lot of things outside our own 4 walls that affected what we are forecasting and we asked ourselves what if we could in some way include these factors in our models. Now we can go beyond times series analysis by using predictive analytics models like simple regression and multiple regression, using a lot more data.

The difference here compared to time series is that time series looks only for specific patterns whereas predictive analytics lets the data figure out what the patterns are. The result is much improved forecast accuracy.

Does Time Series Forecasting a Have Place in the age of Advanced Analytics?

Time series algorithms will always be useful because they’re easy to do and quick. Time series is not going away – people will still be using holt-Winters, Box-Jenkins, times series decomposition etc. long into the future.

What’s the Role of Data in all This?

The problem now isn’t using the models but collecting the data that lies outside our organization. Data these days has different observations. We used to think when we had 200 or 300 observations in a regression we had a lot of data – now we might use 2 or 3 million observations.

“We used to think 200 observations was a lot of data – now we might use 2 or 3 million”

Today’s data is different not only because of the size, but also in its the variety. We don’t just have numbers in a spreadsheet – it may be streaming data, it may not be numbers but text, audio, or video. Velocity is also different; in predictive analytics we don’t want to wait for monthly or weekly information, we want information from the last day or hour.

The data is different in terms of value. Data is much more valuable today than it was in the past. I always tell my students to not throw data away. What you think isn’t valuable, probably is valuable.

Given we are Drowning Data, how do we Identify What Data is Useful?

When the pandemic started, digital purchases were increasing at 1% a year and constituted 18% of all purchases. Then, in the first 6 weeks of the pandemic, they increased 10%. That’s 10 years’ worth of online purchases happening in just weeks. That shift meant we now need more data and we need it much more quickly.

“You don’t need to figure out which data is important; you let the algorithm figure it out”

You don’t need to figure out which data is important; you put it all in and let the algorithm figure it out. As mentioned, if you’re doing time series analysis, you’re telling the algorithm to look for trend, cyclicality and seasonality. With predictive analytics it looks for any and all patterns.

Predictive analytics assumes that you have a lot of data – and I mean a lot

It’s very difficult for us as humans to take a dataset, identify patterns and project them forward but that’s exactly what predictive analytics does. This assumes that you have a lot of data and I mean a lot, and different to what we were using in the past.

Do you Need Coding Skills to do This?

Come to an IBF conference or training boot camp and you will learn how to do Holt-Winters, for example. Do we teach people how to do that in R, Python, or Spark? No. You see a lot of advertising for coding for analytics. Do you need to do that to be a forecaster or data scientist? Absolutely not.

There are commercial analytics packages where somebody who is better at coding than you could ever hope to be has already done it for you. I’m talking about IBM SPSS Modeler, SAS Enterprise Miner, or Frontline Systems XLMiner. All of these packages do 99% of what you want to do in analytics.

Now, you have to learn how to use the package and you have to learn enough about the algorithms so you don’t get in trouble, but you don’t have to do coding.

“Do you need to be a coder? Absolutely not”

What about the remaining 1%? That where coding comes in handy. It’s great to know coding. If I write a little algorithm in Python to pre-process my data, I can hook it up to any of those packages. And those packages I mentioned can be customized; you can pop in a little bit of Python code. But do you need to be a coder? Again, absolutely not.

Is Knowing Python a Waste of Time Then?

Coding and analytics are two different skills. It’s true that most analytics algorithms are coded in R, Python and Spark but these languages are used for a range of different things [i.e., they are not explicitly designed for data science or forecasting] and knowing those language allows you do those things, but being a data scientist means knowing how to use the algorithms for a specific purpose. In our case as Demand Planners, it’s about using K Nearest Neighbor, Vector Models, Neural Networks and the like.

All this looks ‘golly gee whiz’ to a brand-new forecaster who may assume that coding ability is required, but they can actually be taught in the 6 hour workshops that we teach at the IBF.

What’s the Best way to get Started in Predictive Analytics?

The best way to start is with time series, then when you’re comfortable add some more data, then try predictive analytics with some simple algorithms, then get more complicated. Then when you’re comfortable with all that go to ensemble models where, instead of using 1 algorithm, use 2, 3, or 5. The last research project I did at Notre Dame used 13 models at the same time. We took an ‘average’ of the results and the results were incredible.

The IBF workshops allow you to start out small with a couple of simple algorithms that can be shown visually – we always start with K-Nearest Neighbor and for a very good reason. I can draw a picture of it and show you how it works without putting any numbers of the screen. There aren’t even any words on the screen. Then you realize “Oh that’s how this works.”

“Your challenge is to pick the right algorithm and understand if it’s done a good job”

It doesn’t matter how it’s coded because you know how it works and you see the power – and downsides – to it. You’re off to the races; you’ve got your first algorithm under your belt, you know the diagnostic statistics you need to look at, and you let the algorithm do the calculation for you. Your challenge is to pick the right algorithm and understanding whether it’s done a good job.

To add advanced analytics models to your bag of tricks, get your hands on Eric Wilson’s new book Predictive Analytics For Business Forecasting. It is a must-have for the demand planner, forecaster or data scientist looking to employ advanced analytics for improved forecast accuracy and business insight. Get your copy.

I am going to attempt to pull back the curtain and unveil the magic behind the most common algorithms used in predictive analytics for business forecasting, and demonstrate what exactly goes on behind the scenes in the world of AI.

Here I focus on the top methods and algorithms that enable the execution of applications for demand planning and business forecasting. The following are the preferred Machine Learning and Predictive Analytics models of Demand Planners and Data Scientists (in reverse order):

7) Artificial Neural Networks

6) Decision Trees

5) Logistic Regression

4) Naïve Bayes

3) Linear Regression

2) Smoothing and Averaging Time Series Models

1) Simple Ratio Models

7) Artificial Neural Networks: (ANN) are a class of pattern matching techniques inspired by the structure of biological neural networks. ANNs combine logistic regressions into a neural network. ANNs less complicated than they may first appear  – ANNs are a collection of logistic regressions, so if you understand logistic regression, you can easily understand ANNs. Developing the final prediction is generally done by training the model and calibrating all of the “weights” for each neuron and repeating two key steps: forward propagation and back propagation.

6) Decision Trees: The concept and algorithms underpinning decision trees are relatively simple compared to other models. The general purpose of decision trees is to create a training model that can be used to predict the class or value of target variables. Decision trees build classification or regression models in the form of an upside-down tree structure.

5) Logistic Regression: These help demystify neural networks and help answer probability type questions. While it is listed as a type of regression model, it is less a calculation and more an iterative process. This can be used for anything from predicting failures to identifying if the object in a picture is a cat or not, for example.

4) Naïve Bayes: Naïve Bayes are probabilistic, which means that they calculate the probability of each class for a given set of features, and then output the class with the highest observed probability in the data set. In simple terms, based on a bunch of x’s, we’re looking at the odds of Y being y. It can be used for natural language processing, to classification, to simple prediction.

3) Linear Regression: Regression is a simple cause-and-effect modeling that investigates the relationship between dependent (target) and independent variables (predictor). Regression models come in all types and applications. One of the most common is a simple linear regression. The first step in finding a linear regression equation is to determine if there is a relationship between the two variables. This is often still a judgment call for many Demand Planners.

2) Smoothing & Averaging Time Series Models: With these models, one needs only the data of a series to be forecasted. They are simple and widely used by companies that own historical data. Here we assume that the training data set of sales history contains a trend and seasonality, and we extrapolate these patterns forward.

1) Simple Ratio Models: These express the relationship between two or more quantities.  Ratio models are utilized for a range of day-to-day purposes: understanding a seasonality index, calculating the velocity of sales and market penetration, or disaggregating a family level forecast, just to scratch the surface.  This easy-to-calculate statistic is used in various ways to guide decision making and drive forecasts.

What we find is that sometimes the simplest methods provide us a good forecast and are the best use of our time. Fancy techniques are great, but our overriding goal is to select the model that fits our business purposes and the resources available to us. We need to evaluate the model properly to ensure that it can do what we need it to. The most sophisticated techniques and most advanced technologies accomplish little if nobody understands the results. To complete the process we must step back, sometimes simplify, and communicate our analysis effectively. Or just tell them it was AI and leave them in awe  of your magical skills…

Don’t forget to join myself and a host of predictive analytics, demand planning, and forecasting leaders at IBF’s Virtual Business Predictive Analytics, Forecasting & Planning Conference from April 20-22, 2021. At just $499 for this insight-packed 2-day event, it’s an extremely cost-effective way to evolve your skills for the future of demand planning and forecasting.

To add the above-mentioned models to your bag of tricks, get your hands on Eric Wilson’s new book Predictive Analytics For Business Forecasting. It is a must-have for the demand planner, forecaster or data scientist looking to employ advanced analytics for improved forecast accuracy and business insight. Get your copy.

In transitioning from just traditional time-series modeling to predictive analytics, one of the key aspects is utilizing different causal inputs in your forecasting. It is not just relying only on internal shipment data or order history but considering external factors and a multitude of variables that paint a more complete picture. 

Knowing how to integrate this new data and which ones to use can be intimidating and challenging – but if done well, it is rewarding and profitable.

One of the most significant new models I have seen in a while helps bridge the gap between data and insights, and turns multiple inputs into valuable forecasted outputs. It is a powerful model that even inexperienced forecasters and data scientists can use. Using this new methodology, you are almost certain to get the highest fitted forecast or r-squared with little effort or concern. This is a brief introduction into this new method.

It is called Auto Phantom Regression with Integrated Linear Forecasting Operation and Ordinary Least Squared Estimator.

Although the name of the model is long (I am sure they will eventually come up with a good acronym), the name really highlights exactly what it does. Imagine a scenario where you have many predictor variables, or don’t even know what variables exist and you’re not sure how to include them. Because there are so many predictor variables, you would like some help in creating a good model automatically. It will do this by trying and testing many phantom variables during the exploratory stages, building a regression forecast based on ordinary least squares to find the best fit.

The way it accomplishes this is by using a type of stepwise regression in that it selects a model by automatically, adding or removing individual real and dummy predictors, a step at a time based on their statistical significance. The end result of this process is a single regression model, which makes it nice and simple. What makes this model extra special is that every time the model adds or removes predictors based on a statistical test, you also invoke a phantom degree of freedom because you’re learning something from your data set but it does not show up as a degree of freedom.

These phantom degrees of freedom will not impact your number of observations per parameter estimate or affect the predicted R-squared. Instead, it allows the model to perform many statistical tests and try out many models based on actual and dummy variables until it finds a combination that appear to be significant and give you the highest r-squared.

Now there are some serious concerns and words of caution with this new revolutionary model. Firstly, when you try many different models and variables, you’re bound to find one that best fits the data but doesn’t fit reality. Second there is no magic bullet when it comes to building a model. And third, while there are components which are real and useful, I do not actually know of any Auto Phantom Regression with Integrated Linear Forecasting Operation and Ordinary Least Squared estimator model or, as it is known by its acronym, APRILFOOLS!

Lessons To Be Learned

You are training a model, not teaching it to memorize your data

You can learn a lot from trying different variables and bringing multiple data sets into your forecasting process. But be careful. When using regression models, a degree of freedom is a measure of how much you’ve learned. Your model uses these degrees of freedom with every variable that it estimates. If you use too many, you’re overfitting the model. The end result is that the regression coefficients, p-values, and R-squared can all be misleading and, while the model fits the data, it does not serve as a useful forecast.

Before throwing data about every potential predictor under the sun into your regression model, remember that it may not make it better. With regression, as with so many things in life, there comes a point where adding more is not better. In fact, sometimes not only does adding more factors to a regression model fail to make things clearer, it actually makes things harder to understand!

There is No Perfect Model

Yes, we absolutely need to start looking at various inputs to improve our forecast. There is still a place for conducting prior research into the important variables and their relationships to help you specify the best model. When you are using new variables, collect a large enough sample size to support the level of model complexity you need. Avoid data mining for what may work and keep track of how many phantom degrees of freedom you raise before arriving at your final model.

I don’t care if you are using a traditional time-series model or sophisticated machine learning algorithm, when you hear the words “best pick”, be cautious. If you evaluate your model on the same data you used to train it, your model could very easily have overfitting. To help avoid this, have a testing data set or time series hold out periods. This is part of the overall data set you set aside and use to provide an unbiased evaluation of a final model’s fit before putting it into production.

Finally, if a model seems too good to be true and over-sophisticated to the point that you do not completely understand it, it may not be the right model. There is no replacement for experience and knowledge, and learning what is right for your forecasting process.

For further information on choosing the right forecasting model, click here. Alternatively, pick up a copy of my new book, Predictive Analytics For Business Forecasting.

For insight into real predictive analytics models that identify meaningful casual relationships in your data, attend IBF’s Predictive Business Analytics, Forecasting & Planning Conference from April 20-22, 2021. 

The following is an extract from Eric Wilson’s new book, Predictive Analytics For Business Forecasting & Planning, written by Eric Wilson CPF. It is your guidebook to the predictive analytics revolution.  Get your copy. 

Most of what is discussed around predictive analytics is in terms of a new way to forecast where you are not just looking at your internal sales history, but also you are bringing in more data, different drivers, and other external variables to improve your forecast.

We have focused on using predictive analytics to help in discovering “Why do things happen?” and the benefits of translating this into “What could happen if…?” With advances in technology, we now have advanced models and methods that can better enable predictive analytics.

The Demand Planner or predictive analytics professional blends forecasting and business intelligence. They merge techniques and methods including machine learning to support the business’s needs. A common misconception is that machine learning, business forecasting, advanced business intelligence, and all things predictive analytics are synonymous.

The diversity of opinion reflects the fluidity of how we understand the defining language of the field. Business forecasting is the process to extract information and provide insights. Machine learning is a subset or application of AI and is more of an approach than a process. Business intelligence is the different types of analytics and outputs.

Where they overlap is the intersection of process, approach, and insights of predictive analytics.

Business Intelligence

Business Intelligence or BI focuses on infrastructure and output. The use of the term business
intelligence can be traced to the mid- to late-1860s, but it would be a century later before consultant. Howard Dresner was credited with coining the term. His definition was, “Business Intelligence: An umbrella term that covers architectures, databases, analytical tools, applications, and methodologies used for applying data analysis techniques to support business decision-making.” Despite this definition, historically, when people thought of business intelligence, most considered it as a fancy way of talking about data reporting. It has always been much bigger than just a dashboard and, through the years, people have begun to better understand the breadth and uses of BI to inform data-driven business decisions.

Fig b | Infographic depicting intersection of AI, BI, and business forecasting

Predictive analytics in BI has become a natural and needed progression of decision-making capabilities and insights. Where most of BI focused on visualization of data and descriptive type analytics, with predictive analytics we are asking more what could happen or even what we can make happen as an organization. Predictive analytics helps in presenting actionable information to help executives, managers, and other corporate end-users to make informed business decisions. Overall, predictive analytics can help discover why things happen and use this knowledge to reveal what could happen in future.

Business Forecasting

Business Forecasting: The process of using analytics, data, insights, and experience to make predictions and answer questions for various business needs. It is a process of breaking something down into its constituent elements to understand the whole and make predictions. Where BI is about the tools and representation, business forecasting is the analysis and procedures.

Predictive analytics in business forecasting has become a more advanced process that encompasses more and different types of data, more forward-looking causal type models, and more advanced algorithms and technology. It uses several tools, data mining methodologies, forecasting methods, analytical models (including machine learning approaches), and descriptive and predictive variables to analyze historical and current data, assess risk and opportunities, and make predictions.

Instead of just historical sales, we are trying to better understand the factors or the likely purchase behavior of the buyer. Predictive analytics is a new way to forecast where you are not just looking at your internal sales history, but you are bringing in more data, different drivers, and other external variables to improve your forecast.

Machine Learning

Machine Learning involves different approaches and methodologies. It is a subset of AI and is a
collection of different techniques, methods, modeling, and programming that allow systems to learn automatically.10 Machine Learning: An algorithm or technique that enables systems to be “trained” and to learn patterns from inputs and subsequently recalibrate from experience without being explicitly programmed. Unlike other approaches, these techniques and algorithms strive to learn as they are presented with new data and can forecast and mine data independently.

For predictive analytics, machine learning has opened new opportunities and provides more
advanced methods for it to use. Predictive analytics in machine learning is a category of approaches to achieve better forecasts, improved intelligence, automation of processes, and a path to AI. Some new advanced models and methods can be incorporated to further enable predictive analytics.

Predictive Analytics

At the intersection of advanced business forecasting, mature business intelligence, and some machine learning techniques, is predictive analytics. Predictive Analytics: A process and strategy that uses a variety of advanced statistical algorithms to detect patterns and conditions that may occur in the future for insights into what will happen.

Predictive analytics used to be out of reach for most organizations. However, recent advances
in professional skills, increased data, and new technologies, including machine learning and AI
techniques, have made it much more accessible. Predictive analytics utilizes many advanced business and planning processes to provide more information with less latency and improved efficiency. This is not just about advanced analytics outputs and business intelligence; it also offers more mature organizations a view of what and why things occur. Finally, while predictive analytics may use some machine learning techniques, it is only a portion of the planner’s toolbox, along with other statistical and data mining techniques.

This article is an extract from the book Predictive Analytics For Business Forecasting & Planning, written by Eric Wilson CPF. It is your guidebook to the predictive analytics revolution. Get your copy.

Business forecasting and predictive analytics are merging to leverage Big Data as a growth driver.

Predictive analytics does not have to be complicated and Demand Planners can learn these models and methods to drive business insight.

Organizational processes to support the application of predictive analytics insights are arguably a bigger challenge than the models. 

IBF spoke to Eric Siegel, author of Predictive Analytics: The Power To Predict Who Will Click, Buy, Lie, Or Die and former Columbia Professor, who revealed just what predictive analytics is and how it crosses over into business forecasting.

“Predictive analytics is basically applications of machine learning for business problems”, says Siegel. Machine learning learns from data to render prediction about each individual [thing being examined].” That individual thing can be a customer, product, machine, or any number of things.

When asked why predictive analytics is the latest evolution in information technology, Siegel responded “Because predicting by individual case is the most actionable form of analytics because it directly informs decision for marketing, fraud detection, credit risk management etcetera”.

But How Does Predictive Analytics Actually Work?

“Data encodes the collective experience of an organization so predictive analytics is the process of learning from that experience. You know how many items you sold, which of your customers cancelled, or which transaction turned out to be fraudulent.”

Siegel continued, “You know all this – that’s the experience, and you learn from that experience and the number crunching methods derive patterns. And those patterns are pretty intuitive and understandable. They could be business rules. For example, if a customer lives in a rural area, has these demographic characteristics and has exhibited these behaviors, then they might have a 4 times more likely chance of buying your product than the average.”

“That may be a relatively small chance but when improving something like mass marketing, finding a segment that is 4 times more like to buy than the average, that has a dramatic improvement on business performance.”

It is clear then, that by identifying patterns in data, predictive analytics can reduce risk and identify valuable commercial opportunities.

Predictive Analytics Meets Business Forecasting

“There is a continuum between forecasting and predictive analytics”, Siegel notes. But he does highlight key differences in their current applications:

• Forecasting is about a singular prediction, i.e., about sales in the next quarter or who will win a political election.
• Predictive analytics renders a predictive score for each individual whether it is a consumer, client or product, and as such provides insight into how to improve operations relating to marketing, fraud detection, credit risk management etc. more effectively.

Siegel laments the current disconnect between the two fields, “There should be a lot more interaction between what are two very siloed industries but have a lot of the same concepts, a lot of the same core analytical methods, and a lot of the same thinking. Both belong under the subjective umbrella know a as ‘data science’”.

Ultimately, both forecasting and predictive analytics serve to gain business insight but approach it from different starting points. Every business decision starts with a lag between what you know now and what occurs. Whether you’re forecasting sales or the likelihood someone will buy something in response to a marketing initiative, you’re generating a prediction.

Siegel said of the similarities between forecasting and machine learning, “the methods on the business application side include decision, trees, logistic regression, neural networks and ensemble models while forecasting uses time series modeling, but there are ways these two classes really do interact and really build on one another”.

Predictive Analytics Isn’t Scary

When challenged that complex predictive analytics methods can scare people off, Siegel insists that “they’re totally intuitive” and that machine learning and predictive analytics can be “accessible, understandable, relevant, interesting, and even entertaining”. That should reassure Demand planners looking to adopt predictive analytics methods and models.

Talking of the apparent complexity of machine learning models, Siegel commented that even neural networks, which represent the more advanced modeling on the predictive analytics spectrum, are modular and each if its components are in fact very simple.

Even if the model as a whole is difficult to fully understand (even for the people who invented them) you can test them and see how well they work, meaning that regardless of how complicated the models are to understand, their actual application is relatively straightforward.

Whether it’s through his Dr. Data YouTube channel (complete with rap videos), his book, or his Coursera program, Siegel is on a mission to make predictive analytics accessible. When it comes to the data that predictive analytics uses, he again highlights the simplicity, “It can be simple as a two-dimensional table on an Excel spreadsheet where each row is an example and each column is an independent demographic or behavioral variable”.

How Can Demand Planners Start Using Predictive Analytics?

It goes without saying that training in data science and predictive analytics is necessary when it comes to demand planners applying these techniques. Most of the training available on predictive analytics is technical, however, and that’s just part of equation warns Siegel, “There’s another side to machine learning if you’re going to make business value out of it which is the organizational process – the way you’re positioning the technology so it’s not just a cool, elegant model but is actually actionable and will actually be deployed.”

That’s a theme that Demand Planners will recognize all too well and it’ll come as no surprise that supporting process and culture are vital to leveraging predictive analytics insight in an organization, “Organizational requirements like planning, greenlighting, staffing, and data preparation are foundational requirements.”

Click to order your copy now.

One of the key themes raised by Eric Siegel is that forecasting and predictive analytics are merging to meet the business needs of today. To find out more about the future of these fields and how they impact demand planners and forecasters, check out Eric Wilson’s upcoming book, Predictive Analytics For Business Forecasting, published by the Institute of Business Forecasting, which is available to preorder now.

To get up to speed with the core concepts underlying predictive analytics, head over to Eric Siegel’s Machine Learning Course on Coursera.

If we can gain insights from just a small amount of internal structured data, then how much more could we glean from Big Data? I’m talking that external mass of structured and unstructured data that is just waiting to be collected and analyzed.

But there’s a balance between not enough data and too much. What’s the right amount of data to work with as demand planner or data scientist?

There is a debate about how much data is enough and how much data is too much. According to some, the rule of thumb is to think smaller and focus on quality over quantity. On the other hand, Viktor Mayer-Schönberger and Kenneth Cukier explained in their book Big Data: A Revolution That Will Transform How We Live, Work, and Think, that “When data was sparse, every data point was critical, and thus great care was taken to avoid letting any point bias the analysis. However, in many new situations that are cropping up today, allowing for imprecision—for messiness—may be a positive feature, not a shortcoming.”

The obsession with exactness is an artifact of the information-deprived analog era.

Of course, larger datasets are more likely to have errors, and analysts don’t always have time to carefully clean each and every data point. Mayer-Schönberger and Cukier have an intriguing response to this problem, saying that “moving into a world of big data will require us to change our thinking about the merits of exactitude. The obsession with exactness is an artifact of the information-deprived analog era.”

Supporting this idea, some studies in data science have found that even massive, error-prone datasets can be more reliable than simple and smaller samples. The question is, therefore, are we willing to sacrifice some accuracy in return for learning more?

Like so many things in demand planning and predictive analytics, one size does not always fit all. You need to understand your business problem, understand your resources, and understand the trade-offs. There is no rule about how much data you need for your predictive modeling problem.

The amount of data you need ultimately depends on a variety of factors:

The Complexity Of The Business Problem You’re Solving

Not necessarily the computational complexity, (although this an important consideration). How important is precision verses information? You should define this business problem and then select the closest possible data to achieve that goal. For example, if you want to forecast the future sales of a particular item, the historical sales of that item may be the closest to that goal. From there, other drivers that may contribute to future sales or understanding past sales should be next. Attributes that have no correlation to the problem are not needed.

The Complexity Of The Algorithm

How many samples are needed to demonstrate performance or to train the model? For some linear algorithms, you may find you can achieve good performance with a hundred or few dozen examples per class. For some machine learning algorithms, you may need hundreds or even thousands of examples per class. This is true of nonlinear algorithms like random forest or an artificial neural network. In fact, some algorithms like deep learning methods can continue to improve in skill as you give them more data.

How Much Data Is Available

Are the data’s volume, velocity, or variety beyond your company’s ability to store, or process, or use it? A great starting point is working with what is available and manageable. What kind of data do you already have? In Business-to-Business, most companies are in possession of customer records or sales transactions. These datasets usually come from CRM and ERP systems. A lot of companies are already collecting or beginning to collect third party data in the form of POS data. From here, consider other sources, both internal and external, that can add value or insights.

Summary

This does not solve the debate and the right amount of data is still unknowable. Your goal should be to continue to think big and work with what you have, gather the data you need for the problem and algorithm you have.

When it comes to gathering data, it is like the best time to plant a tree was ten years ago.  Focus on the data available and the insights you have today while building the roadmap and capabilities you want to achieve in the future. Even though you may not use it now, don’t wait until tomorrow to start collecting what you may need for tomorrow.

Thanks to Big Data, computational leaps, and the increased availability of analytics tools, a new age of data analysis has emerged, and in the process has revolutionized the planning field. With the explosion of data and the increasing desire to leverage it as a competitive tool, companies are moving from looking in the rear-view mirror to what is in front of them – and even charting their own paths.

In their book, Competing on Analytics, Thomas Davenport and Jeanne Harris describe the competitive advantage to degrees of information, or what they call intelligence. The authors divide these into two quadrants: those that are descriptive, or what I would call traditional or reactive, and those that are predictive, or what I would call revolutionary and proactive. Building on this we can further look at the progression from pure descriptive to past predictive to prescriptive and even what some call cognitive. As we continue along, the graph allows us to see what benefits we  each analytics type provide see (figure 1).

Figure 1:

As you up the X axis and along the Y axis, your competitive advantage increases. But wherever your processes land on the chart, all of these process and outputs are intended to support decision making. Depending on the stage of the workflow and the requirement of data analysis, there are five main kinds of analytics – descriptive, diagnostic, predictive, prescriptive and cognitive.  The five types of analytics are usually implemented in stages and no one type of analytics is said to be better than the other. They are complementary, and in some cases additive i.e, you cannot employ the more sophisticated analytics without using the more fundamental analytics first.

Success lies in reconciling all of these approaches within the same strategic framework. It is important to understand that all levels of analytics provide value whether it is descriptive or predictive, and all are used in different applications. That said, those that are truly leveraging analytics for competitive advantage right now are using predictive analytics, and it is this type of analytics that is driving the revolution happening today in demand planning.

Descriptive Analytics

This is simplest stage of analytics and for this reason most organizations today use some type of descriptive analytics. The easiest way to define it is the process of gathering and interpreting data to describe what has occurred.  For the most part, most reports that a business generates are descriptive and attempt to summarize historic data or try to explain why one event in the past differed from another. In addition to reports, some queries and classification processes can fall into the category of descriptive analytics. We can use advanced machine learning algorithms at this level for more complex data mining and clustering which helps us prepare data for other types of analysis.

Descriptive analytics takes the raw data and, through data aggregation or data mining, provides valuable insights into the past. However, these findings simply signal that something is wrong or right, without explaining why. For this reason, more mature demand planning functions do not content themselves with descriptive analytics only and prefer to combine it with other types of data analytics.

Diagnostic Analytics

At this stage you can begin to answer some of those why questions. Historical data can begin to be measured against other data to answer the question of why something happened in the past.  This is the process of gathering and interpreting different data sets to identify anomalies, detect patters, and determine relationships. Some approaches that uses diagnostic analytics include alerts, drill-down, data discovery, data mining and correlations. This can include some traditional forecasting techniques that uses ratios, likelihoods and the distribution of outcomes for the analysis. Supervised machine learning training algorithms for classification and regression also fall in this type of analytics.

Most Business Intelligence stops short of this stage and is stuck in just reporting KPI’s or historical data. Companies that employ seasoned demand planners go for diagnostic analytics as it gives in-depth insights into a problem and more information to support business decisions. At the same time, however, diagnostic analytics means we are reactive, and even when used in tandem with forecasting, we can only predict what existing trends may continue.

Predictive Analytics

Predictive analytics, broadly speaking, is a category of business intelligence that uses descriptive and predictive variables from the past to analyze and identify the likelihood of an unknown future outcome. It brings together a number of data mining methodologies, forecasting methods, predictive models and analytical techniques to analyze current data, assess risk and opportunities, and capture relationships and make predictions about the future. At this stage you are no longer just asking what happened, but why it happened, and what could happen in the future.

By successfully applying many traditional forecasting techniques to more advanced machine learning predictive algorithms, businesses can effectively interpret Big Data to gain huge competitive advantages. Unfortunately, most companies are still only scratching the surface of the capabilities of predictive analytics and operate solely in the green shaded area of Figure 1, stuck between “what happened” and “what could happen”. Such are the limitations of traditional business forecasting. They miss the bigger picture of predictive analytics being a new, better way to understand business. They haven’t realized that predictive analytics allows you to understand demand drivers and then use that knowledge to proactively respond to the market.

Prescriptive Analytics

Prescriptive analytics is the next step in the progression of analytics where we take:

• The data we gathered in the descriptive stage that told us what happened,
• Combine it with the diagnostic analytics that told us why it happened,
• Combine those with the predictive analytics that told us when it may occur again.

The result is prescriptive analytics that will highlight what you can now make happen. Prescriptive analytics is a combination of data, mathematical models, and various business rules to infer actions to influence future desired outcomes. Some refer to this as demand shaping but it can also include simulation, probability maximization and optimization.

Prescriptive analytics are comparatively complex in nature and many companies are not yet using them in day-to-day business activities. Admittedly, to consistently operate at this level of maturity, this requires new people, process and technology, and an analytics driven culture for the entire organization.  That said, if implemented properly it can have a major impact on business growth and be a competitive game changer. Larger scale organizations like Amazon, Target and McDonald’s are already using prescriptive analytics in their demand planning to optimize customer experience and maximize sales.

Cognitive Analytics

Wouldn’t it be nice if we could take all of the analytics and data and the software learns by itself without us telling it what to do; welcome to cognitive analytics.  Cognitive analytics brings together a number of intelligent technologies to accomplish this, including semantics, artificial intelligence algorithms and a number of learning techniques such as deep learning and machine learning. Applying such techniques, a cognitive application can get smarter and self-heal and become more effective over time by learning from its interactions with data and with humans. With this we may even begin to blur the boundary between the physical and the virtual worlds and automate processes and processing to bring new capabilities to demand planning.

[bar id=”8202″]

Eric will be speaking at IBF’s Predictive Business Analytics, Forecasting & Planning Conference in New Orleans from April 28-20, 2020. Learn more about the methods discussed in this article and how to leverage them as a competitive advantage. Includes special data science workshop.

What if we not only knew what the forecast will be for an item for the next period, but also understood a range of potential outcomes for that item?

What if we didn’t just have a number, but we also had a list of drivers that contributed to many possible numbers?

What if we not only had an outcome, but could use probabilities to discover unseen outcomes?

What if, instead of a single plan, we gave the business alternative scenarios that could play out?

What if we didn’t just forecast numbers but used predictive analytics to understand drivers, paint scenarios, and drive the demand you want?

“What if” Is The Question That Underpins Predictive Analytics

“What if” is perhaps one of the more critical and important aspects of predictive analytics. Predictive analytics and scenario planning allow a business to respond to alternative situations more quickly and effectively. Predictive analytics, using simulation techniques, can increase our knowledge and confidence in making informed decisions. Predictive analytics focused on forecast drivers is information that helps us shape our future by telling us what actions should be taken that will lead to desired business conditions.

The most basic part of forecasting is the assumption. As demand planners, assumptions are more important than numbers. Much of our job is managing them, interpreting them, and turning them into insights. Assumptions are numerous and help us break down complexity and uncertainty. Every business forecast contains assumptions.

Another term for assumption may be “scenario”. A scenario, in this context, is a potential circumstance or combination of assumptions that could have a significant impact (whether good or ill) on an organization. In the messy world of people and behavior, there can be no forecast without a scenario. The only question is whether to make your assumptions explicit (known) or implicit (unknown). You have a choice: pick a single assumption (usually a single number) or use predictive analytics to understand more variables and therefore more assumptions. The latter choice makes the variables known, and allows us to forecast more accurately.

One Choice, Multiple Outcomes

Scenario planning and predictive analytics are based on the premise that for every choice taken, there are several possible outcomes. By accurately identifying multiple variables that contribute to the forecast and preparing for each of these alternative scenarios, it is possible to be reasonably sure that the initial action was the correct one. This level of strategic foresight also allows for the creation of contingency plans that can be activated immediately, if the situation calls for action of that type.

By using predictive analytics and making the assumption known, it is possible to prepare in advance for the several potential outcomes

By using predictive analytics and making the assumption known, it is possible to prepare in advance for the several potential outcomes rather than simply meeting them as they come along. The advance preparation can often save a great deal of time and money, as well as provide the company with intelligence that helps to defuse negative situations while maximizing the benefit from positive ones.

At the core, this is what demand planning and predictive analytics does. Their job is to take the questions that seem almost unanswerable (due to their complexity and the many unknowns) and try to manage the assumptions and develop answers. Each of the questions involves dozens of factors that can change the ultimate outcome. To help, there may be some good analytical approaches to addressing the unknowns and breaking down the complexity posed by such tough forecasting questions.

More than numbers, demand planners manage assumptions and need to understand their individual contribution

We as demand planners live in the world of ambiguity and uncertainty and transform it into insights the business can use. More than managing numbers, we manage assumptions and need to understand their individual contribution. We use weighting and ratios and work towards the best fit of our data sets to the right model to minimize uncertainty and provide answers.

Our world is changing as well, and we need to adapt. Predictive analytics and probabilities just may be the train that is taking us into the future. We have already seen a shift from traditional time series modeling to predictive analytics due to omnichannel and e-planning, much of which is driven by regression models or even more sophisticated machine learning and probabilistic forecasting.

Predictive Analytics Is About Probability

One of the primary goals of predictive analytics is to assign a probability to forecast drivers. With these probabilities you can understand (as unlikely as it may be) the likelihood of the black swan event occurring, or indeed a variety of other more day-to-day outcomes. Predictive analytics can be used to create a number of different what-if scenarios, especially in the areas of risk assessment, customer buying trends, and business. For example, it can be used with a business’s sales history to determine when customers are most likely to make large purchases or which products will perform best. It can also be used in a market as a whole to get an idea of when a business could safely try to expand without taking unnecessary risks.

IBF spoke to Dr. Eric Siegel, author of the best-selling book Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die, and host of the Dr. Data Show on YouTube. He reveals why predictive analytics has become a critical best practice for all businesses and how – when misused – it can further marginalize society’s most vulnerable groups.

What is predictive analytics?

Let me give three definitions. The first is that it’s the business application of machine learning. The second, more common functional definition, is that it’s technology that learns from data to make predictions about an individual person or case in order to improve company performance. The third definition—where predictive analytics meets forecasting— is that it’s the deployment of machine learning methods such as decision trees, log-linear regression, ensemble models and neural networks. These methods serve to predict demand. When applied for forecasting, rather than generating a prediction for an individual customer, corporate client, or transaction, which is what predictive analytics is usually employed to do, it simply provides a forecast for the item and time range in question.

How does predictive analytics work?

Imagine you have rows of data available from which to learn. In the case of a classical business application, each row corresponds to a customer and contains the outcome or behavior about the customer you’d like to predict—such as whether they committed fraud, whether they canceled their subscription, whether or not they turn out to be a good credit risk, whether they defaulted on their debt, or whatever outcome is relevant to your business.

In the case of forecasting, each row includes or summarizes everything you knew up to this point in time about a particular SKU or whatever you’re trying to predict demand for, along with the outcome, i.e., what the actual demand ended up being over the next week or next month. These rows compose your training data from which correlations are discovered between the data points you hold about the product and the outcome.

Are machine learning and predictive analytics synonyms?

It depends on the context, but, when you’re talking about these kinds of business problems, they’re entirely synonymous. A lot of confusion surrounding this technology is due to nomenclature. A few years ago, the industrial world started using the word machine learning. Before that, the term was only common within academia and research and development.

When it comes to this technology and demand forecasting, you can just as well call it machine learning as predictive analytics. Whether we’re predicting the likelihood of somebody defaulting on a loan or predicting demand for a product, the same exact core analytical methods apply.

To complicate things further, what about prescriptive analytics?

Prescriptive analytics is a superfluous term, which has never been agreed on as far as its definition. The way it’s usually used is to imply that predictive analytics is not enough and now you need additional technology in order to translate what you’ve predicted into actual action to be taken. But predictive analytics has all along been intended to directly inform actions to be taken such as whether to make marketing contact, whether to audit for fraud, or approving someone for a credit card application. The term prescriptive analytics implies there’s an entirely new field or technology. There isn’t. Predictive analytics as a field already includes field deployment, i.e., the translation into action.

How does predictive analytics differ to more traditional statistical modeling?

Machine learning is statistical modeling. It’s just that we don’t use the word statistics because it sounds boring and it kind of undersells it because the core methods have advanced so much since the term statistical modeling originated. Predictive analytics is just the use of those methods for business applications. They all make use of historical learning cases, i.e., data, and they all find patterns that map from the things we do know to the things that we’d like to predict.

What impact has predictive analytics had on businesses over the last few years?

Predictive analytics has become a critical best practice for all businesses that operate at scale. It empowers businesses to execute on mass-scale operations more effectively. Business is a numbers game. Mass marketing, for example, is mostly just junk mail. Investigating fraud is mostly looking at transactions that are not fraud. Most resources and time in that sense are intrinsically wasteful.

The question is, how can you make all these mass operations more effective? By predicting. If you can predict for each individual case, that is, put a probability on it, then you know whether or not to risk spending $2 to send somebody a marketing brochure, or the risk of not auditing somebody for fraud. So predictive analytics generate probabilities, which allows us to manage risk. If you take a marketing campaign and improve its return on investment by a factor of three, that’s a valuable business outcome.

How can companies get started in predictive analytics?

E: There needs to be a champion at the company who becomes familiar with the field by reading or undergoing some preliminary training. Then it’s about defining the right pilot project. Where’s the lowest hanging fruit, the greatest opportunity? What’s the large-scale operation that occurs repeatedly— whether it’s a forecasting function or marketing or what have you—that could stand to benefit? Then you just need to define the scope for the project, do a preliminary analysis of the data, and, usually, engage with an external resource like professional services to get that first project off the ground.

What is big data and why is it valuable for businesses?

Big data and data science are simply umbrella terms that mean we have a lot of data and let’s do something useful with them. Big data did originally mean data that is too big for traditional database technology, but that’s not how it’s commonly used now. These days, both big data and data science allude more to a culture than any particular technology or value proposition. When we talk about using big data, we’re just saying, “We’ve got smart people who are working on using data to derive value”.

What kind of data do companies typically hold that could be used for predictive analytics?

The good news is that anything you do repeatedly is the exact thing that creates the data you need to improve that process. In marketing, it’s simple—you’ve got all the demographic profile and behavioral data about that customer, like what they’ve bought up to that moment and other data points. For forecasting, it’s very similar. Everything that’s been going on with the sales of this particular product up to this moment in time, as well as other external factors which you believe could influence demand, can be leveraged by predictive analytics.

One of the most famous cases of a company benefitting from predictive analytics was when Target marketed to women they identified as likely to be pregnant. How did they figure that out?

That’s an unusual case because normally in marketing applications, you’re predicting for a particular product or a set of products. This is a little different because they’re trying to predict something about the current state of the customer rather than their future behavior. If we think about training data with all the things we know about the customer, Target shored that up with data from customers who voluntarily told Target that they were pregnant. Then the analytics process ends up just being the same thing as for any of these other problems we’ve been discussing. They analyzed all the demographic and profile information about the individual customer and the purchases that they’ve been making, especially more recently, to identify which customers are likely to be pregnant. Then they market appropriate products to that group.

When it comes to implementing predictive analytics, is the challenge not so much building or accessing the technology itself but building a culture around it?

Yes. That’s a parallel challenge and is often the greater challenge. Predictive analytics needs to be seen as an organizational process and the business value needs to be the leading motivator. The actual deployment has to be an integral part of the project from inception. To that end, you need to lead with that value proposition. Let’s say we’re selling this idea internally to marketing. We tell them we’re going to market to only the top 30% predicted most likely to buy and in so doing, we’ll save 70% on marketing costs and here’s our estimated ROI.

These metrics are why we’re doing it—you don’t lead with machine learning or the analytical methods, you lead with a value proposition, the functional purpose of the project. Then, to the degree they need to how the underlying technologies work, you explain the technical aspects that we’ve been discussing.

Do businesses need Chief Data Officers?

I’m actually agnostic about that. It certainly can’t hurt in terms of improving visibility of the analytics function and encouraging a general culture of using data and finding ways to do it. Of course, it depends on the individual organization. The very first Chief Data Officer was at Yahoo. It turns out the job title of Chief Data Officer was first suggested as a joke and everybody in the room laughed. But it stuck and it became a whole new thing.

In your book you say it’s impossible to predict with high accuracy. In business, what kind of accuracy do you need?

Well, accuracy is not the right measure. Rather, you want measures that tell you how much better than guessing your predictions are. As previously mentioned, business is a numbers game.  You want to tip the odds in your favor by predicting better than guessing. That can have a dramatic effect even if it’s not “accurate” in the conventional sense.

For example, I can identify the 20% of customers that are three times more likely than average to buy, but it might be that those people overall have just a 6% chance of buying. I don’t have high confidence that any one customer’s definitely going to make a purchase. But knowing that group is three times more likely to buy than the average tells you that the model has learned something. Instead of prediction accuracy, we should ideally measure the value in business terms like return on investment or overall profit of the project or marketing campaign. If you can translate it into those terms, which you often can, that’s the number that matters, not the predictive performance of the underlying model.

There are now off-the-shelf predictive analytics/machine learning tools that are very cheap. Are these black box systems and can they be trusted?

Most of the software products out there are not black box. There are well-known standard machine learning methods such as the decision trees and neural networks methods we have mentioned. The software products have lots of different bells and whistles and variations to them, but more or less they use known methods or algorithms, and they don’t hide how they work. There are exceptions. There are some machine learning products that build branding around having a “secret special sauce.”

For more technical users and experts, including myself, that creates suspicion and raises questions regarding credibility. After all, if you want to sell a product to experts in the field, it doesn’t make sense to keep the core analytics a secret and market the product as some magical tool. Transparency, as well as cost, user interface, and how well it integrates with existing operational systems and databases should be the selling points.

Many people are still forecasting in Excel when these tools are now easily accessible. What do you think is stopping widespread adoption?

There’s a lot of complexity in moving from Excel to proper machine learning. It’s a completely new endeavor and you have to get together training data, apply methods that may be new to you, and then be able to interpret the outcome, results and performance. It’s a big project. The thing is, there has to be a business reason or incentive to do it. How much ROI would there be if your forecasts were 10% better, for example, and how much is that worth to you?

What is the persuasion paradox and why is it important?

The paradox is that, although we can’t perceive or record individual cases of causation or influence, we can predict it. You can’t conclusively observe persuasion, even in retrospect. If you look at a customer, why did they buy the product? Was it because of your marketing initiative? You can never be entirely sure because they might have bought the product anyway. But we can predict the likelihood of causation over large numbers of individuals.

On the other hand, in forecasting, observing influence is generally is not a concern. We just want to know how many widgets are going to be sold next week. We don’t care why we’re going to sell a lot of widgets next week. We just want an accurate estimate of sales.

What are some examples of unethical uses of data?

The first one that comes to mind is using a protected class such as the race, gender, or religion of an individual in order to derive a consequential decision such as whether they get approved for a loan or sentencing in prison. Let’s imagine that a manager is interviewing a candidate for a job, and they use predictive analytics to gauge the candidate’s likelihood of success in the position. The manager might reject the candidate based on the results provided by the system. When the candidate asks the manager for the basis of their decision, they would have to say, “Well, for one thing, our predictive model penalized you by seven points because you’re black.” Normally, protected classes are disallowed as direct inputs, but there are experts, including university professors involved in the legal justice system, campaigning to purposefully introduce then as direct inputs. The risk is real.

What excites you about predictive analytics?

I’m a former academic and I think that technology that learns from examples and, therefore, from experience to draw generalizations that hold in future cases is the most interesting kind of any science and technology. It’s the most exciting science period. The fact that it naturally ends up being valuable only makes it that much more exciting.

To read the full interview with Dr. Eric Siegel, download the latest issue of the Journal of Business Forecasting. When you become an IBF member, you’ll get the Journal delivered to your doorstep quarterly. Get your IBF membership.