Archive for January, 2017

January 30th 2017

Wise Practitioner – Predictive Analytics Interview Series: Halim Abbas at Cognoa

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, Early Screening for Autism By Halim Abbas IMAGE 2Combining Question-Based and Video-Based Predictors at Predictive Analytics World for Business Chicago, June 19-22, 2017, we asked Halim Abbas, Vice President of Data Science at Cognoa, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: Our models screen for autism and other developmental disorders among young children. Close to 300,000 users have completed the Cognoa questionnaire, and this data helps refine and improve our models.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: Predictive analytics power our clinically-validated developmental screening, giving parents peace of mind and enabling them to seek clinical diagnoses and behavioral therapies sooner.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: Our most recent clinical study shows a 100%+ lift in specificity over traditional autism screening methods.

Q: What surprising discovery or insight have you unearthed in your data?

A: We discovered that screening for autism in children younger than four is not only more attainable than previously thought, but even more accurate than with older children, provided that the right signals and behavioral clues are accounted for.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: Predictive analytics on clinical datasets is much more feasible when inconclusive determination is allowed for the cases that are harder to discern using automated systems alone.

———————

Don't miss Halim’s conference presentation, Early Screening for Autism By Combining Question-Based and Video-Based Predictors, on Wednesday, June 21, 2017 from 11:40 am to 12:00 pm at Predictive Analytics World Chicago. Click here to register to attend.

By: Eric Siegel, Founder, Predictive Analytics World

 

No Comments yet »

January 27th 2017

Wise Practitioner – Predictive Workforce Analytics Interview Series: Ben Taylor at HireVue

By: Greta Roberts, Conference Chair, Predictive Analytics World for Workforce 2017

 

In anticipation of his upcoming Predictive Analytics World for Workforce conference presentation, Predicting Safety and Performance from Unstructured Data, we interviewed Ben Taylor, Chief Data Scientists at Ben Taylor IMAGE 2HireVue. View the Q-and-A below to see how Ben Taylor has incorporated predictive analytics into the workforce of HireVue. Also, glimpse what’s in store for the new PAW Workforce conference, May 14-18, 2017.

Q: How is a specific line of business / business unit using your predictive decisions?  How is your product deployed into operations?

A: We have companies using us to accelerate their candidate workflows by concentrating top talent to the top of the queue. One of those examples is with Hilton where they were able to reduce their hiring cycle from 6 weeks to 6 days: https://www.hirevue.com/customers/hilton-hirevue. We have many different sectors including flight attendants, customer service, engineering, and even driver safety using this digital assessment approach. With any predictive solution the saying still holds true "Garbage in = garbage out". In regards to the digital assessment we still see a wide range of performance. Customers who have great questions and performance labels tend to very well, where customers who lack structured interview integrity and have suspect performance labels struggle. Not even the best technology in the world can fix your bad data problem, so data integrity is key start thinking about that now.

Q: If HR were 100% ready and the data were available, what would your boldest data science creations do?

A: I love the data science side of things and I have seen too many examples where the human element brings negative consequences (racism, similarity bias, delays, etc…). We all think we are great at hiring when really in reality we are not because our personal observations are limited and we struggle to generalize our experiences well. My boldest data science creations might include "auto-hire" where the computer is so confident in the candidate's assessment that a hiring decision is made automatically. Imagine having someone start working for you that you have never met, but they are great. I think this is possible someday and we will see this happen first in the high volume low risk positions (i.e. call centers, etc…).

Q: When do you think businesses will be ready for "black box" workforce predictive methods, such as Random Forests or Neural Networks?

A: Never. That might surprise you that I would say that being a machine learning advocate, let me clarify. So default ensemble methods and deep learning nets are black boxes to customers. For expert data scientists they are not and actually tell some great stories behind their features and decisions. So I think the ownership here is not on businesses or industries being forced to adopt these methods, I think the responsibility falls on the data scientists to ensure that approachable stories are being told from these models. I think during the next 3 years you will see a lot of progress with the story telling side where these black box methods will become much more open. As that happens I think all businesses will be ready for this new technology.

Q: Do you have suggestions for data scientists trying to explain the complexity of their work, to those solving workforce challenges?

A: There is a huge language barrier here. I would suggest hiring an IO Psychologist, someone who is much more familiar with bridging that gap and communicating value to workforce challenges. I remember being surprised that the person I was talking to on the phone didn't appreciate the R-value I was talking about from the cross-validation study. Most of what I was talking about was unfamiliar jargon with no clear mapping to actual business metrics. Convincing someone you are smart is no way to sell a product; you have to convince them you have value in terms they understand.

Q:  What is one specific way in which predictive analytics actively is driving decisions?

A: Screening incoming candidate talent is labor intensive process that can be overwhelming for large or popular companies. There are many companies actively using predictive analytics with resume modeling, assessments, mobile games, or video interviews to help concentrate candidates.

Q: How does business culture, including HR, need to evolve to accept the full promise of predictive workforce?

A: Data integrity. The business culture, regardless if they are using predictive analytics now, needs to focus on their data integrity. Are your interviews structured? Are you asking the right questions? Do you trust your performance metrics? How is your intra/inter-rater reliability on these performance scores? Making sure you have great data collection and integrity will ensure that when you are ready that you can get the most benefit.

Q:  Do you have specific business results you can report?

A: Hilton was able to reduce their hiring cycle from 6 weeks to 6 days: https://www.hirevue.com/customers/hilton-hirevue using our predictive interview scores. We have an internal customer support role within HireVue where we think we can reduce our hiring cycle to 1 day, or even automate the hiring decision completely. I am of course pushing to automate the hiring decision in this case because I see that as a historic milestone for HR analytics.

——————

Don't miss Ben’s conference presentation, Predicting Safety and Performance from Unstructured Data, at PAW Workforce, on Wednesday, May 17, 2017, from 3:30 to 4:15 pm. Click here to register for attendance

By: Greta Roberts, CEO, Talent Analytics, Corp. @gretaroberts and Conference Chair of Predictive Analytics World for Workforce

No Comments yet »

January 25th 2017

Wise Practitioner – Predictive Workforce Analytics Interview Series: Andrew Marritt at OrganizationView GmbH

By: Greta Roberts, Conference Chair, Predictive Analytics World for Workforce 2017

 

In anticipation of his upcoming Predictive Analytics World for Workforce conference Andrew Marritt IMAGE 2presentation, The Joy of Text: Building Actionable Models with Perceptions, we interviewed Andrew Marritt, Founder and CEO at OrganizationView GmbH. View the Q-and-A below to see how Andrew Marritt has incorporated predictive analytics into the workforce of OrganizationView GmbH. Also, glimpse what’s in store for the new PAW Workforce conference, May 14-18, 2017.

Q: How is a specific line of business / business unit using your predictive decisions?  How is your product deployed into operations?

A: We have an employee feedback tool called Workometry which provides executives the ability to ask open questions on any topic to their employees. We process the many tens of thousands of open-text responses, usually in multiple languages to, identify issues and target segments to direct action. We think of it less as a survey and more as scaling qualitative methods like interviews. As well as typical employee surveys it’s being used by executives to ask key questions on a range of topics from post-merger integrations, improving customer experience to understanding issues in supply chains.

The tool is also used by People Analytics teams to bring perception data into their models. If we think of the model pipeline typically Workometry will automate much of it and then the most advanced teams will include the metadata we produce for other purposes. Often clients will get us to build those models.

Our view is that it’s better to guide executives to where they need to take action rather than just present data. We use probabilistic and predictive models to do this.

Q: If HR were 100% ready and the data were available, what would your boldest data science creations do?

A: I like to think many of the boldest, cleverest data science innovations should be invisible, or at least the user shouldn’t need to know that you’re using ML. When we presented at SwissText last year one of the Google team were talking about how they ‘curate’ information to answer key questions in search these days. The amount of understanding of natural language sentences is extensive but as users we just get the answers we want.

Q: When do you think businesses will be ready for "black box" workforce predictive methods, such as Random Forests or Neural Networks?

A: I think they are. Personality tests are for the most part black boxes and most in HR are comfortable with them. The issue is that in most decisions you have to optimize against multiple variables. The issue isn’t really the model but the loss function. Whilst we can be rational and define what is best for the company — and the company can probably afford to be rational — the manager often has different goals. Because everyone’s loss function differs they need to be able to interpret the model to make effective decisions.

If the loss function is simple and one which is widely understood it’s easier to use black box methods.

Q: Do you have suggestions for data scientists trying to explain the complexity of their work, to those solving workforce challenges?

A: We do a lot of work teaching empirical decision making to HR managers to enable them to have better conversations with the data scientists. This is a huge help but currently quite rare. We’re also working with graphic designers and data journalists to help identify and tell stories. One of the statements I frequently use is that the most important part of communicating with data is the communication, not the data.

Q:  What is one specific way in which predictive analytics actively is driving decisions?

A: One simple way that we use is to predict variables at a group level, and then compare that with the actual. This can be used to ensure that managers don’t chase after ‘issues’ that are probably at a natural level. For example we’ll often build a predictive model to estimate the level of engagement, or eNPS for a team based on the demographics of the people in the team. We then can guide managers to groups which differ significantly from the expectation. These are usually groups where another variable which wasn’t in the model is causing the difference. These alternative factors are usually ones that you need to address.

Q: How does business culture, including HR, need to evolve to accept the full promise of predictive workforce?

A: I think the thing that people struggle with most is thinking probabilistically. Weather forecasters have struggled with this for a long time. As analysts if we over-promise I suspect that we’ll build disillusionment. I wouldn’t be surprised to see this in 2017 in HR. Predictive modelling isn’t about helping make decisions that are right — it’s about ensuring those decisions are less wrong, or optimizing on the impact of those decisions.

Q:  Do you have specific business results you can report?

A: Sure. With a big financial services firm we built an attrition model of one of their country businesses. We included perception data and a reasonably sophisticated loss function. What we found was that the way to optimize the impact of attrition wasn’t to minimize the attrition. There were significantly different drivers for attrition that affected different groups of the population. In fact some of the factors that would have probably done most to reduce the overall level of attrition would have likely increased the attrition of some of the most valuable segments. What the model, aligned to the loss function let us do was say ‘you need to address these issues in these teams,’ meaning HR could be very specific and targeted in their interventions.

——————

Don't miss Andrews’s conference presentation, The Joy of Text: Building Actionable Models with Perceptions, at PAW Workforce, on Wednesday, May 17, 2017, from 11:30 am to 12:15 pm. Click here to register for attendance

By: Greta Roberts, CEO, Talent Analytics, Corp. @gretaroberts and Conference Chair of Predictive Analytics World for Workforce

No Comments yet »

January 23rd 2017

Wise Practitioner – Predictive Workforce Analytics Interview Series: Sue Lam at Shell

By: Greta Roberts, Conference Chair, Predictive Analytics World for Workforce 2017


In anticipation of her upcoming Predictive Analytics World for Workforce conference Sue Lam IMAGEpresentation, Using Analytics in Recruitment to Identify Improvement Opportunities, we interviewed Sue Lam, HR Diagnostics Manager at Shell. View the Q-and-A below to see how Sue Lam has incorporated predictive analytics into the workforce of Shell. Also, glimpse what’s in store for the new PAW Workforce conference, May 14-18, 2017.

Q: How is a specific line of business / business unit using your predictive decisions?  How is your product deployed into operations?

A: Shell recruitment receives over 100,000 applications for their global graduate programme each year for fewer than 1,000 technical and commercial positions. The foundation of the recruitment process is the assessments because they provide data and insight on which to make fair and unbiased selection decisions. In order to stay competitive in the market with job seekers, Shell recruitment wanted to create a streamlined assessment approach. HR analytics and assessment specialists collaborated to review and analyze the current graduate assessments to understand which areas should be kept and which areas could be streamlined. The goal of the project was to enhance candidate assessment data, boost candidate experience, leverage technology and make structural changes that improve cost effectiveness, scalability and efficiency for variable hire demand levels. Using data analysis of assessment and performance data, we identified areas ripe for change and the new methodology will launch in early 2017.

Q: If HR were 100% ready and the data were available, what would your boldest data science creations do?

A: I work in the employee engagement and leadership assessment space and it would be helpful to understand what personality and environmental characteristics are most related to sustained business performance in our organization in light of the volatile business environment that we are currently operating in.

Q: When do you think businesses will be ready for "black box" workforce predictive methods, such as Random Forests or Neural Networks?

A: Companies need to have clean and accurate data, be fully comfortable with data analysis, and trust that the outcomes are sound before moving forward with them. Some industries will likely be ready sooner than more traditional or conservative industries. However, I think human intervention will always be needed regardless of what methods are employed. For example, a predictive hiring model may be very sound analytically, but it may increase adverse impact in hiring (e.g., hires only men, people of a certain age group). Humans need to be able to intervene to ensure all workforce predictive methods are fair.

Q: Do you have suggestions for data scientists trying to explain the complexity of their work, to those solving workforce challenges?

A: Statistics can appear to be complicated, full of jargon, and scary to HR professionals who aren’t accustomed to working with data all the time. To solve workforce challenges, the biggest hurdle for data scientists to conquer is to get senior stakeholder buy-in. Without executive buy-in, any recommendations from analytics projects will fall flat. We need senior leaders to champion our causes to ensure that a project goes from analytics to action. The first step to getting buy-in is to craft a compelling story about the results. To put it bluntly, most businesses will not care about the details of the analysis so explain the story without jargon.  It helps me to think about explaining the results to a friend of mine who doesn’t know anything about statistics. I try to use real-world analogies and cut out any jargon related to the analysis and I stick to using percentages when I talk about the outcomes. People are typically comfortable with percentages and they’re easily understood, so they can help to drive home a compelling point. Using charts, graphs, and other visuals can help with the complexity as well.

The most important part about getting your results across is to tie them back to business outcomes. Many times, I see really interesting analytics projects but the recommendations aren’t clearly linked back to the bottom line. Finally, always answer the question of “what’s in it for me?” for your customer. Your customer has not come to you with an analytics project just because they think it is interesting or they want to try out a new methodology. They likely want to solve a practical business problem that they have. Before starting an analytics project, understand what questions your customer is trying to answer and what they plan on doing with the results. Find out what will happen if the results that you find are contrary to your customer’s views. For example, will a program or process be dismantled? Will they lose budget or scope? Will it create a lot of work or the mobilization of many employees? Understanding your customer’s motivations will help you as an analytics professional to better craft a story and provide recommendations. If the results are contrary to your customer’s thinking, provide explanations for this and follow-up actions. I see statistics as a tool to help me have a more meaningful conversation with the business and not something to be feared. The analysis is just a starting point.

Q:  What is one specific way in which predictive analytics actively is driving decisions?

A: Predictive analytics is helping businesses understand the best way forward. If the business has a number of different choices that it can take, it would be most fruitful to understand which has the highest likelihood of success. Subject matter experts can provide input on areas of interest and data scientists can test hypotheses based on this expertise.

Q: How does business culture, including HR, need to evolve to accept the full promise of predictive workforce?

A: I think the business and HR can benefit from being more open-minded about what data tells us (and what it doesn’t). Some organizations may be better at looking at data and making logical decisions from the data, but generally speaking, there is a culture of mistrust when data is involved, particularly if the results don’t match our views. When the analysis reveals outcomes that are contrary to our beliefs (e.g., we should change this process that has been around for a long time because there is a better way of doing things), people tend to get defensive. People may get defensive because they’re worried about how it will appear to other people if their work is being challenged or changed. In situations like this, rather than being defensive, business and HR professionals would benefit from being curious about what produced that outcome and what could be improved easily. To support these behaviors, a company needs to have an innovation mindset, where there is no culture of blame.

Similarly, for analytics professionals, it is important to understand why the business and HR may be hesitant to move forward with your recommendations. Conducting analytics is one thing, but to mobilize people and resources to change a process requires patience and understanding the limitations placed on the business when making changes.

——————

Don't miss Sue’s conference presentation, Using Analytics in Recruitment to Identify Improvement Opportunities, at PAW Workforce, on Wednesday, May 17, 2017, from 10:40 to 11:00 am. Click here to register for attendance.

By: Greta Roberts, CEO, Talent Analytics, Corp. @gretaroberts and Conference Chair of Predictive Analytics World for Workforce

No Comments yet »

January 16th 2017

Wise Practitioner – Predictive Workforce Analytics Interview Series: Mike Rosenbaum at Arena

By: Greta Roberts, Conference Chair, Predictive Analytics World for Workforce 2017

In anticipation of his upcoming Predictive Analytics World for Workforce conference presentation, Real World Lessons in Predicting Employee Retention and Engagement, we interviewed Mike Rosenbaum, Founder and CEO at Arena. View the Q-and-A below Mike Rosenbaum IMAGEto see how Mike Rosenbaum has incorporated predictive analytics into the workforce at Arena. Also, glimpse what’s in store for the new PAW Workforce conference, May 14-18, 2017.

Q: How is a specific line of business / business unit using your predictive decisions?  How is your product deployed into operations?

A: Our predictions are primarily used by recruiters and hiring managers in hospitals and senior living facilities to identify the applicants who are most likely to provide certain outcomes, like stay in their role, be an engaged employee, provide high quality care, increase patient satisfaction, or be involved in a medical incident. The benefits of these outcomes are primarily felt by a number of business units, including nursing, patient care, food and nutrition, and housekeeping. Our platform is integrated with the client's Applicant Tracking System (ATS) and potential employees are asked to interact with us through a portal that is also hosted on our platform. We use the application data, some limited third party data, and the candidate's behavior on the platform to customize our models for each client, location, and role and to generate predictions for each applicant.

Q: If HR were 100% ready and the data were available, what would your boldest data science creations do?

A: At Arena our mission is to use data to illuminate talent. We aspire to transform the way the healthcare labor market works in a way that makes employers more efficient and employees more fulfilled. We feel that our focus on hiring is a great place to start, and we are expanding through the employee lifecycle to address areas such as team assembly, promotion decisions, time and attendance, and incumbent attrition. Ultimately, we aspire to help organizations transform themselves to address the challenges of a rapidly changing environment. For example, today health care delivery is going through massive changes, with much of the services that have been provided within the four walls of a hospital moving to clinics, retail outlets, offsite medical labs, and ambulatory surgical centers.  Instead of continuing to have people do work that is no longer needed; we investigate whether they might be a good fit for new the roles that are needed.

Q: When do you think businesses will be ready for "black box" workforce predictive methods, such as Random Forests or Neural Networks?

A: In our experience, business people are more interested in the results than in the predictive methods. Our clients are not as interested in reviewing the models or the statistical techniques as they are in seeing how changing their behavior and decisions will affect their outcomes. Of course we continue to internally investigate additional predictive methods to improve accuracy and outcomes.

Q: Do you have suggestions for data scientists trying to explain the complexity of their work, to those solving workforce challenges?

A: The work of a data scientist is undoubtedly complex, but we have found that it is rarely effective to try and explain that complexity to a business person. We have found a focus on results to be the most persuasive way to convince a business person to use predictive analytics. Many times a simple before and after comparison is enough to get over the first hurdle, and providing a well thought out business case with investments, benefits, and return on investment solidifies the case. Sometimes it can help to use an analogy, like credit scoring or voice recognition or book recommendations to show how complex predictions can easily become a part of simple every day decisions. We also find that giving clear guidance on how to use the predictions can help with adoption; many of our predictions are expressed as percentiles, so explaining that the predictions are mean to rank candidates so the most likely to provide the outcome will be the highest, and the least likely will be the lowest.

Q: What is one specific way in which predictive analytics actively is driving decisions?

A: Many of our clients have open positions that attract dozens, if not hundreds, of applicants. The traditional approach to this would be to have recruiters or hiring managers review every applicant and use their individual judgement to decide which are the best to engage in a hiring conversation. Our platform is being used to replace these judgements (which typically contain biases) with predictions to help them engage with the applicants that are most likely to provide the best outcomes. And by using our platform our clients are also able to remove the personal biases and judgements from their hiring process.

Q: How does business culture, including HR, need to evolve to accept the full promise of predictive workforce?

A: While many parts of the business are comfortable with data and analytics, HR is often behind the curve in these capacities. The typical HR employee does not have ready access to data, does not perform their own analysis, and may not be familiar with the key performance indicators that are used to measure their performance. In order to fully recognize the benefits of predictive analytics, the HR workforce itself will need to develop competencies in data and analytics. Luckily they are not the first to make this journey; their colleagues in marketing have been making a similar journey over the last several years and provide an excellent roadmap.

Q: Do you have specific business results you can report?

A: At Arena our clients use our platform in 400+ organizations to process over 4 million applications per year. The median reduction in first year employee turnover all of our clients has been 38%, and when compared to control groups (either other roles in the same facility or the same roles in other facilities) the median improvement is 162%. Because we have a 100% success rate in improving retention, it is easy for us to provide a guarantee to our clients, and so we provide a guarantee to all clients that if we do not reduce employee turnover by 10% in our initial implementation we will refund all money paid to us.

——————

Don't miss Mike’s conference presentation, Real World Lessons in Predicting Employee Retention and Engagement, at PAW Workforce, on Wednesday, May 17, 2017, from 3:30 to 4:15 pm. Click here to register for attendance

By: Greta Roberts, CEO, Talent Analytics, Corp. @gretaroberts and Conference Chair of Predictive Analytics World for Workforce

No Comments yet »

January 13th 2017

Wise Practitioner – Predictive Analytics Interview Series: Darryl Humphrey at Alberta Blue Cross

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, Claim Pattern Anomalies – Making a Mole Hill Out of a Mountain at Predictive Analytics World San Francisco, May 14-18, 2017, we asked Darryl Humphrey, Senior Data Scientists at Alberta Blue Cross, a Darryl Humphrey IMAGEfew questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: Our scope is health, dental, and pharmacy benefit claims submitted by plan members and providers.  Our objective is to reliably gauge the probability that a claim, or series of claims, is / are fraudulent or represent abuse of the benefit plan.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: Obtaining strong evidence of fraud or plan abuse most often requires an on-site investigation and other labor intensive activities.  The result of our analytics materially increases the probability that these efforts will have a positive ROI.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: We have increased the financial recovery per investigation as the analytics indicates which of the behavioral measures are anomalous which facilitates more specific lines of investigation.

Q: What surprising discovery or insight have you unearthed in your data?

A: Random forest analyses indicates that some behavioral measures long-held to be important actually don’t contribute much to the differentiating provider claiming patterns.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: Developing an in-house capability (analytics skills, business experience, and tools) has been more cost-effective than using a third-party and is providing a greater analytics depth, breadth, and agility.

———————

Don't miss Darryl’s conference presentation, Claim Pattern Anomalies – Making a Mole Hill Out of a Mountain on Wednesday, May 17, 2017 at 3:30 to 4:15 pm at Predictive Analytics World San Francisco. Click here to register to attend.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

January 9th 2017

Wise Practitioner – Predictive Workforce Analytics Interview Series: Feyzi Bagirov at 592 LLC and Harrisburg University of Science and Technology

By: Greta Roberts, Conference Chair, Predictive Analytics World for Workforce 2017

In anticipation of his upcoming Predictive Analytics World for Workforce conference presentation, Enhancing the Quality of Predictive Modeling on College Enrollment, we interviewed Feyzi Bagirov, Chief Business Officer at 592 LLC and Analytics Instructor at feyzi-bagirov-imageHarrisburg University of Science and Technology. View the Q-and-A below to see how Feyzi Bagirov has incorporated predictive analytics into the workforce of 592 LLC and Harrisburg University of Science and Technology. Also, glimpse what’s in store for the new PAW Workforce conference, May 14-18, 2017.

Q: How is a specific line of business / business unit using your predictive decisions?  How is your product deployed into operations?

A: One of the ways enrollment departments in higher education are using data science is  identifying students who are most likely to enroll, less likely to enroll and unlikely to enroll. This helps prioritizing marketing efforts. 

Q: If HR were 100% ready and the data were available, what would your boldest data science creations do?

A: Identifying and hiring the right student candidates.

Q: When do you think businesses will be ready for "black box" workforce predictive methods, such as Random Forests or Neural Networks?

A: When businesses will answer these three questions to themselves: 

     1. What questions need to be answered to achieve our objectives? 

     2. What data do we need to answer them? 

     3. How do we get that data?

Q: Do you have suggestions for data scientists trying to explain the complexity of their work, to those solving workforce challenges?

A: Ask them to talk about their business and look for the pain points. Once identified, give a 10,000 feet overview of how the data insight can help in making a decision. 

Q:  What is one specific way in which predictive analytics actively is driving decisions?

A: A monetary outcome (either making or saving) or a public/private benefit of a decision. 

Q: How does business culture, including HR, need to evolve to accept the full promise of predictive workforce?

A: Managers need to realize that:

  1. Data science insights will SUPPORT, but will not MAKE the decisions for them
  2. Utilizing organizational data is more than running descriptive dashboards. Getting into a predictive component quickly is important.
  3. Very few Data Scientists know everything about Data Science. 
  4. Maintaining data quality is important if you want quality insights. Sacrificing data quality for the sake of moving forward needs to be an exception.

——————

Don't miss Feyzi’s conference presentation, Enhancing the Quality of Predictive Modeling on College Enrollment, at PAW Workforce, on Wednesday, May 17, 2017, from 10:15 to 10:35 am. Click here to register for attendance

By: Greta Roberts, CEO, Talent Analytics, Corp. @gretaroberts and Conference Chair of Predictive Analytics World for Workforce

No Comments yet »

January 3rd 2017

Wise Practitioner – Predictive Analytics Interview Series: Craig Soules at Natero

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, Using Predictive Analytics to Improve Customer Retention, at Predictive Analytics World San Francisco, May 14-18, 2017, we asked Craig Soules, CEO & Founder at Natero, a few questions about his work craig-soules-imagein predictive analytics.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: Natero helps its customers by predicting two kinds of potential customer behaviors.  The first is customers who are likely to churn or stop using a given service.  The second is customers who are likely to upsell or purchase more of a given service.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?
 
A: Predictive analytics is a key way in which our customers decide which customers to reach out to and work with.  By focusing on the customers who are likely to change their use of the service (either churn or upsell), they can have the most positive effect on the health of their business.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?
 
A: Customers using Natero have been able to reduce their customer churn by as much as 24% month-over-month.  Although churn reduction is ultimately achieved through the efforts of the customer success team and their engagement with the customer's needs, knowing which accounts to spend time with is a critical factor in spending those efforts wisely.  Predictive analytics play a key role in driving their attention and efforts in the right ways.

Q: What surprising discovery or insight have you unearthed in your data?
 
A: One surprising discovery is the role that individual user data plays in understanding account churn.  A lot of customer success teams today rely on high-level metrics such as DAU and MAU to understand account health, but those are almost never enough to be truly predictive.  In the end, the behaviors of individual users and the changes in those individual behaviors are often required to build accurate models of churn outcomes.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: Predictive models really need to be tuned not just to the use case, but to the individual scenario.  As such it's critical to gather feedback from the users of the model results on an ongoing basis to continue to tune those results toward the specifics of their use case.

———————

Don't miss Craig’s conference presentation, Using Predictive Analytics to Improve Customer Retention, on Tuesday, May 16, 2017 from 10:55 to 11:15 am at Predictive Analytics World San Francisco. Click here to register to attend.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »