Bronfman company builds bridges between artificial intelligence & business intelligence
Quote from Timothy Fitzpatrick on July 3, 2022, 20:20Jul 1, 2022,07:00am EDTZohar Bronfman is the CEO and co-founder of Pecan.ai, a predictive analytics platform built to solve business problems.
Every company gathers data and lots of it—customer data, market data, competitor data and industry data. Cloud systems, software as a service (SaaS) and business intelligence (BI) tools process zettabytes of data each year. But how many companies are able to make the best use of this data using the tools and teams they have today?
Data can be a company’s most valuable asset, providing the basis for predicting everything from future revenue to buying behavior and customer retention. Many companies have well-established BI teams that review and analyze historical data for performance and management trends. But when companies want to move beyond traditional historical analysis to incorporate predictive analytics and artificial intelligence (AI), they face challenges in finding the talent and tools they need. Data scientists are hard to hire and are trained to focus more on research and model accuracy than on specific business results.
But is there a way to bridge the gap by evolving business analyst teams into a new breed of AI analysts? After all, BI teams have many important strengths: They know the business, know what’s important to the stakeholders and the lines of business they support, and they understand the data they’re working with better than anyone else. And although they aren’t as statistically experienced as data scientists with building and maintaining predictive AI models, there are technological innovations that can help bridge these data science knowledge gaps.
When businesses want to use the data, tools and teams they’ve already built today to start generating more useful predictions about the future, how should they prepare? And what steps can they take to prepare to use their data to make accurate AI-based predictions?
The core of this challenge is bridging the chasm between data science and BI. Both domains analyze data to propel the business forward, but they each have their own strengths and limitations.
Classical BI is well understood: It’s mainly focused on interpreting past events and trends and presenting them in easy-to-digest aggregated reports and dashboards. A limitation of BI is that the insights generated are usually hypothesis-driven, meant to explain why a particular trend or behavior happened in the past by looking at a large segment of people sharing the same common denominator. Without the right level of machine learning (ML), BI isn’t equipped to provide precise nongeneralized, hyper-granular insights down to the individual customer level.
At the same time, built-in human bias in selecting which variables or data points to analyze can also limit or influence the results with preconceptions while ignoring the unexpected insights that an ML system can uncover. Those patterns of behavior only AI can see. Finally, most BI teams don’t have the deep statistical analysis training needed to implement predictive modeling. That’s where data science is supposed to help.
So far, data science has often fallen short of that promise for many businesses. Predictive and prescriptive models are hard to deploy, and most projects never make it to production. At the same time, companies are challenged to quantify the business impact that their ML and AI investments have generated.
To address the shortage of talent and the disconnect between data science and business priorities, there are new advanced analytics solutions that can help companies leverage the business analytics talent they already have. Business analysts typically work closely with specific departments or lines of business, so these professionals know how their organizations capture data and how they create and measure business value. Many of today’s business analysts are eager to have access to automated statistical analysis, machine learning and data cleansing so they can focus on interpreting and applying predictive models that provide more value to the company.
BI teams know the data and what’s important for the business. So, ask them these questions: What metrics are you looking to improve? Are you trying to grow revenue, reduce churn or increase customer lifetime value? These different goals will point to unique approaches to analyzing data.
Adding AI to BI data moved analytics from looking at the past in aggregate to predicting the future of an individual customer and highlighting marketing opportunities. There are many questions this could help answer: How often should a mobile game publisher offer a specific promotion to a player to bring them back to the game? How much discount should an e-commerce company offer to win back a customer who hasn’t made a purchase in the past two months but whose predictive lifetime value puts them in the VIP category? If the customer is 90% likely to return on their own, should the marketing team spend their marketing dollars to retarget them or divert the funds to a different program or campaign?
BI can only show you that there’s a connection between players and customers receiving special offers and returning to play or buy again—but that connection only reveals that people like free stuff and discounts. It doesn’t tell us which customers will really like a particular offer at a specific moment in the future. Instead of making the same offers to a large cohort of people, predictive intelligence can identify which customers are most likely to return on their own and which need the nudge of a promotion. With this information, a company can target its marketing to the specific customers who will respond best to this nudge at the right time.
Business efficiency anchored in precision and automation is key to gaining and maintaining scale, especially in times when resources are limited by challenging market conditions. Predictive models provide a glimpse of customers’ future, and by bringing business intelligence and data science together, they could become accessible to many companies. The chasm between data science and business analytics needs to close if we want to maximize the opportunities for very capable, data-rich BI teams to bring more value to the enterprise.
In part two of this article, I’ll describe a few specific steps companies can take to prepare for these opportunities.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Zohar Bronfman is the CEO and co-founder of Pecan.ai, a predictive analytics platform built to solve business problems. Read Zohar Bronfman's full executive profile here.
Source: https://www.forbes.com/sites/forbestechcouncil/2022/07/01/building-a-bridge-between-artificial-intelligence-and-business-intelligence-to-maximize-business-outcomes-part-one/?sh=6585150cf186
Zohar Bronfman is the CEO and co-founder of Pecan.ai, a predictive analytics platform built to solve business problems.
Every company gathers data and lots of it—customer data, market data, competitor data and industry data. Cloud systems, software as a service (SaaS) and business intelligence (BI) tools process zettabytes of data each year. But how many companies are able to make the best use of this data using the tools and teams they have today?
Data can be a company’s most valuable asset, providing the basis for predicting everything from future revenue to buying behavior and customer retention. Many companies have well-established BI teams that review and analyze historical data for performance and management trends. But when companies want to move beyond traditional historical analysis to incorporate predictive analytics and artificial intelligence (AI), they face challenges in finding the talent and tools they need. Data scientists are hard to hire and are trained to focus more on research and model accuracy than on specific business results.
But is there a way to bridge the gap by evolving business analyst teams into a new breed of AI analysts? After all, BI teams have many important strengths: They know the business, know what’s important to the stakeholders and the lines of business they support, and they understand the data they’re working with better than anyone else. And although they aren’t as statistically experienced as data scientists with building and maintaining predictive AI models, there are technological innovations that can help bridge these data science knowledge gaps.
When businesses want to use the data, tools and teams they’ve already built today to start generating more useful predictions about the future, how should they prepare? And what steps can they take to prepare to use their data to make accurate AI-based predictions?
The core of this challenge is bridging the chasm between data science and BI. Both domains analyze data to propel the business forward, but they each have their own strengths and limitations.
Classical BI is well understood: It’s mainly focused on interpreting past events and trends and presenting them in easy-to-digest aggregated reports and dashboards. A limitation of BI is that the insights generated are usually hypothesis-driven, meant to explain why a particular trend or behavior happened in the past by looking at a large segment of people sharing the same common denominator. Without the right level of machine learning (ML), BI isn’t equipped to provide precise nongeneralized, hyper-granular insights down to the individual customer level.
At the same time, built-in human bias in selecting which variables or data points to analyze can also limit or influence the results with preconceptions while ignoring the unexpected insights that an ML system can uncover. Those patterns of behavior only AI can see. Finally, most BI teams don’t have the deep statistical analysis training needed to implement predictive modeling. That’s where data science is supposed to help.
So far, data science has often fallen short of that promise for many businesses. Predictive and prescriptive models are hard to deploy, and most projects never make it to production. At the same time, companies are challenged to quantify the business impact that their ML and AI investments have generated.
To address the shortage of talent and the disconnect between data science and business priorities, there are new advanced analytics solutions that can help companies leverage the business analytics talent they already have. Business analysts typically work closely with specific departments or lines of business, so these professionals know how their organizations capture data and how they create and measure business value. Many of today’s business analysts are eager to have access to automated statistical analysis, machine learning and data cleansing so they can focus on interpreting and applying predictive models that provide more value to the company.
BI teams know the data and what’s important for the business. So, ask them these questions: What metrics are you looking to improve? Are you trying to grow revenue, reduce churn or increase customer lifetime value? These different goals will point to unique approaches to analyzing data.
Adding AI to BI data moved analytics from looking at the past in aggregate to predicting the future of an individual customer and highlighting marketing opportunities. There are many questions this could help answer: How often should a mobile game publisher offer a specific promotion to a player to bring them back to the game? How much discount should an e-commerce company offer to win back a customer who hasn’t made a purchase in the past two months but whose predictive lifetime value puts them in the VIP category? If the customer is 90% likely to return on their own, should the marketing team spend their marketing dollars to retarget them or divert the funds to a different program or campaign?
BI can only show you that there’s a connection between players and customers receiving special offers and returning to play or buy again—but that connection only reveals that people like free stuff and discounts. It doesn’t tell us which customers will really like a particular offer at a specific moment in the future. Instead of making the same offers to a large cohort of people, predictive intelligence can identify which customers are most likely to return on their own and which need the nudge of a promotion. With this information, a company can target its marketing to the specific customers who will respond best to this nudge at the right time.
Business efficiency anchored in precision and automation is key to gaining and maintaining scale, especially in times when resources are limited by challenging market conditions. Predictive models provide a glimpse of customers’ future, and by bringing business intelligence and data science together, they could become accessible to many companies. The chasm between data science and business analytics needs to close if we want to maximize the opportunities for very capable, data-rich BI teams to bring more value to the enterprise.
In part two of this article, I’ll describe a few specific steps companies can take to prepare for these opportunities.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?