Business Intelligence Archives

Go Beyond Metrics

Often, our clients come to us believing that they already apply analytics to their business, when in fact, they are actually doing a fantastic job at applying metrics. Metrics provide measurements of variables, while analytics explain the relationships between variables. Although both metrics and analytics work together to provide you with the insights necessary to grow your business, metrics and analytics are not synonyms. To date, metrics have dominated the business intelligence field in the form of dashboards and scorecards. Dashboards display metrics as the measurement of key variables. For this reason, they are very useful for monitoring current business activities and results. While metrics are still very important, analytics goes beyond measurement to provide explanations of the deep relationships behind business metrics. Instead of reporting that something happened, analytics seeks to explain why something happened, when events can be expected to occur, and how we should act to create certain desirable outcomes.
The shift towards analytics is growing in importance because today’s executives need to be able to scale their decision-making in an increasingly complex world with little room for error. While many executives understand the value of data-driven decision making, decision-makers can become overwhelmed when faced with the volume and complexity of available data provided by metrics and dashboards. Analytics provide value to by cutting through the clutter and complexity and providing decision-makers access to the most imports variables, the relationships between variables, and even the accuracy of the analytic model. When using analytics, executives are able to avoid sorting through non-information, and instead, focus on the information which drives their best decision-making.

July 19, 2011 Nate Watson News 5 comments

Presenting Business Intelligence

At CAN we spend a lot of time developing better ways to present and communicate our ideas, because our systems only produce results if people understand how to use our insights. Presenting business intelligence is based on the business question, the audience and what is important. The following is an explanation of how we decide how to present:

Select a Business Question

We start every project by defining the business question to be answered. We do not worry about what techniques or methodology we are going to use to answer the questions. Instead, we determine what is the most important question to be edited. We only start a project once our client is able to answer “What question, if answered, would have the largest impact on your business in the next 6 months?”. (Related Post on ROI)

Select an Audience

Who you are presenting to determines what information you present and in what order. There are three presentation styles, but they depend on who you are talking to. The graphic above outlines each presentation.
When presenting to generalists you want to start with the conclusion and then explain how you reached that conclusions. By starting with the conclusion, you provide generalists, who are typically decision makers, a filter to understand and develop questions during the rest of your presentation.
When presenting to experts, you want to start your presentation by providing background information, and then explain step by step how you reached your final conclusion. This is useful because the experts want to know that you know what you are talking about, and presenting background information before you reach conclusions will help you establish trust with the experts.

Justify Each Element

The best presentations start on time, end early and leave plenty of time for questions. The key to success is to curate your presentation by justifying the existence of each sentence, graph and data table. If you can not provide a good explanation for something you included in your presentation, you must take it out.

Focus on Results

After a successful presentation, your audience will know exactly what the next step is; who they need to call, what they need to buy and what they need to do. If there is no clear action they must take after hearing your presentation, then your research and presentation have been wasted.
Learn how you can use business intelligence to make data-driven decisions.

June 30, 2011 Nate Watson News 3 comments

Zombie Marketing: Go from Infection to Infestation

At CAN, we love trying to figure out why things work or don’t work. This often leads to continuous experimentation. I decided to unleash this post about Zombie Marketing because it is our latest marketing strategy, it is lethally effective, and, well, it’s about ZOMBIES!

For full-out zombie apocalypse, the key is for the infection to reach critical mass before the authorities are able to exterminate the Differently Animated. For a Zombie outbreak to take over the world you don’t announce it to the masses. Instead, you infect a few key people and let the virus spread quietly. In sales and marketing, as with Zombies, the key is to select the right people to contact, but the ‘right people’ aren’t who most would people usually think of.
When most people think about influential people, they tend to focus on the leaders of social groups, people at the center of their social group with many people listening to them. Although they are the center of their social group, leaders are very difficult people to infect because they typically have a specific platform to protect, and their attention is fragmented because so many people want their attention. Even if you manage to infect a leader without first reaching an overwhelming critical mass, infecting them won’t have as much impact as if you even had already established a small but concentrated infected user-base. A prime example of using leaders as infectious agents is when you get a celebrity endorsement for a product with an established user-base, or a celebrity endorsement for a brand new product that no one has ever used.
Instead of infecting leaders, the key to a successful Zombie Marketing strategy is to infect the influencers in a group. Influencers are people on they edge of many social groups, they don’t have a platform, and they move easily from one social group to the next infecting people with their ideas. Influencers easily develop horizontal friendships, which are friendships between people with different interests, geographies, races, ages, education backgrounds, income levels, and lifestyles. People with horizontal friendships are especially valuable because they are constantly being exposed to new ideas, perspectives, products, and people outside of what would be considered their typical sphere of influence. (Check out a related post on the value of horizontal friendships)
Since influencers are infinitely curious and open to new friendships and ideas, I find them quite enjoyable and relatively easy to infect. However, they require a different method of infection compared to leaders and the masses. To infect leaders you have to get recognized, with the masses you have to push your message, and with influencers you have to allow them to discover something novel. People become influencers because they enjoy discovering new, novel and previously undiscovered ideas.

Focus Your Content on Solutions to Specific Needs: Influencers enjoy connecting people to solutions to their ideas. In Zombie Marketing you want to make spreading the infection as easy as possible. You want influencers to know that if they find someone with a specific need you have the solution.
Focus Your Analytics on the Infection: You want your focus to be on developing content that infects influencers, and your content will be guided by the feedback that you receive. So don’t focus on visits or clicks. Instead focus on time on site, conversion, comments and loyalty.
Provide Consistent Value: When people invest their money or time in something they expect to get consistent results. If you want people to keep coming back to your site you want to provide them with consistent value. This doesn’t mean that you have to post everyday, but that if you post once a week you continue to post once a week and preferably on the same day.
Promote your Work Sporadically during Off Times: While you need to continually develop great content, promoting your work sporadically and during off times can work in your favor, at least until you gain a significant user base. Not posting all of your content encourages people to keep checking back in with your website to get the latest updates. Post during off times, like Friday, Saturday and Sunday night when most people aren’t posting and yet their are still people hungry for new content.
Let the Virus Run It Course: In addition to being patient, you need to develop content that is viable over a long period of time. This means that it should be memorable and solve a need that people will have today, tomorrow and a year+ from now. That way once you infect someone with your ideas, they are useful and contagious for as long as possible.

Disclaimer: Zombie Marketing might not be for you if you operate in a perfectly competitive market place with many sellers, many buyers and an undifferentiated product. However, if you operate in a monopolistic competitive market places then a Zombie Marketing strategy might be the right one for you.

June 25, 2011 Nate Watson News 1 comment

Refocusing Business Intelligence on Client Needs not Wants

When my partner, Tadd Wood, and I started Contemporary Analysis we decided that we were going to focus on developing a great final product, instead of focusing offering our products at a low cost or focusing great customer service. Not that our prices or our customer services is lacking, but that they are not our focus. We did this because I learned from owning previous businesses that people care about price before they sign the contract, they care about customer service when the product is being delivered, but once the final product is delivered all they care about is the quality of deliverable.
We identified three key principals that we use when building our final deliverables:

We focus on providing answers to our clients most important business questions.
Our methodology is designed to provide reliable answers to real businesses.
We provide answers to our clients’ questions so that they get immediate tasks to improve their business in addition to greater understanding.

The core of these principals is to focus on our clients needs and not their wants. This has been really difficult because what our clients need is answers to their business questions. While it is easy to get distracted by flashy tools, cool visualization and interactive interfaces, at the end of the day our client’s just want answers to their business questions. This is why we decided to refocus business intelligence as presented in the business industry.

June 18, 2011 Nate Watson News 1 comment

How to Build an Analytical Culture

CAN was started in 2008, to make predictive analytics applicable and accessible to businesses of almost any size. To fulfill our mission we have become experts in helping our clients adopt predictive analytics. The first step in adopting predictive analytics is to start building an analytical culture. In an analytical culture, people:
Know the Numbers: Everyone should know what is happening and whether certain strategies were successful. The objective of knowing the numbers is not to judge individuals or the team, but learn from the past and refine future attempts at innovation.
Examine the Numbers: In order to effectively discuss and examine the numbers, team members should be familiar with the numbers, their assumptions, and their application.
Use the Numbers: The ultimate goal of adopting predictive analytics is to make better decisions. This is the hardest and most complex phase of becoming analytical. However, when achieved, this stage provides the highest Return on Investment (ROI). It requires participating members to understand factors driving the numbers, historical paradigm shifts, and the assumptions used to make decisions. Once the factors and assumptions are understood, they can be utilized properly to optimize processes and predict outcomes of specific decisions.
The Analytical Portal
The key to building an analytical culture is to make analytics easy to use, understand, and accessible. The easiest way to do this is to develop an analytical portal. The purpose of the analytical portal is to allow everyone in your organization to know, examine, and use the numbers without having to contact the accounting and information technology departments every time a report or numbers are needed. Your analytical portal should allow people to have access to 5 business intelligence tools:
Dashboards will be used to quickly and easily report key metrics. The key to building successful dashboards is to choose the correct metrics and provide quick and easy access. Dashboards as well as scorecards will allow team members to see the big picture and drill down into finer details on a need to know basis. Dashboards differ from slicing & dicing data in that dashboards are predetermined visualizations of key metrics that are the same across individuals and time.
Detailed Reports are to be read before using analytics to make decisions. These are designed to ensure that statistics, equations, metrics and other reports analytics are fully understood. This will ensure that recommendations are properly implemented.
Slicing & Dicing Data allows your employees to ask questions about the data and get quick responses. People should be able to quickly pull any data that they want from a variety of data sources such as the census, economic data, news sources via text mining, social media, surveys, and internal databases.
Predictive Analytics allows your employees to understand the relationship between outcomes and the variables that drive them. The results of predictive analytics can be used to simulate the outcome of specific decisions and create strategies to maximize outcomes.
Alerts & Notifications are the watchdogs of analytics. They provide a dependable and timely way to bring awareness to team members about important events, but are not distracted by smaller, less significant changes. If properly implemented, managers will spend less time monitoring employees and more time addressing key issues. Alerts & Notifications can be sent via email, dashboard widget or text message.

June 18, 2011 Nate Watson News No comments yet

Why Predictive Analytics is Important

CAN was started in 2008 , to make predictive analytics applicable and accessible to businesses of almost any size. We are working diligently because predictive analytics has the unique potential to allow executives to scale their decision making as organizations and decisions become increasingly more complex with ever thinner margins for error. For example small businesses are now having to compete around the world as if they were multinational companies, and multinational companies are having to develop and deliver products and services to ever more fragmented consumers as if they were small businesses.
Introduction to Predictive Analytics
Predictive analytics is different than other Business Intelligence tools because it goes beyond visualizing data and human assumptions, instead it combines data, theory and math to make forecasts about the future using current and historical facts. While executives and decision makers also analyze current and historical facts to make forecasts about the future, what separates executives from predictive mathematics is predictive analytics ability to find detailed relationships from hundreds of possible variables and millions of observations, and also produce an estimate of the error of the prediction. Predictive analytics is able to tell you that a independent variables, e.g. males, has a .3% impact on a dependent variable with 95% confidence. While executives and managers still have to make the final decisions, predictive mathematics can provide insight that isn’t possible with other business intelligence technologies.
A Brief History of Predictive Analytics
It wasn’t until the middle part of the last decade that we had the combination of mathematical techniques, data storage capacity, data processing power and data creation that we need to make predictive analytics accessible to businesses of all sizes. The mathematical techniques have existed and improved since the founding of the Econometric Society in 1930. The data storage capacity has been around since 1977 with Oracle’s commercialization of the relational database. The processing power has existed since IBM commercialized business computing with the IBM 360. SaaS and Social Media provided the final requirement, data creation. Both of these technologies have made is common and inexpensive for people to quickly and easily collect and publish information about their lives and jobs.
The Limitation of Predictive Analytics
Predictive analytics is not without flaws and if people become over reliant or trusting of predictive models you can have disasters such as the subprime mortgage crisis. To avoid hubris and ensure that our customers can trust our models, CAN has developed the philosophy of Less Wrong and the CAN Three-Way Test.
The philosophy of Less Wrong is that the goal of every model that we produce is to make our clients less wrong than they were, but not to assume that we will ever make our clients 100 percent right. In business their are often clear winners with clear metrics such as profit keeping score. However, no business is ever perfect, the winners are only Less Wrong than their competitors. Also, business success and failure is only temporary. The ability to help companies continually become less wrong is what attracted CAN to predictive analytics. By combining data, theory and mathematics CAN helps develop predictive models to help our clients make decisions that are Less Wrong.
Our philosophy of Less Wrong has allowed us to develop business model that allows us to quickly and inexpensively provide results to our clients. Similar to Software as a Service (SaaS) we provide Analytics as a Service on month-to-month contracts, and we use our ongoing relationships to continually help our clients becomes Less Wrong. While we provide our clients with results within the first 30 days, we are continually improving and adapting our systems to an ever changing business environment.
While the philosophy of Less Wrong helps to establish reasonable goals and allow room for continual improvement, our models provide the foundation of our client’s business, and so we work hard to make sure that our models reflect reality so that they generate reasonable predictions. One of the best ways that we do this is the CAN Three-way test. The CAN Three-way test requires that each model we produce can be backed up with theory, data and math.

The Theory: Without a sound theory, you have to rely on randomness, and randomness decreases the long-term usefulness of the model. CAN relies on the expertise of our clients executives, managers and employees to build the theories around our models.
The Data: CAN makes sure to fully understand your data, because data is the raw materials used to build predictive models. CAN will examine patterns in your data to determine which equations to use to model your data, and also determine the quality of your data. CAN is able work with imperfect data, we just make sure that we understand the flaws and limitations of your data, such as potential bias etc.
The Math: CAN’s analysts go through hundreds of different models to select the right one that fits the data as well as the theory. This doesn’t mean that we always select the equation that has the best line fit, but instead the equation that fits both the data and the theory.

June 13, 2011 Nate Watson News 3 comments

Insurance Client Stories-Contemporary Analysis

CAN has been fortunate to help a number of Life and Health Insurance companies work smart. It has been a great relationship because clients from Life and Health Insurance companies understand what we do as they are experts in using mathematics to forecast human behavior on a biological side, and CAN is an expert in applying predictive analytics to forecast human behavior on a marketing and sales side. Our customers invest in developing new products, and we use math to make sure they can distribute them most effectively. The following are some of the solutions we have developed:
Marketing ROI with Terrain: Our clients wanted to know the ROI of their marketing campaigns, but had been struggling to develop a system that was able to determine the impact that their marketing was having on sales. We agreed that it would be too difficult to explain the impact that a company’s marketing was having and instead we used CAN’s Terrain system to explain the impact that the economy and seasonal fluctuations had on sales. With every predictive model their is a residual x-factor that can not be explained in detail. So instead of including the economy, a major driver of purchasing decisions, in the X-factor, we included the company’s marketing in the X-factor. While they did not exactly accomplish their goal of measuring the ROI of their marketing activity, we were able to explain how economics and seasonality impacted sales and the residual could be attributed to marketing, competitors and other factors. Now our clients can get a reasonable idea if marketing is impacting sales.

Client Management: Using Pulse, CAN helped our clients understand and improve their client portfolio by modeling the factors of client profitability, loyalty and claim activity. Each week we provide our client with a list of the clients that are at risk of leaving within the next 90 days, and a list of sales leads that have a high probability of being profitable and loyal clients. Each month we provide them with a report on their client portfolio relative to the population and their goals.

Sales Management: Using Beacon, CAN has worked to connect agents and brokerages to our clients home offices. Beacon has been used to focus the development of leads and marketing materials to best serve the needs of agents and brokerages. Beacon also has been used to improve communications with training, regulations and promotions between home offices and agents by segmenting agents by learning styles and providing communication guidelines to our clients.

June 13, 2011 Nate Watson News No comments yet

The Steps to Implement Predictive Analytics

The following is the step-by-step process that CAN would use to implement predictive analytics into your business systems:

Step 1

We select from our catalog of intellectual property which mathematical formula or survey questions will be most effective to explain the activity being studied, and develops a list of potential dependent and independent variables.

Step 2

We will meet with individuals on your team to learn more about your business, employees, and customers. This meeting is intended to be a discussion to determine the independent variables that are most likely to be significant and their correlation to the dependent variables. To encourage discussion, before the meeting, participants will be given a list of proposed dependent and independent variables, a description of the hypothetical model, and how the results will be applied.
At the beginning of the meeting, the we will give a brief introduction of how the hypothetical model will be developed as well as an overview of the proposed dependent and independent variables. After the introduction, one member of our team will work to moderate the discussion and take notes. The outcome of the meeting with be a list of:

Important Historical Events: Understanding historical events help to identify outliers and their cause(s), such as errors in the data, changes in business paradigms, or significant events outside of the business.
Assumptions: No model can be perfect, and thus require a certain number of assumptions. Assumptions are non-conscious theories about how the world works. It is important to understand the assumptions so that results can be accurately produced and interpreted.
Dependent Variable: The Dependent variable represent the desired outcome that is being studied. The predicted dependent variable is explained by the mathematical mixing of the independent variables. The value of the dependent variable will change whenever the independent variables change, and that is why it is called the dependent variable.
Independent Variables: Independent Variables are variables that are hypothesized to have an impact on the dependent variable. When the independent variables change, the dependent variable changes as well.

Step 3

After the second step, the our team will collect the data for the dependent and independent variables from internal and external sources. CAN will work with your team to securely transfer the data from your internal sources. If the necessary data is not available CAN will develop a survey. Once the data has been collected it is loaded into the CAN database, cleansed, and connected to the modeling environment.

Step 4

Contemporary Analysis will study the data and present an overview to your team to ensure that we are correctly interpreting the data.

Step 5

The data will then be run through mathematical models to determine which variables have the most significant impact on the activity being studied. Models are selected based on statistical robustness, standard error, and how well they represent the data.

Step 6

A research report will be developed to provide information on the variables and why they were selected, the quality of the data, the quality of the model, and suggestions on how the research can be applied. The report is developed both as a comprehensive study and an executive summary to facilitate those who are non-specialists, but want to gain an understanding, and for those requiring a greater level of detail.

Step 7

CAN will meet with your team to present the results of the research, the report, and suggestions of how results can be implemented. The CAN team will also work with the management team to break down results into specific and actionable tasks.

Please contact us to learn how we can help implement predictive analytics into your organization.

May 30, 2011 Nate Watson News No comments yet

How to Structure a Survey

We rely on existing data when available, however sometimes we have to collect primary data using surveys. Over the years we have developed many surveys — including a bi-lingual survey to measure the employment and consumption preferences of Omaha’s first generation latino population. The following outlines the basics of our methodology for how to structure a survey:

Incentives: The best surveys use social incentives instead of monetary incentives. Using social incentives requires that surveys are designed to make it easy for people to feel the satisfaction of giving their opinions and helping people. This means that surveys should be short with provided answers that clearly and completely answer the survey questions.
Questions: Every question needs to designed to provide the most information for the least amount of effort from the participant. This requires that only essential “business questions” are researched, and that each response provides multiple dimensions of information.
Answers: The best surveys provide answers that are able to completely answer the survey questions, because respondents will lose motivation if they can not truly express themselves. This doesn’t mean that you create pages of possible answer, instead it should focus how the question is answered to limit the possible answers. The format of the answers should stay consistent as much as possible so respondents don’t have to re-read the questions, and instead can focus on understanding the question and answering appropriately.
Security: Surveys are valuable because they provide people a protected and ambiguous line of communication. When respondents feel safe they are able to provide information that they wouldn’t disclose in normal communication.

Please let us know if we can help your collect and analyze data for your organization.

May 30, 2011 Nate Watson News 3 comments

Keeping Business Intelligence Simple

CAN does more than optimize your sales, marketing, customer services, management and strategic planning; we optimize how you learn. Our job is not to provide you with more information, data or work. Our job is to provide you with supported recommendations that help you run your business more effectively. CAN helps you with business intelligence to get the insights you need so you can get back to your life.
CAN keeps business intelligence simple by:

No Extra Data Collection: CAN’s systems are designed to only require access to your accounting data, because it is the most accurate data that companies already have. Any additional data is provided by CAN either through surveys or by using CAN’s databases with data on economic variables, demographics, psychographics, real estate, consumers, and businesses.
Implementation: CAN systems are designed to be simple enough that businesses without computers could implement our systems, and businesses with more advanced infrastructure are able to implement using a minimal amount of effort. We did this because we wanted the success of our systems to be independent of a company’s IT department, software trends, or industry whims.
Reporting: We have resisted the temptation to wow you with complicated visualizations and data tables, and instead CAN provides you with just the necessary information to answer your important business questions.