Essential Analytics for Solo and Small Business Owners

In which I will briefly explain the purpose, goals, and skills which will be covered in Section 2.

Topics include: laying the groundwork for the rest of the course: choosing a data set, simple data visualization techniques, data set characterization, and identifying outliers; and a specific psychedelic example of the lingo you will have mastered at the end of the section.2.

In which I appeal to students to broaden their thinking on what to measure and track, and provide many ideas and angles for consideration.

Topics include: why it's an important first step; selecting something other than profit as a key indicator; a list of considerations that may help you choose the right positive and negative measures unique for your business; and the value of starting simple.

In which I talk about practical considerations of data collection such as clear definitions, verifiability, repeatability, and convenience.

Topics include: clarifying what makes a good data set vs. a poor one; explanation of objectivity and subjectivity as applies to data collection; and practical considerations for setting up a data collection approach.

In which I explain what characterization is, and why it's important to invest effort into creating a model of how a business behaves.

Topics include: basics of building a model; step-by-step instructions, real-world examples, and exercises to support learning; letting the data speak.

In which the student will realize that variability is ubiquitous, and we are all used to compensating for it. I'll explain how it both hides and reveals useful information.

Topics include: an illustration of the variability in my morning commute translated into statistical terms; some examples of variability revealing and alternately hiding information; our human predisposition to misinterpret data; and what understating your business variability profile buys you.

In which I will provide a common-sense example of variability, and illustrate some ways to look at graphical data to extract clues from it.

Topics include: a simple example for visualizing data, and interpreting the clues found in its variability; and a contextualized first peek at run charts, trend lines, and scatter plots.

In which I will demonstrate some basic charting skills in Excel: run charts, and scatter plots, and trend lines. I will also talk about visually and mathematically identifying data correlations.

Demonstration includes: step-by-step instruction for creating a scatter plot and from a column of data in Excel; creating a run chart on the same data set; interpreting the charts; how to identify outliers; two different types of trend lines; tips on improving chart readability; simple conditional formatting; the correlation function; and a real world example from Sassy Granny's Smoothie Emporium.

In which the student may practice and master the charting techniques previously demonstrated.

Exercises include: an invitation for the student to repeat the charting demonstration steps from the previous lecture; three separate exercises covering trend lines, conditional formatting, and correlation strength. Answer key is provided.

In which I provide step-by-step instructions on how to install and access the free Analysis Toolpak add-on (Excel).

Topics include: step-by-step installation instructions demonstrated; where in the toolbar to access its functionality once it's installed; and how to find additional help if you need it.

In which I will explain the concept of frequency distribution (shape of a data set), and talk about key insights it provides about a data set. Topics include: definition of frequency distribution; what understanding your data's shape buys you; different types of distributions (what they look like, and specific examples of what kind of data sets might fall in each category); discussion of symmetry, asymmetry, and skew including how to determine the amount of skew in your data set; and what your outlier count indicates

In which I explain, illustrate, and extol the virtues of the histogram, a special type of chart that helps identify the frequency distribution of a data set.

Topics include: the definition of histogram, and what it looks like; why it's the tool of choice in determining data set frequency distribution; how it differs from a bar chart; and two simple and practical examples - one with candies, and one with product manufacturing efficiency.

In which I demonstrate step-by-step how to create a histogram of a data set in Excel.

Demonstration includes: step-by-step instruction on how to quickly find the minimum and maximum values in a data set, and the data set range; tips on choosing an appropriate histogram bin count; a comparative illustration of using differing numbers of bins and its effect on the ease of interpreting results; some simple chart adjustment tips; and several iterations of using the histogram data analysis function in Excel.

In which the student may practice and master the histogramming techniques previously demonstrated.

Exercises include: an invitation for the student to repeat the demonstration steps from the previous lecture; and three separate exercises which will allow the student to apply histogramming steps in order to reveal differing data distribution shapes. Answer key is provided.

In which I will explain what to do if a data set does not have a "normal" frequency distribution.

Topics include: the three options when a data set turns out to be non-normal; criteria for determining which path to take; a discussion of skew and what it looks like in a histogram; what multimodal and gapped distributions are, what they mean, and how to adjust for them.

In which I will explain what an outlier is, and why it is critical to identify and investigate them.

Topics include: how to most easily to identify outliers; alternatives to dealing with them; discussion of their value in shaping business improvement strategies; and specific psychedelic and smoothie emporium examples.

In which we will visit our Salon project business, and practice the techniques learned in Section 2.

Demonstration includes: data set selection and how to organize the information in Excel; keeping raw data separate from the worksheet area; using Excel to calculate the mean, median, max, min, range, standard deviation, and skew of a data set; visualizing the data using a run chart; histogramming the data set, and interpreting the salon daily client traffic data set characterization.

In which we will visit our Web based project business, and practice the techniques learned in Section 2.

Demonstration includes: data set selection and how to organize the information in Excel; keeping raw data separate from the worksheet area; using Excel to calculate the mean, median, max, min, range, standard deviation, and skew of a data set; visualizing the data using a run chart; histogramming the data set, and interpreting the web instructor's weekly course sales data set characterization.

In which the student may practice and master skills learned with the Salon data set, or use the workbook as a template.

Exercises include: an invitation for the student to repeat the demonstration steps from the previous lecture; and a "try it yourself" tab to allow the student's own business data to be copied in and the information characterized using the same steps.

In which the student may practice and master skills learned with the Web Instructor data set, or use the workbook as a template.

Exercises include: an invitation for the student to repeat the demonstration steps from the previous lecture; and a "try it yourself" tab to allow the student's own business data to be copied in and the information characterized using the same steps.

In which I will briefly explain the purpose, goals, and skills which will be covered in Section 3.

Topics include: the importance of establishing a baseline for business performance against which future results can be compared; the process of planning a change as an organized experiment; focusing and scoping the change properly; how long you will need to collect data (based on the baseline variability) in order to be able to clearly identify success vs. failure; and a specific psychedelic example of the lingo you will have mastered at the end of the section.

In which I will explain the immediate and recurring value of a baseline data set, which characterizes your current or starting state.

Topics include: the definition of baseline; its relevance and importance in driving improvements; how it can be used to gauge true change from false trends; and several examples of sensible baseline data sets.

In which I will encourage students to think more broadly about business improvement strategies, and introduce a fishbone diagram to assist in this type of thinking.

Topics include: the advantage of looking before you leap; methods to expand your thinking when it comes to change strategies; definition of a fishbone analysis; thoughts on root cause investigation - how to turn a hindsight tool into a foresight assist; 18 separate viewpoints to help broaden thinking; and a specific example - considering a web instructor's plan to increase sales from all 18 separate viewpoints.

In which I will emphasize a methodical approach, provide a planning checklist, and walk through an example. I will also touch on proactively addressing risk factors during planning.

Topics include: researching change strategies; budgetary considerations; establishing a timeframe; identifying recognizing constraints, limitations, and risks; setting clear, measurable goals; the value in investing in changes focused on reduced variability; and a specific example of this planning process applied to our salon project business.

In which we will talk about the importance of limiting external influences as much as possible in order to maintain focus on the change implemented - and also what to do when despite best efforts, things go awry.

Topics include: the value of method and patience; protecting the experiment by staying focused on a single end result (resisting the urge to add variables); balancing short term goals against long term goals which may seem opposed at times; and the occasional return to square one.

In which I explain a rule-of-thumb method to "guess" how much data will need to be collected (minimum duration of the experiment) in order to determine if a statistically significant change has been achieved.

Topics include: the trade-off of minimizing the time dedicated to an experiment, and maximizing confidence in statistically significant results; a calculation (which is based on your data set's baseline variability) provided to give a ballpark estimate on a sensible experimental duration; a fun bowling example; and some ground rules when deviating from the suggested experimental timeframe.

In which I will bore the non-mathematicians with the details behind the rule of thumb previously presented.

Explanation includes: a bunch of math, with some charts and examples. The general idea is that the estimated minimum number samples needed to recognize statistically significant results depends on the baseline data set's variability, and also increases as expected confidence level increases.

In which we will revisit our Salon project business, and practice the techniques learned in Section 3.

Demonstration includes: establishing a goal % change for salon client traffic; allowing the Excel workbook auto calculate the number of data points needed to recognize statistically significant change; and a discussion of how expected results "confidence level" affects the amount of data you will need to collect.

In which we will revisit our Web based project business, and practice the techniques learned in Section 3.

Demonstration includes: establishing a goal % change for weekly course sales; allowing the Excel workbook auto calculate the number of data points needed to recognize statistically significant change; and a discussion of how expected results "confidence level" affects the amount of data you will need to collec

In which the student may practice determining the appropriate duration of experiments based on confidence levels.

Exercises include: an invitation for the student to repeat the demonstration steps for determining appropriate experiment duration for both the salon and web instructor projects; and also, using the "try it yourself" tab, on the student's own business data and/or on a fun paper airplane project.

In which I will briefly explain the purpose, goals, and skills which will be covered in Section 4.

Topics include: executing as closely as possible to an established plan; the value of identifying and tracking adjacent data sets; situational awareness and protecting the integrity of the experiment; the correlation vs. causation trap; and a specific psychedelic example of the lingo you will have mastered at the end of the section.

In which I will talk about the stages of implementation, and practical considerations for executing to plan.

Topics include: setting the plan in motion; best ways to protect the integrity of the data; what may go awry; appropriately gauging your reaction to inevitable complications; and an exercise playing out "what if..." scenarios.

In which I will share ways that data collection can go wrong, and share ways to avoid traps and pitfalls. I will also nag you to remember to create data backups.

Topics include: the importance of consistency, integrity, and independent data verifiability; planning a process-oriented collection approach that minimizes subjectivity; the importance of stakeholder involvement; and some fun science fair methodology examples.

In which I will share ways to raise your awareness - keep your wits about you during the measurement period, in order to protect the integrity of the data set.

Topics include: the importance of vigilance during any measurement period; keeping contextual notes along with your data; the value of anomalies; looking past surface explanations when anomalies are identified; and the "5-why's" analysis applied to an illuminating bakery waste example.

In which I will inundate you with examples of the correlation causation fallacy, and explain why we are pre-wired to make judgment errors where patterns are concerned.

Topics include: why humans are predisposed to jump to conclusions; how correlation and causation are different (with many examples); and why this confusion is potentially the biggest threat to accurate critical analysis of data.

In which I will briefly explain the purpose, goals, and skills which will be covered in Section 5.

Topics include: a brief historical background of our most advanced analysis technique - the t-Test; discussion of what the t-Test does (and does not do); the value of variance comparisons; further elaboration on confidence levels; multiple step-by-step t-Test analysis demonstrations in Excel, with instruction (and a handy shortcut to) correctly interpreting the results; and a specific psychedelic example of the lingo you will have mastered at the end of the section.

In which I introduce our headliner act: the t-Test. I will explain what it is, how it works, introduce its inventor, reveal its historical association with beer, and explain its power and relevance (even for non-beer-related analyses).

Topics include: how serious the beer industry was, even in the early 1900's; William Sealy Gosset, possibly our favorite statistician ever; t-Test benefits and limitations; how data set variability factors in; the incredible value of the t-Test in business analytics - how it can save you time and money, and allow you to see simple hidden truths to assist in data-driven decision making.

In which we will bring back the concept of variability, and learn how to measure it across a data set. I will also discuss why smaller variability is better, and make a case for change experiments specifically aimed at variance reduction.

Topics include: the definition of variance; what it looks like in a run chart and a histogram; how to calculate it (the hard way, and the easy way); why we like smaller variance; what we can learn from looking at the comparative variances of data sets; and how variance relates to the t-Test.

In which I will explain (with confidence) what the term "confidence level" means, why it's important, and how our expectations for acceptable level of confidence depend on the context of an experiment.

Topics include: inherent inaccuracy in variable data sets; the two things that affect confidence level; how much confidence is reasonable to expect; necessary trade-offs; false positives vs. false negatives and which is worse; typical confidence levels used in statistics; and the definition of "alpha" (which indicates the amount of doubt we're comfortable with.)

In which we will take a test drive of the t-test. I will demonstrate step-by-step in Excel how to compare two data sets to see if a statistically significant difference can be supported. And I will explain how to easily interpret the results that Excel provides.

Topics include: hands-on step-by-step t-Test examples using Excel; reprise of our bowling experiment - comparing performance before and after lessons to determine if there's a statistically significant change; how to use the provided Excel workbook to give a handy color-coded "Success" or "Fail" t-Test result, and also how to interpret the results that Excel provides; and several iterations of the t-Test example, varying the settings.

In which the student may practice and master the t-Test techniques previously demonstrated.

Exercises include: an invitation for the student to repeat the demonstration steps from the previous lecture; and a separate exercise covering item sales variability comparison, multiple t-Test execution, and results interpretations. Answer key is provided.

In which we will revisit our Salon project business, and practice the t-Test analysis.

Demonstration includes: copying raw client traffic data into the workbook; deciding how many post-change data points to collect and analyze based on expected confidence level; comparison of mean, variance, and skew; a quick peek using a run chart and trend line; a t-Test analysis of before and after data sets; P-Value interpretations; and a caveat on relying on smaller data sets than recommended.

In which we will revisit our Web based project business, and practice the t-Test analysis.

Demonstration includes: copying raw sales data into the workbook; deciding how many post-change data points to collect and analyze based on expected confidence level; comparison of mean, variance, and skew; a quick peek using a run chart and trend line; a t-Test analysis of before and after data sets; P-Value interpretations; and an illustration of the perils of peeking too early.

In which the student may practice the t-Test analysis.

Exercises include: an invitation for the student to repeat the t-Test steps from the previous lectures on both the salon and web instructor projects; and also, using the "try it yourself" tab, on the student's own business data and/or on a fun paper airplane project.

In which I will briefly explain the purpose, goals, and skills which will be covered in Section 6.

Topics include: a case for incorporation constant collection into your business process; when to re-baseline; practical considerations in business experiments including the point diminishing returns and break-even point; an introduction and demonstration of the Pareto analysis technique; a look-forward to multiple factor experiments; and a specific psychedelic example of the lingo you will have mastered at the end of the section.

In which I will talk about expanding your data collection approach, and establishing a growth strategy that fits your business needs. I'll also explain the need to re-baseline at logical points, as a way to support staged improvements (or respond to business climate changes).

Topics include: establishing a long-term strategy; revisiting the value of reducing variability; examples of concurrent data sets for the two project businesses; leading vs. lagging indicators; and recognizing when a baseline has outlived its usefulness.

In which I will recognize that as data takes time and effort to collect and analyze, practical considerations should dictate where that effort is best applied.

Topics include: practicality as a key consideration in any change strategy; cost/benefit trade-offs - culling your data sets; recognizing the value of failures; finding the indicators right for your business; the importance of return on investment; a practical example on calculating a break-even point for an initial change investment; how variability affects both break-even and ROI; and some practical examples.

In which I will introduce the Pareto analysis tool, which will help you prioritize problems, risks, and issues.

Topics include: definition of the Pareto analysis technique; an example of Pareto in action: merchandise returns; what a Pareto chart looks like; how it differs from a bar chart or histogram; and its evolving nature which keeps your biggest problems at front and center, saving you from focusing on minutiae

In which I will provide a step-by-step demonstration of the simple, elegant Pareto tool.

Demonstration includes: dog park shenanigans; a step by step illustration of creating Pareto charts in Excel

In which the student may practice and master the Pareto analysis technique.

Exercises include: an invitation for the student to repeat the demonstration steps from the previous lecture; and a separate exercise exploring psychedelic product returns by product category and season. Answer key is provided.

In which I introduce ANOVA and Factorial analysis techniques which, while out of scope of this course, provide a logical next step for students wishing to build on skills learned in this course.

Topics include: an exploration of t-Test limitations; Analysis of Variance (ANOVA) benefits, usefulness, and applications; the Factorial experiment benefits, usefulness, and applications; and examples of where these techniques can take you.

In which I will review what we have covered in this course, thank the student for taking this journey with me, and itemize the supplemental material provided in Section 7.

In which I will share a bit of my background and experience, as relates to this course offering.

In which I will list the equipment, software, templates, and other resources used to develop this course, including a home-grown progress dashboard that let me know, hour by hour, how very slowly I was progressing.

In which I will share additional links and suggestions that may assist the inspired or curious student on their journey to business analytics mastery.