Marketing Analysisfor Everyone.
No installs. No coding. No cost. Yes to insights.
90 tools, AI-powered study buddy, interactive reading modules, and 60+ badges โ all browser-based, all free.
See what MktPraxis can do for you
Choose the experience built for your role.
You have the data. Now which test?
Turn the moment of "I have no idea where to start" into a defensible first step. Answer two plain questions about your data, then see the method, the tool, and why the tempting alternative would mislead you.
Find your next stepAn AI that can actually see your spreadsheet
Praxis Pal meets you inside the analysis, with the file and output already in view. Ask why a coefficient is negative or what a result means, and get help connected to the variables you are actually working with.
See contextual helpLeave with something you can actually show
A transcript records a grade. Your work should show more: analyses you can export, badges earned through practice, and a Praxis Pro certificate with a link others can verify.
See your evidence of workRuns on the laptop you already have
Your course should begin with the question, not a software scavenger hunt. Open a browser, drop in a spreadsheet, and start working on the device already in front of you.
Start with your dataBetter than watching the recording at 1.5x
Make a decision, see its consequences, and argue about what to do next. Live rounds, market simulations, and interactive chapters turn marketing analytics into work you have to think through.
Step into a live sessionRead it. Try it. Use it.
The explanation, interactive decision, and check-in belong on the same page. Learn the judgment behind modern marketing work while the example is still alive, not two editions behind.
Open the living textbookStart from a real course, not a blank page
Begin with a teachable sequence: modules, readings, tools, quizzes, and practice already connected. Put your judgment into the examples, emphasis, and pace instead of rebuilding the plumbing.
See the course foundationReach students before they drift away
In a large class, silence can look like coping until it is too late to help. Activity, module progress, and quiz history give you an earlier signal and a reason to reach out while a conversation can still matter.
See the earlier signalA teaching assistant for every student
You set the standard; Professor Mode gives students the practice path to reach it. It guides each move in the tool, while Praxis Pal helps them interpret their own work when your office hours are closed.
See guided practice scaleNobody has to approve this
No lab image, seat license, purchase order, or student software bill. The platform lives in the browser, and the accessibility and student-data details are ready when your department asks.
Get the practical detailsMake the room worth being in
Give students a reason to make a choice in the room. A projected code becomes a shared live decision, or a market simulation becomes the running argument that connects one class meeting to the next.
See the room come aliveSee what's possible
Getting Started
Data Import Wizard
Upload a messy CSV or Excel file and get guided, one-click fixes for common formatting problems โ then export clean data ready for any tool.
Univariate Analyzer
Explore a single variable's distribution, outliers, and summary stats. Always start here.
Tool Finder
Not sure which tool to use? Browse all 90 tools, use our guided wizard, or find answers to common questions.
Pearson Correlation
Check if two continuous metrics move together (e.g., ad spend vs. revenue).
Data Visualization Studio
Explore bivariate relationships through 8 chart types with rich narratives that teach what each chart reveals and hides.
Categorization Studio
Transform continuous data into meaningful categories. See how the same data tells different stories.
Missing Data Lab
Inject MCAR, MAR, and MNAR missingness, apply four imputation treatments, and watch each one distort your regression slopes.
Hypothesis Testing
Welch's t-test
Compare averages of two independent groups (e.g., Control vs. Treatment).
Paired t-test
Compare before/after scores for the same people (e.g., Pre vs. Post).
One-Way ANOVA
Compare averages across 3+ groups (e.g., 4 different landing pages).
Rank-Based Tests
Compare groups when averages mislead: skewed spend, a few huge orders, or 1โ5 ratings. Mann-Whitney, Wilcoxon and Kruskal-Wallis in one place.
A/B Proportion Test
Compare conversion rates between two variants.
Bandit vs. A/B Race
Race an A/B test against adaptive algorithms on the same visitors. See what each one earns, and what each one lets you learn.
Chi-Square Test
Check if two categorical variables are related (e.g., Device vs. Signup).
McNemar Test
Analyze "switchers" in paired yes/no data (e.g., before/after purchase).
One-Sample Hypothesis Test
Test whether a mean or conversion rate meets a business SLA or industry benchmark, with Big Data effect size exploration.
Chi-Square Goodness-of-Fit
Compare categorical frequencies against equal or custom benchmark distributions to identify standout consumer drivers.
Predictive Modeling
Bivariate Regression (Simple)
Predict an outcome using one variable (e.g., Spend โ Sales).
Multiple Regression
Predict a numeric outcome using multiple drivers.
Transformation Lab
Find out whether re-scaling a variable is actually warranted, then get the formula for Excel, Sheets, Python, R, SPSS or Stata.
MLR with Interactions
Model non-linear effects and interaction terms.
Logistic Regression
Predict binary outcomes like Churn vs. Retain.
Multinomial Logit
Predict choice among 3+ options (e.g., Brand A vs B vs C).
Mediation Visualization
Visualize direct, indirect, and total effects in mediation and moderated mediation models.
Residual Diagnostics Studio
Audit the 4 classic LINE regression assumptions (Linearity, Independence, Normality, Equal Variance) and Cook's distance outliers.
Text Analysis
Qualitative Analyzer
Manually code and analyze interview transcripts or focus groups.
Theme Extractor
Automatically discover hidden topics in reviews using AI.
Sentiment Lab
Score text as positive, negative, or neutral.
Marketing Attribution
UTM Campaign Tracking Lab
Tag a campaign, run it for 30 days, then find out how much of your dashboard was telling the truth.
Web Analytics Sandbox
A safe GA4-style account with two years of real traffic. Segment it, diagnose the funnel, and find the problems hiding in the data.
Paid Search Sandbox
A Google Ads account nobody has audited. Move a live auction, find the wasted spend, and defend next month's budget.
Shapley Value Attribution
Cooperative Game Theory. Assign fair credit to channels based on their marginal contribution to the team.
Markov Chain Attribution
Sequence-based attribution. Analyze how the order of touches acts as a bridge to conversion.
Marketing Mix Modeling
Estimate what each channel really adds to sales, including carryover and diminishing returns, then find a better split of the same budget.
Choice, Preference & Pricing
Conjoint Study Creator
Design conjoint analysis experiments to measure customer preferences and willingness to pay.
Conjoint Analysis
Analyze conjoint study results to reveal customer preferences and part-worth utilities.
Basic Conjoint Analysis
Simplified conjoint analysis for quick preference insights.
Conjoint Experiment Simulator
Take a real choice study in 3D, then analyze the data it produces with your own answers inside it.
MaxDiff (Best-Worst Scaling)
Rank messages, features or benefits on a scale that sums to 100. Forces the trade-offs that rating scales let people avoid, where everything comes back "important".
TURF Analysis
Pick the 3 flavors, products or messages that reach the most customers. The best set is rarely the 3 most popular, and this shows you why.
Price Optimization Sandbox
Estimate demand curves from data, find revenue- and profit-maximizing prices, and compare pricing strategies side-by-side. Upload your own data or explore built-in scenarios.
Tiered Demand Curve Estimator
When you sell Basic / Standard / Premium to one market, estimate per-tier demand and the full own/cross-price elasticity matrix. Berry (1994) aggregate logit with optional 2SLS for endogeneity.
Customer & Lifecycle Analytics
Basic Customer Lifetime Value
A gentler way into CLV: two small models, one question at a time. Learn what a customer is worth over the months ahead, and why a quiet month is not always a lost customer.
Customer Lifetime Value (CLV)
Model customer retention and conditional spend to forecast CLV by segment. Identify at-risk customers, evaluate marketing levers, and benchmark acquisition spend with CLV/CAC analysis.
Retention & Survival Curves
See how long customers stay, which sign-up cohorts hold on best, and why a simple churn rate often gets it wrong.
Panel Dynamic Logistic Regression
Your repeat buyers keep buying. Find out whether that is habit or loyalty, and what a free trial is really worth.
Market Basket Analysis
Find which products are bought together, and tell a real pattern from a coincidence, before you plan a bundle or cross-sell.
CRM Sandbox
A working B2B sales CRM for four reps. Diagnose the pipeline, tell a real gap from luck, and build a coaching plan you can defend.
Huff Gravity Model Simulator
Simulate retail trade areas on an interactive map. Adjust store attractiveness and distance-decay parameters, run what-if competitor scenarios, and watch market-share boundaries shift in real time.
Causal Inference & Policy Evaluation
Difference-in-Differences
Did the campaign really work? Compare the markets that got it with ones that didn't, and check the assumption the answer depends on.
Propensity Score Matching
Match treatment and control customers on their observed characteristics to estimate a causal effect when you could not randomize.
Segmentation & Measurement
K-Means Clustering
Segment customers into distinct groups based on behavior.
k-Prototypes Clustering
Cluster mixed attributes (Numbers + Categories) like Spend + Region.
Latent Class Segmentation
Probabilistic customer segmentation using LCA/LPA. Discover hidden segments with soft class membership.
Perceptual Mapping (MDS)
Visualize how brands are perceived using multidimensional scaling โ find competitive battlegrounds and white space.
Scale Reliability & EFA
Check whether your survey items hang together, find the dimensions underneath them, and build scale scores you can defend in a report.
Advanced Machine Learning & Forecasting
Decision Tree Classifier
Build and interpret classification trees to segment customers and predict outcomes.
Neural Network Playground
Build and visualize neural networks to solve complex marketing classification problems.
Neural Network Lab
Train a neural network on your customer file, watch it learn, and find out on customers it never saw whether it beats a regression you can explain.
Forecast Starter
Your first time series tool. See what a series is made of, extend it with a rule you can say out loud, then find out whether that rule beats doing nothing at all.
ARX Forecasting (Intro)
Learn time series basics with ARX modelsโeasier than ARIMAX, no moving average terms.
Dynamic Logistic Regression
Forecast yes/no events like stockouts or sell-outs from yesterday's result and your plans, and test it the honest way: on days it never saw.
ARIMAX Forecasting
Forecast sales over time with external predictors like ad spend.
Marketing Resource Optimizer
Allocate limited budgets or resources across channels to maximize ROI.
CRM Sandbox
A working B2B sales CRM for four reps. Diagnose the pipeline, tell a real gap from luck, and build a coaching plan you can defend.
Territory Optimizer
Draw sales territories on a map, balance workloads, and optimize rep assignments.
Model Drift Lab
Watch a prediction model degrade as customer data shifts. Diagnose feature drift with PSI & KS statistics, retrain to recover, and learn when to set monitoring alerts.
Gen AI Cost Estimator
Estimate what it actually costs to deploy an LLM in marketing โ token pricing, conversation accumulation, model tier trade-offs, and ROI against human agents.
Study Design & Probability
Sample Size (Single)
Plan n for one mean or proportion.
Sample Size (A/B)
Plan n for comparing two groups.
Sample Size (Reg)
Plan n for correlation/regression.
Multi-Arm A/B
Plan n for A/B/C/D tests.
Sampling Visualizer
See how different sampling methods work.
Confidence Interval Simulator
Draw repeated samples from skewed revenue distributions to understand frequentist coverage and parameter capture.
Selection Lab
Explore selection probabilities.
Probability Calc
Calculate likelihood of repeated events (Binomial/Poisson).
Understanding Distributions
Eighteen short demos covering the distributions analysts actually use, each one attached to a real decision: staffing a chat queue, setting a free-shipping threshold, judging whether an A/B result is worth shipping.
Distribution Calculator
Calculate shaded areas, tail probabilities, and critical thresholds for Normal (Z), Student's t, Chi-Square, and F curves.
Bayes' Theorem & Natural Frequency
Visualize Bayesian updating, 1,000-dot natural frequency trees, and the base-rate fallacy in customer retention and fraud alerts.
Odds-O-Matic
Probability, odds, odds ratios and relative risk for two customer segments, drawn as meeple pictographs so the difference is visible before it is calculated.
Model Fitting Foundations
Understanding Regression
Ten short demos: where Rยฒ comes from, correlation as a slope, reference categories, interactions, ANOVA, and omitted variables.
Response Model Lab
Name the response shape, pick a form, then calibrate it by hand against MAE or MSE.
MAE Calibration Lab
Manually fit linear & quadratic models by minimizing error. Discover what "model fitting" really means.
Log-Loss Classification Lab
Manually calibrate logistic regression by minimizing log-loss. Learn why probability calibration matters for marketing.
Overfitting Explorer
Explore bias-variance tradeoff through polynomial regression. Compare training vs. holdout performance across marketing scenarios.
Rare Event Calibration Lab
See why models trained on rare-event data (churn, fraud) produce unusable scores โ and apply the King-Zeng correction to fix them.
Classifier Comparison Lab
Compare five binary classifiers side by side โ Logistic Regression, Decision Tree, k-NN, Naive Bayes, and SVM โ on the same data.
The Hundred-Heads Test
A stranger calls a hundred coin flips correctly. Work out why the airtight arithmetic proves nothing โ then find the same three failures in your own A/B tests.
Classification Threshold Lab
A classifier gives you a probability, not a decision. Move the cut-off and watch profit trade off against cost until you find the threshold that actually pays.