Start a New Analysis
Do you have a problem in mind?
What's driving my sales effectiveness? Which kind of leads close and which dont? Can I prioritize my leads for better closure?
Driver Analysis (Binary Outcome)
Define a binary outcome (e.g., churn, failures). And discover why/when that outcome happens. Also clusters where the probabilities of the outcome happening are very different.
Driver Analysis (Continous Variable)
What are the chief drivers for your outcome? And how do they impact the outcome? Are there pockets of data (clusters) where the outcome is very different from the rest?
Time Series Analysis
Everything to do with time-series data. What are the seasonalities? Where are the anomalies? How are time-series related to each other? Are there clusters of time-series?
Discover categories in your data, that you can use to categorize new data as it comes in. Or flag anomalous words as they come, or do a bunch of things with text.
Impact Analysis (Continous Outcome)
What is the impact of a certain action on your (continous) outcome? Are there pockets where impact is higher/lesser?
Impact Analysis (Binary Outcome)
What is the impact of a certain action on your (binary) outcome? Are there pockets where impact is higher/lesser?
Forecasting / Time-Based Driver Analysis
See drivers for target after first adjusting for time. Auto Periodicity Detection, Correlations, etc.. And what actions can impact the target?
Discover profiles at a higher level in any hierarchical data, e.g., profile shopping baskets based on items, profile doctors based on the kind of procedures they do, etc. You can also rollup the data at a higher level for use elsewhere.
Marketing Spend Mix/Attribution
What's the impact of marketing spend mix? How quickly do marketing spend effects decay? And what will be incremental sales due to fresh marketing spend?
Discover clusters in your data, where the each column in the data is an attribute, e.g, cluster customers from profile data (gender, age, income, etc.)
Select an ID Column (e.g., transaction ID, machine ID, etc.), and generate patterns where events occur together, e.g., 'what items get bought together?', 'event A happened - how likely is it that event B will happen?', etc.