When? For how long? More or fewer than before?
Time is relevant to almost every analysis. The ability to look forward and backward and drill down to weeks and days from months and years is fundamental.
Any business intelligence solution should let you move through time like a DeLorean. Working with time should be easy and intuitive. Tableau’s built-in date and time functions let you:
- Drag and drop to analyze time trends
- Drill into days, hours and seconds with a click
- Analyze time by day of week
- Perform time comparisons like yoy growth, moving averages and trailing-twelve-months (TTM)
- Find trends and do forecasting
Dates have a natural hierarchy: years, quarters, months, days, hours, and so on. Your software should understand this and make time series analysis easy. Filter to the most interesting dates. Drill down with a click. View the running total with a table calculation. Need IT to get involved? Then your software isn’t doing its job.
In this video we see how easily Tableau deals with dates. Do simple time series analysis by dragging and dropping. Add running totals and moving averages with a few clicks. Drill into dates and do seasonal analysis in seconds.
Trends throughout the workweek can give you that extra bit of intelligence about your business. Are customers more active on Mondays than Fridays? Do your emails get opened more when you send them on a Wednesday? Is that blip in your production data simply a result of five weekends in a month instead of four?
In this visualization, we look at website traffic over the course of several months. Looking at the complete time trend tells us there’s a slight increase in traffic and a general downward trend in conversion. But what action can we take? The day-of-week analysis suggests that weekend visitors, though fewer, convert at a higher rate. Perhaps offer a promotion to those visitors.
Applying basic calculations to time series data can yield not-so-basic insights. Running totals, difference from last or first value, and the ability to set a common baseline are examples of simple transformations that can be very useful with time data. They allow you to compare one period to another and do quite sophisticated analysis with little effort.
This analysis takes the three Toy Story movies, which came out over a period of years, and sets them to a common baseline: the total gross by week since opening. This lets you see the movies from a common starting point rather than over an absolute timeline, which is much more useful for comparison.
Simply by understanding patterns throughout time you can make more informed decisions about the future. Tableau’s trend line model is easy to apply and can be configured to fit linear, logarithmic and polynomial models.
Here we see GE’s stock price over a period of years. Looking only at the last year suggests the stock may be on a downward trend; extending the date axis to the future suggests that the stock may hit $15 in the near future. Change the date range of the analysis to see different trend lines.