How to Compare Your Data Across Time

Looking at what happened today can be useful. Comparing it with another period can tell you much more.

A comparison helps answer a different question: what actually changed?

You can compare one day with another, this week with last week, or a recent period with a longer history. The important part is choosing a comparison that makes sense for what you're trying to understand.

Choose the Right Time Period

Start by deciding what you want to compare.

If you're interested in a recent change, comparing today with yesterday may be useful. If you're trying to understand a broader shift, a weekly or monthly comparison may give you a clearer picture.

Short periods show more detail. Longer periods can make broader patterns easier to see.

Neither view is automatically better. They answer different questions.

Compare Similar Periods

Whenever possible, compare like with like.

A partial day may not be directly comparable with a complete day. A few hours of information may look very different from an entire week.

Before comparing two periods, check:

  • whether they cover similar amounts of time;

  • whether similar types of information were collected;

  • whether there are large gaps in either period; and

  • whether anything unusual affected one period but not the other.

A fair comparison makes the differences you find more meaningful.

Start With the Overall Difference

First, look at the big picture.

Did the information generally increase, decrease, or remain similar?

Then look more closely at where the difference came from.

A change in an overall result may reflect many small differences throughout the period, or it may come primarily from one unusual event.

Understanding which is happening can tell you more than the total difference alone.

Look at When the Difference Happened

Timing can reveal details that totals hide.

Two days could end with similar overall results even though the information was distributed very differently throughout each day.

Likewise, two periods with different totals may actually look very similar except for one particular time.

Look at when measurements or events occurred and ask:

  • Did the difference happen throughout the period?

  • Was it concentrated around a particular time?

  • Was there one event that accounted for much of the change?

  • Did the usual pattern happen earlier or later?

This helps you understand where a difference came from instead of simply noticing that one exists.

Check the Information Around It

Once you find where something changed, look at the surrounding information.

Your connected devices and sensors may have recorded other events or measurements around the same time. Information you entered may provide additional context.

For example, if one period looks different from another, check whether your recorded events, inputs, or other measurements were also different.

You may find that several pieces of information changed together.

Don't Let One Unusual Point Define the Comparison

A single unusual measurement can sometimes make two periods look more different than they really are.

If something stands out, look at the rest of the data before deciding what the comparison shows.

Ask what the periods look like with that measurement in context.

If most of the information is similar and only one point is substantially different, that tells a different story from a change that appears repeatedly throughout the period.

Compare Again Over Time

One comparison gives you two points of reference.

Repeating the comparison can tell you whether the difference continues.

If this week looks different from last week, see what happens the following week. If today's pattern changed from yesterday's, look at additional days.

Repeated comparisons can help distinguish a temporary difference from a change that continues over time.

Use Comparisons to Ask Better Questions

The purpose of comparing periods isn't simply to decide whether one was “better” than another.

A useful comparison helps you ask more specific questions:

  • What changed?

  • When did it change?

  • How large was the difference?

  • Was the change spread throughout the period or concentrated in one place?

  • What other information changed at the same time?

  • Have I seen the same difference before?

  • Does it continue when I look at another period?

Those questions can turn a simple before-and-after comparison into a much clearer understanding of your data.