Visitors: Properly calculate totals over time

A common mistake that many users make when first using content analytics is improperly attempting to add visitor totals over a period of time.

Web analysts call this the “hotel problem,” and it refers to the fact that you cannot add visitor totals for a range of dates in a period and have that total equal the actual number of visitors in the period. So you can’t, for example, add visitor totals for Monday, Tuesday and Wednesday and have that equal the correct total for the three days as a whole. You can’t add visitors for every day in a month and match the correct total visitors for the month. And so on.

A chart often best illustrates the problem. Imagine these guests are staying at the Parse.ly Inn:

Guest	Monday	Tuesday	Wednesday
John	x	x
Dawn		x
Fred	x	x	x
Daily Visitors	2	3	1

As you can see, if you add the number of visitor totals for each day, you get six (2+3+1).

But there are only three visitors who stayed at the Inn, not six. You cannot add the daily totals and get the correct total for the full date range.

Let’s look at another example. Here are actual daily visitor totals for a small site tracked by Parse.ly:

Metric	Mon.	Tue.	Wed.	Thu.	Fri.	Sat.	Sun.
Visitors	423	257	181	228	168	166	148

Adding the numbers from each day provides a total of 1,571 visitors. But the actual total, when looking at a report for the entire week, is 1,231. That means some visitors came to the site more than once during the week, and adding the totals for each day means you are counting those visitors multiple times during the week instead of just once.

The key point to remember is that you cannot add numbers for all dates in a period and have that be the correct total. You must run a report for the entire period to get the correct total, be it a week, month, quarter or year.