Results in CSV report compared to HTML report

Avatar
  • updated
  • Answered
Hello,

I generated an HTML formatted report for all of 2014. I then generated just a December report as a CSV file sent to my email. The HTML report included the dropdown so I could choose December 2014 if I wanted to, along with November and October.

I compared just the summary numbers from the CSV file for December only to the dropdown December in the HTML report. Only the timeframe was changed, no other modifications in OPTIONS or filters. I have two different sets of numbers.

Can you tell me why they would be different and which I should chose as the accurate set?
I have version 8.6 Pro.

CSV:

Summary
Hits
Total Hits 390010
Visitor Hits 390010
Spider Hits 0
Average Hits per Day 12580
Average Hits per Visitor 26.1
Cached Requests 15466
Failed Requests 14775
Page Views
Total Page Views 42190
Average Page Views per Day 1360
Average Page Views per Visitor 2.82
Visitors
Total Visitors 14941
Average Visitors per Day 481
Total Unique IPs 9944
Bandwidth
Total Bandwidth 8.33 GB
Visitor Bandwidth 8.33 GB
Spider Bandwidth 0 B
Average Bandwidth per Day 275.26 MB
Average Bandwidth per Hit 22.40 KB
Average Bandwidth per Visitor 584.82 KB

HTML:
Summary
Hits
Total Hits 390,267
Visitor Hits 390,267
Spider Hits 0
Average Hits per Day 12,589
Average Hits per Visitor 25.95
Cached Requests 15,466
Failed Requests 14,895
Page Views
Total Page Views 42,323
Average Page Views per Day 1,365
Average Page Views per Visitor 2.81
Visitors
Total Visitors 15,040
Average Visitors per Day 485
Total Unique IPs 9,960
Bandwidth
Total Bandwidth 8.33 GB
Visitor Bandwidth 8.33 GB
Spider Bandwidth 0 B
Average Bandwidth per Day 275.31 MB
Average Bandwidth per Hit 22.39 KB
Average Bandwidth per Visitor 581.08 KB

Thank you for your help.

Gina
Avatar
Gina Price
Michael,

Thank you VERY much! This explanation helped me out a great deal. I wasn't quite sure what you had meant and I was going to do some testing.

I sincerely appreciate your help.

Gina
Avatar
Michael
I'd like to provide a bit more information on what could cause the difference if it indeed appears because of visitor filters.

Example: a visitor made the following requests:
November 30 - page1.html
November 30 - page2.html
December 1 - page3.html
December 1 - page4.html

Let's assume that you have created a filter to include visitors that accessed file page1.html.

If you analyze the whole year, this visitor will be included as it accessed page1.html, and the program will analyze all his hits. The program will then count 2 hits for November and 2 hits for December, and will show them such way if you view a subreport for December.

However, if you analyze December statistics only, the program will analyze only requests for page3.html and page4.html. As it won't know that the visitor requested file page1.html, it won't include any requests of this visitor, and count no hits for December.

In most cases difference caused by such situation should be very minor like in your case, so you can use statistics from any report.
Avatar
Gina Price
Michael,

Thank you. I will check the visitor filters and make sure nothing in the background is slightly different. I appreciate the guidance and will keep your advice in mind if I see this again.

Sincerely,
Gina
Avatar
Michael
Do you use any visitor-based filters? Using such filters may have such side effect because of visitors that had hits both in November and December.
Avatar
Gina Price
Hello,

I actually ran the files one right after the other. I used "previous month" rather than choosing the settings manually. I used "last year" for the first report and "last month" for the other report. I didn't change anything else. As soon as the year report was done, I ran the individual month report.

Any other ideas I should try or test? I left the time defaults alone.

Thank you again. I appreciate this community for guidance.

Gina
Avatar
Michael
It's strange, it seems that there is a minor difference in program settings, log data or time range that causes this difference in reports:

1. Maybe you have updated the country/city/state database between creating these reports, it may affect reports if you use filters based on this data. Maybe there was a also a change in log files for this period.

Such changes could have occurred if you haven't created the reports immediately one after another. In such case you should reanalyze both profiles and check if it solves the issue.

2. When you have chosen the December period, have you set the time range to "Last month" or set it manually, or have you just analyzed logs for this month? If you have analyzed the December logs without setting the time range, you should try to analyze all logs but set the time range in the profile properties.