Take the 2-minute tour ×
Webmasters Stack Exchange is a question and answer site for pro webmasters. It's 100% free, no registration required.

My website is largely targeted to US business users. Not surprisingly, the number of visits during the weekdays is much larger than on Saturdays and Sundays. In addition, on holidays like July 4 or Christmas, the number of visitors drop dramatically. I also have a feeling that internet usage decreases slightly in summer months while people are on vacation/outside. The result is that in Google Analytics (pic below), my daily visitors go up and down and appears relatively erratic. It's hard to determine if any increase or drop in visits is attributable to something on my site or is part of a trend across the US.

Is there a way to normalize the number of visitors to my site against internet usage in general? Or are there tools other than Google Analytics that do this?

If normalized, the chart below would look much flatter and if there was an increase, I would know it would be attributed to something on my site rather than to the internet in general.

Visits to my site: Notice the weekly up and down and the drop around July 4

share|improve this question
add comment

migrated from superuser.com Aug 15 '13 at 19:48

This question came from our site for computer enthusiasts and power users.

4 Answers

up vote 2 down vote accepted

Google Trends is a good tool to give you an indication of the market fluctuations in your industry. To normalise the data you are probably going to have to do this manually. You can specify the country in Google trends as well as the keywords that you want to evaluate the relative search volumes for. Once you have this you can compare the fluctuation with those seen on your site. To smooth out your weekly fluctuations you will likely need to use a 7-point moving average.

If you have enough data you could use a moving average that looks at an entire year (365 point moving average). This would remove any weekly fluctuations as well as any seasonality fluctuations to give you an idea of your websites normalised performance, as i say though, the downside is that you will need a moderate amount of historical data.

share|improve this answer
add comment

Great question and the answer is time series analysis. You've spotted one of the common problems in analyzing time series. Not sure what language you use, but R and python have good libraries for this.

Here's an intro: http://www.itl.nist.gov/div898/handbook/pmc/section4/pmc4.htm

And here is a paper about using the technology for forecasting. http://blogs.adobe.com/digitalmarketing/analytics/forecasting-adding-value-to-web-analytics-part-1/

It's work to set it up, but the information is actually useful. Without it, you're just eyeballing the graphs.

share|improve this answer
add comment

I don't know of any tools within or without Google Analytics to compare hits with the rest of the internet in general, much less to specific countries / demographics / etc. I generally use the Week view to iron out the weekend variations, and compare to Previous Year to make sure things are on the right track.

There will always be some variation, of course, especially the lower your traffic levels are, but if you have sharp increases/decreases outside of those trends, sometimes the only thing to do is to drill down to see if the change was persistent across different browsers, traffic sources, pages... or if you can isolate the change to certain factors.

share|improve this answer
add comment

Thats' quite normal trend, I would say. It does not seems to have any bad image on your site part since its a US culture where people are more concentrated on there personal life rather than spending time on internet & website. Even good corporate brand site have similar pattern.

I suggest to focus on more Weekdays' data.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.