Impression Variation

Impressions vary wildly every 4-5 days

[warrisr]

Hello,

I recently generated a 4 month graph of impressions from one of my AdWord campaigns and was surprised to discover a huge cycling of impression rates.

The graph consistently cycles up and down every 4-5 days from a low of 1,000 impressions a day to a high of 6,000 impressions a day.

I have been using AdWords for about a year now with good success, but have just started trying out some of the reporting tools that Google has available. This seems to be a very significant swing and probably partially explains a similar cyclical swing in orders that I have noticed over the past year, but chalked up to other factors.

Does anyone else see this kind of variation in impressions? Is there something I should be doing to get a more consistent day-to-day and week-to-week impression rate? I have gone through all of the documentation published by Google and don't see any mention of this, or strategies on how to deal with it.

Ron

[qfguy]

That does not seem unusual to me. I chalk it up to work week rythm. Why is it a 'problem'? The average is the average.

[Syzygy]

Weekends - starting from Fri lunchtime (GMT) - always see a huge drop in impressions for me, thus graphs also indicate a 4-5 day cycle. As we aim at professional/qualified individuals in certain sectors it is, of course, to be expected that our clicks come Mon-Fri, on a global 9-5.

Syzygy

[eWhisper]

All of the sites that I track have a very consistant weekly peak rate, both in time of day and in days of the week. Depending on the sites purpose, often you'll see a huge difference in hits between weekdays and weekends.

[warrisr]

Thank you all for the re-assurance. The cycling is indeed directly related to weekdays/weekends, however it just seemed that such a huge difference (4-6 times more impressions during the week then on weekends/holidays) was a bit extreme. But from what people are saying, I guess this is not all that unusual.

Thanks!

Ron

[qfguy]

wait for big news and then you'll see traffic tank. When SH was captured we had no hits for a week!

[dragonlady7]

Haven't you ever observed this with sites you yourself visit? I used to visit a particular entertainment site weekly in the early evening, because they had a notice that they were updated on that one weekday.
The site would be slow and the large files (Flash animations) would take forever to download. I simply chalked it up to their being large, and would make myself a sandwich or whatever.
One week i forgot to go on the weekday night, and instead went two days later.
The site absolutely flew.
Duhh... everyone else had noticed the date it was updated, and went there on that day.
I bet their logs were crazy one day a week.
I'm sure other traffic is similar. i'm not a creature of habit but my web browsing habits more set than anything else, because there are certain times when it's convenient and I'm thinking about browsing.
So I wouldn't be too surprised by these weekly fluctuations. I hadn't thought of them until the incident with the popular site, but it's been very interesting to me in running my own site now to see when people think to visit it. :)

[AdWordsAdvisor]

This is slightly off topic, but worth mentioning I think.

Traffic can also be tremendously effected by current events, and big spikes can happen when something you sell is featured on TV, on the radio, in periodicals, etc.

I think my favorite example revolves around a certain stainless steel bodied car, with gull-wing doors.

I used to speak frequently with an advertiser who sold parts for these DeWidgets. And every time the movie 'Back to the Widget' played on TV (in which a DeWidget plays a prominent role), he had a big spike in both impressions and clicks - without an increase in ROI.

Not too surprising, really.

And you should see what happens to impressions for some magazine related keywords, when Britney Widget makes a splashy appearance in that magazine.

And let's not even talk about things featured on the Oprah Widget, at least here in the US...

AWA