Weeding out irrelevant analytical data to see truer conversion rates

[dickslee23]

Here is the scenario. We have many brick and mortar store locations as well as an ecommerce website. It's hard to get exact, but my estimates seem to be that approximately 1/3 of the visitors to our website are interested only in obtaining information about the brick and mortar store locations and not interested in ecommerce transactions. Of course this kills the conversion rate. We use google analytics and I'd like to somehow be able to quantify with more accuracy what the "real" conversion rate might be. Is there some method to weed out specific pages/traffic (like brick and mortar landing pages) from being taken into account when conv. rate is calculated? The number that matters for conv. rate of course is "visits" and not unique pageviews, so I'm not sure that really would do anything helpful.

Any tips?

[Bevelwise]

Use advanced segmentation and setup a custom segment.  Set that segment up so that it includes data from all sources, but excludes users who trigger a page view on whatever that specific brick and mortar page is.

I say this as a solution to your direct question, but you also assume that users who visit these specific pages are not going to trigger an e-commerce sale, which they could.  I totally see where you are coming from here because in most cases you may be correct.  You know your audience better than me.

[dickslee23]

Thank you, that helps point me in the right direction. You are right that it's still hard to say for sure what the visitors are doing and if they will trigger a sale or not, but at least for us it will be a closer estimate than NOT filtering the data in some way.

[KeriMorgret]

This YouMoz post about setting up complex advanced segments for brand names might give you some ideas. http://www.seomoz.org/blog/guide-to-setting-up-advanced-segments-in-google-analytics-for-complex-brand-names

That post focuses on how to look at just branded or just non-branded queries. You might find it useful to take a similar approach but instead of branded filter for keywords that include location, hours, etc.