Bill,

I think that this is actually more of a traffic analysis and landing page optimization issue then a strict usability issue - although usability problems are one potential cause of the drop in conversions - and there is no simple answer.

Fundamentally, there is no reason why an increase in traffic should cause a drop in conversions.  What really matters is the source of your traffic and how they engage with your site when they arrive.  It sounds like you have increased traffic while also making some changes to the site, and teasing apart the effect of each factor is part of the challenge.

On the traffic side, I would look at the following:
 • Conversion rate by traffic source - the source(s) of your increase could have a lower conversion rate then the average mix before the increase
 • Conversion rate by keyword - if search or PPC traffic is driving your increase, you make be ranking well on keywords that aren't associated with purchase activity.  For example you might score well for something like "frequent flier programs" which is probably someone only doing research, whereas a keyword like "cheap airline ticket" is more strongly associated with an intent to purchase.
 • Time on site or number of visits until purchase - It may be that it takes multiple visits by a user until she converts and your recent increase in traffic portends a later increase in conversion.

On the site/design side, I would look at the following:
 • Bounce rate on initial landing page - especially as this compares to your historical rate. This is an immediate indicator of whether your landing page is compelling.  Breaking this down by traffic source is also useful.
 • Compare the design and messaging of your ads to the design and messaging of your landing pages - incongruity between ads and landing pages will cause users to quickly abandon your site
 • If you're not already, try a tool like ClickTale to watch users interacting with your site.  This can be extremely insightful, especially on complex purchase processes

Finally, I would aggressively split test both your landing pages and your ads.  Even if you identify a weak traffic source or a design problem, you can realize significant gains in conversion through aggressive testing.  Netflix has a great philosophy on this:
http://www.quora.com/What-types-of-things-does-Netflix-A-B-test-aside-from-member-sign-up

I realize that this isn't quite the answer that you were looking for, but hopefully it's helpful.

-Ivor

On Feb 16, 2011, at 11:07 PM, Phillips, William [HDQUV] wrote:

> Hi - Although I am not a usability expert I have worked on web sites for
> over 15 years and I have learned much from this group.  I have a
> question I was hoping the group might have some insight on or perhaps
> even point me to a white paper or literature on the topic.  
> 
> As our web site traffic has increased (mileageplus.com), we have seen
> the actual conversion decrease, although the overall YOY conversion is
> higher.  I was under the impression this is standard; since we're
> driving more traffic to the site and have optimized it with new
> features, more users are looking and browsing and ultimately purchasing.
> Since the numbers were so low in the beginning, only those users really
> interested in purchasing were using the site.  My management finds the
> drop in conversion worrisome.  Is there any particular school-of-thought
> on this?  
> 
> Many thanks,
> Bill
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~
> Bill Phillips
> Project Manager | Mileage Plus
> Email: [log in to unmask]
> Phone or Fax: 312.997.8166 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
>    --------------------------------------------------------------
>        Tip of the Day: Forward out-of-office replies to
>                    mailto:[log in to unmask]
>            CHI-WEB: www.sigchi.org/resources/web/faq.html
>              MODERATOR: mailto:[log in to unmask]
>    --------------------------------------------------------------

    --------------------------------------------------------------
           Tip of the Day: Postings must be in plain text
            CHI-WEB: www.sigchi.org/resources/web/faq.html
              MODERATOR: mailto:[log in to unmask]
    --------------------------------------------------------------