Brad Geddes / PPC Geek
Official AdWords Seminar Leader.
Author of Advanced Google AdWords.
Co-Founder, Adalysis.
(312) 884-9017
Brad Geddes's Theories on Marketing Overall Conversion Rates are Useless

Overall Conversion Rates are Useless

Overall conversion rates just let you know the percentage of people who completed an action. They do not contain any context about the actual visitors. It’s the context surrounding conversion rates which is the important aspect to pay attention to.

This is some of the analysis we did our our own site a few months ago before we conducted an entire redesign; it was  this type of thinking and data that lead to decisions that have had a very positive impact (more on that in another post).

Here’s a snapshot of overall conversion rates on one site:


0.54% doesn’t sound very good from a conversion rate standpoint.

This happens to be the conversion rates for this entire site for just the AdWords seminars goal. Since this site has a blog, information about our company, etc – and the seminars are location based – overall conversions doesn’t mean much as most of the readers do not live in or near a city where we speak.

Conversion Rates By Region

If we were to examine overall conversion rates for the seminars by region, we see a much different set of numbers:


In a case of regional items, overall doesn’t matter – examining conversions from those who are eligible to actually attend your event (or come to your store, etc) is much more important.

Conversion Rates by Source

If you don’t have a geographic focus; then examine conversion rates by source:


Or better yet, segment by both source and region (this could be keyword, campaign, etc):


If you have more than one conversion type, then don’t just segment by total conversion rates; again that’s useless. Conversion rate just shows you the percentage of people who converted; which might not have anything to do with revenue.

My favorite metric to work with in Analytics is Per Visit Goal Value

I know exactly what a seminar registration is worth. I know exactly what a feed subscription is worth (which is a lot less), a subscription, and I know what a contact is worth. In analytics, each of these items is set with its own value. Now instead of just looking at conversion rates – I can see what source leads to the highest total value for the site.


Again, this could be keywords, referring sites, campaigns you are running, etc.

Now we have something to work with – actual revenue and dollar amounts.

How do we get more traffic from the sits sending us visitors worth $67, and not waste our time focusing on sites that are sending us little traffic.

Monetizing Your Site

You don’t have to just focus on traffic coming to your site. Examine the traffic within your site.

The $ index will show you the value of that page view.  Examine your own  site for you pages with the most monetization value.


Now, compare them to the site average:


As you can see, our top monetization pages have many less page views than the site average. This means we need to think about how to drive more traffic into those sections.


Overall conversion rates are useless.

Focus on at least one segmentation of traffic. It could be region, traffic source, campaign, etc. But do not look at overall numbers because they don’t tell you anything actionable.

Segmenting by two items is even more useful if you still have enough data to make decisions, and you have the time to not just do the analysis – but also the work.

If you spend all your time figuring out what to do; and don’t do it – then maybe you should have done less analysis and more doing.

Always remember, traffic does not convert the same by source.

I went through this (and a lot more) analysis of our site, traffic sources, etc before we did a fairly major (and inexpensive) redesign a few months ago. You can read about the actual redesign process here.

Now that it has been a few months since we conducted the redesign, and have enough data about what the redesign has actually done for the site; I’ll be writing up those finding and publishing them soon.


  1. Chad Summerhill
    September 15, 2010 at 8:16 am · Reply

    Love this: “If you spend all your time figuring out what to do; and don’t do it – then maybe you should have done less analysis and more doing.”

    P.S. My boss just told me to sign up for Certified Knowledge and I’m looking forward to giving it a try.

    Thanks for the great post!


  2. Geno Prussakov
    September 15, 2010 at 10:25 am · Reply

    Very good point, Brad, and good illustrations/examples too. Any generic or average across-the-board metric is that way (misguiding and non-concrete). It’s one of those measurement pitfalls all online marketers should watch out for.

Leave a Reply