How to create a traffic map


New member
Jul 17, 2012
The steps you’ll follow to create your traffic map are as follows:
1. Collect site-wide statistics.
2. Identify the most popular page on your site.
3. Find out how most people get to your site.
4. Identify follow-up pages.
5. Repeat for your sales funnel pages.
6. Identify any “cross-feeders.”
7. Calculate theoretical CPA (non-revenue-generating funnels).
Once you’ve constructed a traffic map using these steps, you’ll need to interpret it.An invaluable
metric for this purpose is the $/Index (“dollar index”). This is best demonstrated with the
following tutorial.

We’re now going to walk through the process of creating a simple map.We’ll use real data from
a real site,, to illustrate the process.

Step 1: Collect Site-Wide Statistics
Begin by picking a date range long enough to capture a few thousand visits in your analytics
reports. The date range you select should also reflect ordinary traffic conditions. You want to
eliminate any one-time traffic bursts, such as holiday promotions, mentions on Slashdot, etc.
Visitors who arrive during these periods may be more or less motivated than your regular visitors
and may throw off your findings.
Next, consult your analytics reports to pull the following data:
- Total visits
- Total absolute unique visitors
- Site-wide bounce rate
You’ll find this information on the “Visitors Overview” report in Google Analytics (click on
“Visitors” in the left navigation bar)

Your primary website analytics metrics include Visits, Absolute Unique Visitors, and Bounce Rate.
These can be found on the “Visitors Overview” report

I chose a six-day period for Methuselah Foundation: February 2-7, 2009. The statistics I
recorded were:
Total visits: 3,596
Total absolute unique visitors: 2,904
Site-wide bounce rate: 57.42%

Step 2: Identify the Most Popular Page on Your Site
For many domains, this will be the homepage. There’s no need to guess, however. You can conclusively
identify this page by consulting the “Top Content” report in Google Analytics (found
under the “Content” menu on the left navigation bar)

The “Top Content” report indicates that the most popular page on Methuselah- is the root page (“/”). This report also contains three mandatory statistics that
we’ll collect for every page on our map: unique pageviews, bounce rate, and exit rate. Jot the
information you’ve gathered so far on a piece of paper like this:

All Pages
Visits: 3,596
Unique Visitors: 2,904
Bounce Rate: 57.42%
UP: 2,038
Bounce 54.4%
Exit: 54.8%

Step 3: Find Out How People Get to Your Site
Most of your visitors are probably arriving on your site through your homepage (the same page
we identified in step 2). However, some sites make heavy use of landing pages, and we want to
find this out now.
To get this information, visit the “Top Landing Pages” report in Google Analytics (again,
under the “Content” menu). This report will indicate what page is making the first impression
with your visitors. If it’s the same page as above, record the entrance rate on your map by drawing
an inbound arrow on the first box. If it’s a different page, add that page to your map, fetch
the same statistics we gathered in step 2, and record that data as well

The first row of the “Top Landing Pages” report will tell you how most of your visitors are arriving
at your site.

Here’s how our map looks at this point:
All Pages
Visits: 3, 596
Unique Visitors: 2, 904
Bounce Rate: 57.42%

Step 4: Identify “Second Step” Pages
Now we want to identify what pages your visitors are going to after they arrive at your site. There
will likely be dozens, maybe even hundreds, of these follow-up pages. Dont waste time overanalyzing
the situation, though. Instead, focus on just the top five.
Retrieve this data by navigating back to the “Top Content” report and clicking on the page in
the first row. This will bring you to the “Content Detail” report. Click the link that reads
“Navigation Summary.” The right column of the page will tell you where your homepage visitors
are going

The first row is often the same destination page as the source. This indicates that users may
be reloading the page. Dont worry about it unless it’s significantly higher than 10% (in which
case, you should investigate further to find out why people are finding it necessary to refresh the
Ignoring the first row, record the next five pages on your map (leaving space for the additional
statistics). Draw an arrow from your most important page to each of the “second step”
pages and record the percentage of clicks shown on this report.
Feel free to write down any notes while you’re constructing the map if ideas begin to jump
out at you (or if you simply need to remember what a cryptic URL points to).
To complete this step, go back to the “Top Content” report and fill out the missing statistics
for each of the pages you just added.

Step 5: Map Your Funnel Pages
When constructing a traffic map, it usually becomes apparent at this point that there are two sets
of important pages on our sites: the pages that visitors are actually visiting and the pages we
want them to visit (our conversion funnel pages).
If these two sets of pages dont overlap (which will usually be the case for all but single-page,
long copy sites), you’ll need to repeat the above process for your funnel pages.
Start with the first page in your funnel and retrieve the clickthrough rate from the page
before, as well as the unique page views and bounce/exit rates. Continue to do this for each page
in your funnel until it’s complete (use the “Top Content” and “Page Navigation” reports), then
continue the process for any remaining funnels you have

Step 6: Look Out for “Cross-Feeders”
Notice that we havent filled in the percentage of traffic coming from the email newsletter signup
form to our “thank you” page. That’s because we’re now going to work backwards to figure out
how people are joining our email list. This allows us to identify “cross-feeder” pages.
A “cross-feeder” is a funnel page that is receiving significant traffic from two or more pages

Developers and designers like using cross-feeders because it saves them time (they only have to
code the page once). However, they cause no end of headaches for search marketers. They require
you to manually separate out your traffic statistics for each source, and if there are any significant
differences in clickthroughs or conversion rates, you will be forced to unravel your data to answer
questions like the following:
- Why is Page A converting at so much higher of a rate than Page B?
- Why is Page B driving so much more traffic?
- What makes Page A visitors different from Page B visitors?
It can be very difficult to answer these questions with any confidence, so when I design a sales

funnel, I avoid using cross-feeders if at all possible. Instead, I duplicate these important pages
and give them distinct URLs. Good developers dont like to do things this way, so expect some
If you arent building a site from scratch, you dont have that luxury. You’ll need to use the
“Navigation Summary” report on your funnel pages to determine if you have any cross-feeders.
Click the link for your sales funnel page on the “Top Content” report.You’ll be directed to the
“Navigation Summary” report. Look in the left column to determine if there are any significant
sources of traffic feeding the page that arent already on your map

This report allowed us to identify a second page sending traffic to our newsletter signup
“thank you” page. In essence, we have two sales funnels being merged into one.We’ll now add
this page to our map along with the statistics from the “Top Content” page.
We can also use this page to figure out the conversion rate for each of the feeder pages. The
“Navigation Summary” report above indicates that only 3.12% of the 32 visitors (1 visitor) came
from the first email newsletter signup form. The remaining 93.75% (31 visitors) came from
another page, which is not currently on our map.
Use the reports in a similar fashion to figure out how people are getting to this second
newsletter signup page. Add this information to the map as you collect it (Figure 5-9).
It’s obvious that we have a real mystery on our hands here. In this case, the two pages were
virtually identical, yet one had a much higher conversion rate (29.2% vs. 0.6%). I had to sort
through a number of different reports to try and figure out what the difference was. It turned
out that the reason was because most visitors arrived on the second email signup page only after
clicking through various other pages on the site. As a general rule, the more clicks it takes to get
to a page, the fewer visitors you’ll get, but the ones who make it there will be very motivated! I
call this “Darwinnowing.”

Step 7: Calculate Theoretical CPA
If your conversions arent measured in actual dollars (for instance, you are generating leads or
signups for an email newsletter), then you’re almost done. If you arent tracking order amounts,
then your job is considerably easier than the one outlined in step 8.
Instead, you’re going to figure out how much each conversion would cost you if you paid
$0.25 per visitor. This is a lower cost-per-click price than you would typically pay on Google
AdWords. Being conservative here will help to ensure that we can indeed be competitive when
it comes time to start our PPC campaign.
In our example, 32 people signed up for the email newsletter during the six-day analysis
period.We had 2,904 visitors total, so the site-wide conversion rate is simply:

32/ 2,904 x 100% = 1.10%

This means that on average, 1.10% of the site visitors sign up for the email newsletter. Divide
the conversion rate into $0.25 to calculate how much it would cost you (on average) to get an
email newsletter subscriber.

$0.25/0.0110 = $22.72

What this tells us is that it will cost us nearly $23 to get someone on our email list from
Google AdWords, and that’s using a conservative $0.25 cost-per-click estimate. That’s far too
high a price for the Methuselah Foundation (a nonprofit) to pay. This exercise has shown us that
starting a PPC campaign at this point would be premature.
Last edited:
Older threads
Newer threads
Recommended threads

Referral contests

Referral link for :


Latest Blog ArticlesMost Viewed Threads

Popular tags

You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.