Forum Moderators: DixonJones
The column usually spans 3 or 4 pages, and we track drop off in page-views from page to page:
Column #1
Page 1
Page 2 down 23%
Page 3 down 21%
Page 4 down 14%
Column #2
Page 1
Page 2 down 27%
Page 3 down 17%
Column #3
Page 1
Page 2 down 37%
Page 3 down 22%
Page 4 down 7%
Column #4
Page 1
Page 2 down 31%
Page 3 down 21%
Page 4 down 8%
Column #5
Page 1
Page 2 down 26%
Page 3 down 7%
Page 4 down 9%
Column #6
Page 1
Page 2 down 27%
Page 3 down 14%
Page 4 down 6%
Column #7
Page 1
Page 2 down 19%
Page 3 down 11%
Page 4 down 12%
Column #8
Page 1
Page 2 down 35%
Page 3 down 29%
Column #9
Page 1
Page 2 down 62%
Page 3 down 36%
Column #10
Page 1
Page 2 down 68%
Page 3 down 9%
I'm looking at the last two columns and wondering why the train wreck.
Where would you look to explain such a major drop off? The feedback we're getting about the content is that it is among the BEST he's ever written, so the cause is probably something in the page code itself rather than the content.
#5 - August 1, 2001
#6 - September 4, 2001
#7 - October 14, 2001
#8 - November 30, 2001
#9 - January 11, 2002
#10 - February 9, 2002
> maybe slowdowns in page download time these past two times
Starting with Column 8, the articles became more graphic intensive. The extra graphics mostly come after the first page, and I use javascript to preload the images -- so rendering times are pretty decent on my 56k connection (4 to 7 seconds)
One thing I did notice -- with the extra graphics the articles LOOK longer, and each page takes more screens full, definitely more scrolling. So this month I split the newest article (and most graphic intensive to date) into 7 shorter pages. It went live today -- hope that makes a difference. If it does, I'll go back and chop up the two problem children.
The audience is mostly medical professionals and the columnist is widely respected in his field. So as his columns become more scholarly, better footnoted and more thoroughly illustrated, we expected a BETTER reception.
The number of people digging into the back columns every month is growing, as is the search engine traffic.
Can anyone suggest another metric I could check? I did look at incomplete downloads - very ordinary and steady, nothing helpful there. People were just deciding not to click through.
Column #11
Page 1
Page 2 down 18%
Page 3 down 9%
Page 4 down 10%
Page 5 down 6%
Page 6 down 14%
Page 7 down 7%
Total drop-off 50.5%
1. First of all, the time between page views. If people had become tired while reading a column, they would have decided to put off reading.
2. How many people have bookmarked the column? If a column is cool, but people haven't got enough time to read it carefully, they could bookmark it to read a column later.
3. Therefore you should check a number of returning visitors. The more this number is, the more likely people read columns twice over.
4. Did the structure of column's traffic change? Visitors could be worse targeted than readers of other columns. You should check sources of traffic, geographical distribution of visitors and so on.
I hope it'll help you.
Best regards,
Alexander Sadovsky.
It's possible, though, that any given column won't be as well-written or compelling -- even given a targeted audience -- as another. There are columnists I read regularly, but once in a while they write a snoozer and I bail out early... and I suspect I'm probably not alone on those days!
So at least part of what you're seeing may just be quirks in the content, not a factor of navigation or design.
