|Paging Or Scrolling|
Surfer Perceptions - usability study from wichita.edu
| 9:23 am on Feb 8, 2003 (gmt 0)|
I was checking out the usability studies and I was pretty suprised that people don't seem to mind scrolling.
I really hate pages that scoll much past the browser chrome and I've always advocated putting the important content above the fold. I've seen a marked increase in conversion rates when the important info and the money link is above the fold. 2-5% better.
The second reason I tell other webmasters to put the important content above the fold is more subtle. People tend to lose focus if they are allowed to write many paragraphs and put them on the page, by asking them to put important content above the fold each page ends up stressing the important info near the top.
I also prefer short, focused pages for SEO purposes. More tightly focused pages, more phrases to target.
The results of the study I was reading indicated that people really don't mind scrolling. They had a slightly better comprehension rate if the paging technique was used but it took them longer to find specific information.
I use the "full" technique more often than not but the study didn't really show any significant improvement in any areas other than search time and comprehension for any method.
With all studies I'm sceptical of the results and moreso with studies that use small samples. I'm also wondering about the actual text used. If they just broke up the text and created new pages then the entire reason for creating short pages (from my standpoint) is lost. Pages that require people to click to get more content need to be designed specifically for paging rather than just being arbitrarily broken into chunks of text.
Just wondering how other people interpret these results and if anyone else is conducting studies on their own.
I can see where people might perceive longer pages to be more informative but I think everyone is in agreement that people scan web pages more often than read them word for word and it seems like the paging and full techniques would facilitate the scanning method.
[edited by: tedster at 10:39 pm (utc) on Feb. 10, 2003]
[edit reason] make a working URL [/edit]
| 12:44 pm on Feb 8, 2003 (gmt 0)|
Last year I did a study for one of my clients on this topic. Over a three month period we did a split run test in several critical site areas. We saw that their audience preferred scrolling over frequent clicking - at least up to a point.
One flaw in our study was that we measured screen resolution and not availHeight. Over 90% of visitors were running between 800 and 1128 screen resolution, so we assume that great the majority were running at or near full screen windows. But we didn't actually measure this, so the question of where the "fold" actually fell in each case isn't known.
This site has about 1100 information pages, and about 100 e-commerce pages. We tested both types of pages and the results were nearly the same.
For example, a monthly column with a strong readership was configured three different ways. We ran a 3-page version, a 5-page version and a 9-page version. The choice of which version a visitor got was determined by a random function.
Measuring click-throughs to the final page of the article, we saw an overwhelming preference for the 5-page version. On a full screen 1024 monitor, the 5-page version ran about 5 screens-full per page.
In fact, we had the strong suggestion of a very STEEP bell curve. At 9 pages, only about 13% of the readers made it from the first page to the last. But at 5 pages it was 49%. For the 3 page version, the number of people completing the article fell off again to 21%.
If one of my clients had the interest, I think it would be worthwhile to try serving up different configurations based on available height of the window. I would aim to give each visitor a version of the article that required no more than, say, 6 screens of scrolling. Alas, no one so far wants to open their wallet for that experiment.
| 12:49 pm on Feb 8, 2003 (gmt 0)|
>>In fact, we had the strong suggestion of a very STEEP bell curve. At 9 pages, only about 13% of the readers made it from the first page to the last. But at 5 pages it was 49%. For the 3 page version, the number of people completing the article fell off again to 21
Just basing my opinion on that it would seem to indicate that the "full" condition version in the usability study is the preferred method.
Thanks for the info. Looks like I need to study up, and provide a working link :)
| 1:26 pm on Feb 8, 2003 (gmt 0)|
Just from personal experience: I hate news articles split up onto several pages. Waiting for the layout, banner ads, menu etc of the next page to load totally breaks the reading flow and makes it hard to get a quick overview.
If your text has very distinct chapters, like a tutorial (1., Install that program - 2., Do this and that with it), pages make sense: they communicate a sense of achievement (phew, done with step X, halfway through). But other than increasing page/banner views, I don't see any advantages.
I'm willing to bet that without this condiditon, they would have seen different results:
"The layouts were stored on a local server, virtually eliminating download time in each condition."
| 2:36 am on Feb 9, 2003 (gmt 0)|
>>The layouts were stored on a local server, virtually eliminating download time in each condition
I see your point, but for the sites I have input on the pages are virtually the same save for the body text, images are cached and there aren't any delay issues.
One of the issues they didn't address was text format, headers, bold, etc, all things that can change how fast information is found. The didn't mention the eye travel involved left to right either. I'm starting to use multiple columns on a few sites so that there is very little left to right eye travel and the Newspaper style seems to reduce scan time.
| 4:21 am on Feb 9, 2003 (gmt 0)|
|One of the issues they didn't address was text format, headers, bold, etc |
Agreed - and if those screenshots represent the acual test pages, they used much thicker blocks of text than I normally would.
| 8:34 pm on Feb 10, 2003 (gmt 0)|
Thanks for all the comments so far on the study! I was just as surprised at the results as some of you seem to be! I'd just like to make a few comments and hopefully clarify a few issues for you.
>>I really hate pages that scoll much past the browser chrome and I've always advocated putting the important content above the fold. I've seen a marked increase in conversion rates when the important info and the money link is above the fold. 2-5% better.
Absolutely! There’s plenty of research that suggests important information should be kept above the fold. I don't mean for this research to contradict those studies in any way.
>>I use the "full" technique more often than not but the study didn't really show any significant improvement in any areas other than search time and comprehension for any method.
Some of our other research shows that full is probably the best way to go, especially when people are scanning. The first one deals with "news" -like content, and the second with search results.
>>With all studies I'm sceptical of the results and moreso with studies that use small samples. I'm also wondering about the actual text used. If they just broke up the text and created new pages then the entire reason for creating short pages (from my standpoint) is lost. Pages that require people to click to get more content need to be designed specifically for paging rather than just being arbitrarily broken into chunks of text.
Yes, the sample is a little small (such is the nature of recruiting). As far as the text used, they were passages from sample ACT tests, and each had four paragraphs. So, there wasn’t really an “arbitrary” breaking of the pages; we used the natural logical breaks already provided for us.
>>I can see where people might perceive longer pages to be more informative but I think everyone is in agreement that people scan web pages more often than read them word for word and it seems like the paging and full techniques would facilitate the scanning method.
And in fact, other research we’ve done suggests exactly that. We’ve got several studies on scanning web pages. In this case, I wanted people to examine whether paging/scrolling had any effect on comprehension. (It didn't.)
>>I'm willing to bet that without this condiditon, they would have seen different results: "The layouts were stored on a local server, virtually eliminating download time in each condition."
You’re right, and that’s something we have to consider every time we do a study, but in this case, we decided it would add too much of a nuisance variable.
>>One of the issues they didn't address was text format, headers, bold, etc, all things that can change how fast information is found. The didn't mention the eye travel involved left to right either. I'm starting to use multiple columns on a few sites so that there is very little left to right eye travel and the Newspaper style seems to reduce scan time.
In this case, I didn’t want to add any features to the format that would help participants remember particularly well where specific information. I just wanted to know purely whether paging/scrolling had any impact.
>Agreed - and if those screenshots represent the acual test pages, they used much thicker blocks of text than I normally would.
Those are actual screenshots of different conditions. The blocks of text were approximately 70 characters across. Since the study has been completed, I’ve come across some research that suggests closer to 50 char/line is more optimal – so that’s something I may look into in the future.
Again, thanks for all the great comments!
| 10:18 pm on Feb 10, 2003 (gmt 0)|
Welcome to WebmasterWorld, Ryan_Baker. It's good to have you here.
We've taken a good look at many of the usability stidues on wichita.edu -- I'm very pleased to see such valuable information being make widely available.
|There’s plenty of research that suggests important information should be kept above the fold. |
Certainly key information about the page should be above the fold. That way the visitor sees "where they are" and "what they can do from here" on their first glance. But no less an authority than Jakob Nielsen has conceded that VERTICAL scrolling itself is no big deal today, although it was in the earlier days of the web when people weren't used to it.
I bring this up because having important information above the fold is not the same as having the entire page fit above the fold - the way it did in this study.
One major factor a designer must consider if they are concerned about "above the fold" is what resolution are we talking about here? In the old days (when several often-quote studies were done) there was only 640x480. Today that antique resolution represents only 3% of users.
However, with the nearly certain rise of web browsing on wireless phones, where the fold is rendered becomes much trickier! Try Opera 7 browser on its "small screen" setting and you'll see what I mean.
| 10:44 pm on Feb 10, 2003 (gmt 0)|
Back in those days the first wheeled mice were fairly expensive. It's now almost impossible to find a mouse without a wheel. This makes scrolling more comfortable and much smoother than using the scroll bar (the old-fashioned way). We were continually adjusting our eyes as to where they should focus amongst all those fits and starts.
I tend to think that the wheel also makes the longer page easier to read than multiple pages -- you don't have to find the "Next" button and go through everything needed to click it. And then don't have to adjust your point of view and find your place on the new page.
And if the navigation isn't set up well, when you're six pages deep in an article and want to go back...
| 2:28 am on Feb 11, 2003 (gmt 0)|
I have to wonder how Amazon's increasingly byzantine page layout is holding up. I find it annoyingly long to download; it looks like every department has its fingers on a link somewhere on the page.
| 4:28 am on Feb 11, 2003 (gmt 0)|
Can anyone explain, then, why a lot of sites insist on using VERY long pages to sell from?
I've seen sales pages for lots of e-books, services, etc, where it's all in one VERY long page (say, at least a thousand words) and goes through everything about the service, and is really just a 'rally' to get you so stirred up that you'll buy.
Lots of the Internet marketing people are using these extremely long 'sales letter' style pages.
If they keep doing this, doesn't it prove that scrolling really isn't a problem in online sales?
| 5:22 am on Feb 11, 2003 (gmt 0)|
I think those long selling pages operate under the assumption that you don't want to lose anyone in the middle of your pitch by making them click.
Overall, I think it's important to remember that "one size does not fit all." ;) Different content calls for different layouts. Different HTML strokes for different folks. ;)
Certainly many magazine type sites are using multiple pages for a single article so that they can increase their pageviews/ad impressions since they are selling on a CPM basis. I hate that especially when I want to print out an article for reading later. Some sites even keep the multiple page version intact on the printable version. Yuck!
In the affiliate space, we have a whole different challenge, and more than pagination, the use of graphic sub-heads and other typographic devices make a page scannable, and thus increase visitor time at our sites, as well as "comprehension" while browsing. Do any of you think that breaking up longer articles into more pages helps affiliate sales?
And given that for Google, at least, we're looking for at least 300 - 500 words, as I understand the requirement, it's pretty tough to keep it all above the fold anyway, isn't it?
Just my 2 cents...
| 8:10 am on Feb 11, 2003 (gmt 0)|
Recent discussion on long sales pages
| 2:10 pm on Feb 11, 2003 (gmt 0)|
I think it depends on what content you have. Lets say is a source code site good for scrolling or not, And i think it is because you want a continues flow of data and thats what people hould think about continues flow.
| 2:53 pm on Feb 11, 2003 (gmt 0)|
I went and found some long MLM pages, not hard to do as they all seem to be in that format and read the copy and it looks like the call to action words are sprinkled throughout the entire page and the goal is get people to read the same call to action phrases three times. :)
Since MLM is a successful industry,(anything that appeals to base greed and easy opportunity is successful) the copy and the format seems to work.
What I didn't find was large text blocks. The entire page was formatted in small paragraphs with call to action headers or headers anticipating the mind's objections. They followed that up with a reason for action or information designed to quell objection.
What I didn't find was quick info regarding how the process worked, or the monetary consideration involved. Seemed to all be hype followed by, well hype. The only link off most of the pages was to a CC form.
In other industries the "full" condition, with proper headers seems to be the norm. "Add to Cart" was almost always seen next to "More Info". :)
Another reason for the MLM "long page" treatment might be that those sites are almost all template based and copied word for word from the affiliate program. The only difference is the affiliate ID.
| 5:49 pm on Feb 11, 2003 (gmt 0)|
I happen to be a usability junkie, because all of this is based on adjusting our site design and structure towards common sense and making sure it is usable to the users.
They are at point a, and they want to go to point c, so we have to have ways to easily go to point c, either directly, or step by step. or we may offer, at point c, that point d is equally as valid or even point d.
BTW I highly recommend Jakob Nielsen's book on HomePage Usability, because the book has color photos of websites as he deseminates the usability aspects.
I find the biggest struggle is the competition between marketing people who want to find fads, or cool gui design, that have nothing to with what customers are looking for.
This is a major problem, because we let the desire to creatively sell interfere with common design standards and principles.
Here is my standard principle.
If the customer can't easily figure out how to use it, then let's not waste our time, to create something cool that can not be actually useful.
| 11:07 pm on Feb 11, 2003 (gmt 0)|
In 2001 I worked as copywriter on a website that was completely above the fold - that's what they wanted, down to dictating 781 characters of text as a per-page limit. So I did it.
The website did them almost no good at all. It looked beautiful, but when it came down to people visiting, taking in the information, and then responding -- nada. Server logs showed very little extended reading. After all how many times are you willing to click around a site to take in all the fragments and learn what you want?
There's another practical problem with using such short pages. If you have something of substance to communicate, that creates SO many pages. Navigation quickly becomes quite complex and usability suffers.
My take on it today is that the "above the fold" website is anattempt to duplicate the print brochure online. It's not a "native" use of the web.
| 1:19 am on Feb 12, 2003 (gmt 0)|
We have a very long site with many news articles. We realized that almost everyone:
1. Hates waiting for new pages to load
2. Has a scroll mouse (except Mac users, the fools!)
So we made our pages long and it works great. No-one has ever complained in two years.
As the Tablet PC becomes more common I believe that pages will become standard A4 shape ie taller and that web-pages in general will become more like magazine pages. This will lead to more interstitial pages and more ads at the bottom of pages.
We are already making everyone come through a special advertising page BEFORE they enter our site and will shortly be extending this idea to full interstitials between headlines and stories. No one has complained and the advertisers are very happy.
That's partially because we publish a magazine and these ideas work well in print.
| 12:55 am on Feb 14, 2003 (gmt 0)|
Boy, this is timely. We are about to launch (~ 2 weeks) a split server test on just this issue.
I fought for scrolling, I compromised by saying, "Let's test ..." so I'll know in about 4 weeks what works for our site.
| 2:26 am on Feb 14, 2003 (gmt 0)|
One of the strategies we've developed for in depth content is what I call modularization. Essentially the first page has several more about [this topic] [example.com] links spread through the content. Those links bring up the longer pages, and the main page remains more of an overview.
We've found it also helps to style the page very clearly. For instance, one bit of content included testimonials for a seminar presenter. We created a div with a subtle background color that held a group of one-liners with the names of the people who made that comment. And at the bottom we had a link "read the full testimonial".
The page visitor could easily see where the testimonial section began and ended. If they wanted to scroll past it, it was an easy job. If they wanted to dig deeper, that was also clear but we didn't force all the content on every reader.
Similar treatment was given to an overview of the seminar topics, and so on. So we ended up with just some scrolling for the lead page, and lots more for all the detail.
This strategy proved so effective (the seminar sold out in record time) that we've begun doing it a lot more - and it requires some re-training of the copy writer.
Today when I first get a content draft for editing, I often lift out 4 to 6 paragraphs from here or there to use for the "more about" link -- and move them to their own page. As an additional benefit, those tightly focused "more about" pages make great search engine food.