May 1, 2007
From the desk of Jane Galt:I have an Andy Rooney moment
What the hell is with multiple page web articles? The reason books have pages is obvious; it's hard to carry around a single 110x80 foot sheet of paper, much less unfold it to read. Not so much for web articles. Does someone actually find this preferable? I hate having to click and wait . . .
Posted by Jane Galt at May 1, 2007 4:57 PM | TrackBack | Technorati inbound linksMore page views, more ad revenue?
I detest having to click through 3 or 4 pages to read a single article. However, multiple pages means multiple ads, thus more opportunities for revenue for the site, so I understand the motivation behind it.
A decade ago, "usability expert" Jakob Nielsen (then employed by Sun) started claiming very strongly that users prefer multi-page web articles because scrolling continually is... I don't know, arduous? Confusing? Tending to cause middle finger strain?
This was the result of extensive usability testing he did (so he says). When he describes it, it's basically focus group testing. We all know how accurate focus groups are.
I hate multi-page articles too.
If the wait for one page out of three is annoying, imagine how much more annoying it would be to wait for a all-one-big-page version of the same thing to load. At least with multiple pages, if you choose to quit reading halfway through, you won't have been forced to wait for the bits you didn't read.
Just click on the "Print version" links. Then you lose the annoying ads, AND everything's on one page.
Personally, I prefer a single long page on my screen: indeed, when I read an article on a site that deliberately divides it into multiple pages (e.g. the New York Times) I always click the "One-Page Version" button before proceeding.
The slowest (and, invariably, the first) things to load are the ads: I'd rather get that out of the way once and for all when I first load the page, so as not to have my continuity of thought interrupted.
Having a single long page is nothing new even in the physical world, of course: it was good enough for the Greeks and Romans for millenia. Multipage documents are largely an artifact of the printing revolution, not a change for the convenience of the reader.
The thing I really hate are articles with multiple pages where you can't easly locate the "single page" button or, even worse, there isn't any.
...
I almost thought you were having Kate guest-blog for a minute there.
Even Homer nods, I guess.
There's two ways to do it: you can read the printer friendly, which is less distracting because there are fewer ads and pictures or you can open the second page in a new tab (or window if you don't have tabs) while you read the first page.
Readers may not like it, but the economic motivation is clear. A different set of ads can be served on each page, therefore generating higher revenue.
The worst is the ad which appears separately between the pages and cannot be bypassed until IT gives you the option to go to the page you requested.
Note that the more media sources do that, the more others must follow. Advertisers began to mentallly adjust the reported numbers downward (to account for clicking through), and so the one-click pages have to jump to the industry standard. So there is even a negative externality to the practice you so abhor.
Justin JJ's snark is just a bit off-target; Dr. Nielsen's findings incorporate elements of common sense.
To judge by personal experience, there's a pretty strong correlation between a user's comfort with the written word and their dislike of article pagination.
To judge by still more personal experience, tons o'scrolling is bothersome for folks whose eyesight leaves something to be desired. Like, just about everyone over 50 years of age.
And the rest is for the sake of being able to serve more ads, to get the maximum leverage from the F-scanning phenomenon.
Besides, the painless approach isn't to the single-page version, but to the print version. How did I manage to be the first to bring that up?
Someone already mentioned ad revenue; another possible motivation may be to conserve bandwidth.
If, say, 15,000-50,000 visitors click on any given story link, and less than 20% of them are likely to read past the lede paragraph, AND you divide the story into two or even three pages (of which the user's browser can be instructed to pull most of the repetitive elements from local cache after loading the first page, meaning that only new elements are served for each successive page click)...then the "800 bytes here, 1.3kB there" effect could add up into the gigabyte range before the monthly bandwidth bill comes due.
At any rate, I generally do like having one page, also.
It's mostly ad revenue. "above the fold ads" pay the most and if you space out an article you potentially get twice as many page views per story.
If Nielsen's work was done for Sun, that long ago, it was probably based on text, without advertising. And long enough ago, may have been a justification for keeping documents in separate files, to reduce cost of storage and manipulation with early word processors (time = money, it took a while for stuff coming from and writing to a 360k floppy).
Ideally, I think, the pagination should bear a bit of resemblance to dividing the article into sections or chapters. That is, a logical break in the document should coincide with each physical page change. Perhaps, before web browsers all implemented internal link destinations uniformly, the only way to jump to a new section of a document would be to put each section on a page.
And for a group project, leaving each section/page would allow simplified management of independent updates.
Personally, though, I think it is simply about getting more ad impressions.
Multiple pages do make it easier to cite specific passages in an article.
me is incorrect. If a page takes a long time to load over a broadband connection, it's the fault of the graphics, ads, buttons, icons, cutesy patterned background, links, etc. A few thousand words of text in a single field loads almost instantaneously.
I prefer single page, but multi-page is good for reading long things where you may want to stop reading part way through and come back to it later. Just bookmark the page you are on when you stop.
Obviously, this could be fixed by having a bookmarking mechanism based on a given line or word selected in the text (this is supported by HTML but not, as far as I know, any browsers).
A disadvantage of multiple pages is in searching for text, which one "page" makes easy.
If multipage is tough on older eyes, there are two easy ways to help:
1. White space.
2. Larger font sizes.
Multipage is for the ad revenue. Last gripe: why are there no longer two spaces after a period at the end of a sentence? It makes everything more readable.
Sigh. This thing strips all the spaces after a period.
ben: What does eyesight have to do with scrolling? I'm 53, wear bifocals, and also have a second pair of bifocals with an intermediate prescription in the main lens for computers and working on electronics hardware. I can see how arthritic fingers might make scrolling harder (like everything else), but not old eyes.
I think Nielsen's research must have been with computer newbies that didn't know what scrollbars were. In any case, with on-line newspaper or magazine articles, I've almost always got to scroll anyway to reach the bottom of the page, _then_ click and wait for page two. In some ways it is better to get a quick-loading first part of the article so I can decide if I even want to read all of it, but couldn't web designers provide that in a one-page format just by sending the text of the first paragraphs _before_ loading the fiftyeth advertisement?
Multiple pages do make it easier to cite specific passages in an article.
If you want to guide readers to a specific passage on-line, tell them to search for a particular phrase - but as others have mentioned, first they've got to find that button for the single-page version.
Thank you! And what's with pages that refresh themselves. In the middle of a sentance, the screen resets and moves me back to the top of the page. (I'm sure it has to do with the prominence and frequency of certain types of ads in both cases.)
Hmmm... is there a correlation with successful on-line revenue operations and article format. I don't think the few successful online papers out there do the multipage articles or persistent refreshes and add pings, coincidence?
[Chris:
There's two ways to do it: you can read the printer friendly, which is less distracting because there are fewer ads and pictures or you can open the second page in a new tab (or window if you don't have tabs) while you read the first page.]
The new tab bit works fine *if* you know up front that it's a multi-page article, and if you're not on dial-up(*). I bring in a bunch of stuff to read and disconnect. Finding a continuation at the bottom of the page just means I may not bother with the rest of the article.
[Aaron:
Thank you! And what's with pages that refresh themselves. In the middle of a sentance, the screen resets and moves me back to the top of the page. (I'm sure it has to do with the prominence and frequency of certain types of ads in both cases.)]
Insty and Dr H. have started this recently. Once again, if you're on dial-up and not connected, the refresh results in an error message.
(*) Yes, dial-up is out of the 90s, but if I had broadband I'd never leave the computer. Count me in the 'no podcasts, Jane' group.
The "more page views/ad revenue" and "more space for ads" rationales probably don't capture it. Advertisers are smart enough now to know that racking up page views isn't a very significant measurement anymore.
I think the adds change, and they also get data on how long and where on the page you are. How far you read into what type of content and authors etc.
Yes, I hate that. I also dislike the trend towards "click to read the whole thing" blog entries, and various hover-over links and widgets that pop up when you don't expect them. It's just an endless series of tiny barriers between you and the stuff you're reading.
Sometimes I use a text-to-speech program so I can listen to web articles while I go about mindless chores. Multi-page articles really make that a pain. It occurs to me that some handicapped people, who use tools like that out of necessity or otherwise have a hard time with content that comes in tiny little chunks, are probably exceptionally put out.
Only the propriatrix of a seldom-visited, negative revenue, substance free site would ask such a stupid question.
One space after a period has become "industry standard", allegedly because word processors do proportional spacing which typewriters couldn't. Having grown up w/ the two-space rule, I also find the one-space practice a bit odd too. But it really does depend on what the printer is doing.
...Regarding the vision issue, consider the eyestrain and vertigo that result from being forced to find where your former place in the copy, every time you scroll. Doing it once or twice on a page is one thing; doing it umpteen times gets old, fast... or maybe I just notice this because I have the Uncorrected Astigmatism From Hell and I read those kinds of pages practically all day, every day.
Dr. Nielsen bolted from Sun as soon as he was certain he could make a better living as an indy. He's very good at communicating to business decision-makers, enough so that other shops have adopted his approach.
Quite often the load time on pages has nothing to do with the actual bandwidth, but with the amount of heavy lifting the computer has to do. On a typical BigMedia page you're looking at connections to three or more servers, numerous images (each of which carries connection overhead since each is handled via its own HTTP transaction), perhaps Flash (which can be resource intensive), and finally poor production practices such as nesting of HTML table elements and failure to specify the heights and widths of images. When you account for all of the issues I just mentioned, you're looking at a lot of bottlenecks.
Firefox users, meanwhile, can fix what are typically the worst of these issues by typing about:config into their Location field, pressing Enter, and changing the values of the following keys that are displayed:
- network.http.max-connections-per-server
- network.http.max-connections-per-server
The low default max-connections-per-server value is particularly brutal on load times.
Giving the proprietress a hard time about the fact that she does not get avaricious about her revenue is in bad taste, to say the least. Where's the gratitude? I've got a bit to spare, should I share it?
Ad impressions *do* matter with respect to revenue; meanwhile, click-throughs matter even more. If you're a BigMedia site with a tendency to throw mud at the wall with respect to your ad coverage, then it's seemingly logical to conclude that the more ads you serve to any given visitor, the more likely you are to get click-throughs. Just sayin'.
"Read More" links signal to the reader that the entry or article in question gets windy, quite possibly on a subject about which the reader could really care less.
The absence of leading on long passages is a sure sign that the site hasn't been redesigned in a while; common and pain-free support for increased leading has been around for more than five years.
...And the second of those two Firefox config keys was meant to be network.http.proxy.pipelining BTW.
D'oh.
ben - thanks for the tips. I agree that scrolling has its problems, but I still think it is by far the lesser evil compared to loading a new page. A fast page load is only slightly annoying, but the problem is that there is a very long tail on the distribution of page load times, and long load times (from many seconds to never) are common enough to cause a feeling of helpless dread in those first couple of seconds waiting for any page. And multiple pages make it inconvenient to search, copy, or backtrack.
If better coding practices and configuration settings could help alleviate these negative experiences, then perhaps people would become more accepting of multi-click content. The Winds of Change blog is exemplary in this respect -- when you click "read more," the new text appears instantly, and is exactly positioned with respect to the part you've already read. But I would still prefer to scroll.
not fond of on-line newspapers that place an article on multiple pages & have a link to get you to jump ΒΌ screen past an ad, but those don't bug me nearly as much as blog postings that show part of the article and then have a "below the fold" section. Do bloggers make you click for the same reason - ad revenue?
THS on May 2, 2007 9:07 PM:
One space after a period has become "industry standard", allegedly because word processors do proportional spacing which typewriters couldn't. Having grown up w/ the two-space rule, I also find the one-space practice a bit odd too. But it really does depend on what the printer is doing.
It is every writer's duty to resist and reverse the emerging typographic barbarism, which holds that a single space should separate sentences. No. The experience of centuries has taught typesetters to insert a wider space after sentence endings, a visual gap between ideas to reflect the natural inclination to pause for breath while reading the text aloud. Extra space also serves to distinguish ends of sentences from abbreviations.
When setting a line of movable metal type, the standard practice is to compose a line by inserting quad (half en) spaces between words, and a full en space between sentences. When the compositor reaches the end of a line (ie, the next word won't fit), s/he inserts extra bits of space -- first between sentences, then between words -- until the line is justified against the right margin. On a monospace typewriter, you'd approximate the typesetting standard with two spaces after the end of a sentence. (em = width of capital "M"; en = 1/2 em; quad = 1/2 en = 1/4 em).
A glance at the pages of any book typeset before 1990, and most books 1990-2000, will demonstrate this convention -- and, I believe, its aesthetic wisdom. Proportional fonts and extra space between sentences have been typographically normative since approximately the time of Gutenberg. It is programmer's laziness, not the advent of proportional spacing in word processors, that explains the aesthetic decline. The unfortunate trend toward word-spaced sentences was accelerated when the first web browser appeared, for it was far easier to render all the whitespace (tabs, spaces, linebreaks) in HTML'ed text as a single space.
I blame this decline in aesthetic standards on the General Decline of American Public Education. The junior high school curriculum required (working class) boys of my generation to attend classes in the "industrial arts": a semester for each of drafting, graphic arts, metal, and wood shop. We enjoyed this daily respite from academic tedium while girls suffered through an ordeal called "home economics".
Astute observers might have noticed the double-spacing between sentences in this comment. To accomplish this trick, I appended a non-breaking space (in HTML, " ") to the period at the end of each sentence. For comparison, the initial quote by THS has single-spaced sentences, common to contemporary webbed text.
Boycot.
No. Not the publications. The articles.
I will no longer click through multiple links or the print link. I will just close the window.
Hopefully the publishers will begin to include a print version link right next to the standard article link at the least. Or at best, provide full length articles only.
db: Can't agree. As someone who reads old typographical manuals for pleasure, I can assure you that the better class of typographer has been railing against extra space between sentences for at least a century. Typographers tend to fetishise an evenly textured page, and the extra space creates a white hole in the text.
David Hecht: No, the Greeks and Romans scrolled sideways, with the text in short, parallel columns - very much like the web convention we're discussing.
Personally, I like it for longer articles, since I have trouble finding my place again when I scroll vertically.
I intended the phrase "better class of typographer" to have a hint of sarcasm. I mean the artier ones, associated with fine-press movement.
I have been an advocate of two-spaces-at-the-end-of-sentences for a long time, for better readability. I've had people actually assert that a larger space at the end of sentences is built into word processors.
I always ask "How does a word processor know the difference between a period at the end of a sentence and a period after Mr. or Dr.?"
It's fun, the dead silence that follows.
Comments are Closed.