OT: Thanks swa66. My old paid version of Urchin doesn't have those links or anything about connection speeds. I had thought that Google Analytics and Urchin were nearly identical, but I guess not.
Time to upgrade to the free version LOL!
|It's certainly easy to A/B test fast/slow entry pages. Wouldn't that be a better way to test what consumers REALLY want? |
Right. It is also easy to guess the results, depending on your demographics.
|Buttom line: people do not want to wait AT ALL. Everything over 1 second feels like waiting. So IMHO this is the goal. |
|ok, I'm a bit cranky at the moment. That the customer won't wait 4 seconds is unreasonable, as most of us seem to agree. But does it matter what we think? Where do you make your money? From the customer! Any other answer is wrong. |
Agreed. Me the geek... I don't like waiting 2 seconds. Me the webmaster, I market everything to anyone as fast as possible.
This thread is another example of a familiar sentiment expressed at webmasterworld: that all web users are like us and that trends are more important than the current situation.
I don't know how many times people here have claimed AOL is worth "nothing" because it has a declining user base. (AOL is worth billions of dollars.) Gosh, I don't know anyone who uses AOL, why even consider it? says the naive webmaster.
Here we see the assumption that since fewer people are using dial-up, we should go with the trend and design for broadband. And not just DSL or cable, but T1. And that users are looking more for an experience like television than the web circa 1997.
If you all want to do 200K landing pages, it's fine with me. You're going to lose customers, but that just means all the more customers for the sites that provide a better experience.
|Agreed. Me the geek... I don't like waiting 2 seconds. |
I don't think anyone is suggesting that sites should be slow. Clearly the pages should render as quickly as possible but there is a limit. That limit is the time it takes your brain to process information.
You guys who are quoting one second and two seconds are clearly not being realistic. It is very difficult to make a decision not to wait for a second or two because it takes the brain that amount of time to process what is happening on the screen. Yes, this applies even when nothing is happening on the screen. The brain must recognises this and make the decision to go for the back button.
It's just not practical to do this in less than two or three seconds. Those of you who are claiming that you do must be surfing with your mouse pointer poised over the back button getting ready to click if nothing happens instantaneously. Can I just point out that you guys are missing out on a lot of Internet content? ;)
I don't know... I think I lean the other way... because I am a geek, I've been a webmaster, now mainly a network/systems admin I have a bit more patience for slow loading sites. The 2 second rule was for the most part as unrealistic then, as it is now. but it was/is a good goal.
Seriously, I do think knowing what all has to go on, how many electrons have to flow from here to a data center thousands of miles away call a process then output data back to electron from back to me... I do have more patience. Of course my browsing habits tend to have an average of ten tabs open and when I click a link it may be several minutes before that tab regains focus again.
One thing that may have had more effect on the stats and research, at least then, would be what was loading in those critical few seconds of human attention. Many people who quote those stats use them as one reason why google and other 'fast loading' sites had/have grown in popularity.
But in designing sites to load fast, the user interface was normally slimmed down and simpler. Sites like google and many others had clean, simple interfaces, which happened to load fast. In contrast search 'portal' sites had rows and rows of data and options for the user. Took a long time to load, yes. But more importantly, especially before everyone was web savvy, the result was a busy and confusing interface. Which in the end, I think had more to do with making those 2-4 second stats than the actual load time which was just a byproduct of clean pages.
One other thing that is lost in the explanation for this stated data;
"75% wont shop there again if the site has tech problems"
How much of online buying is impulse? We're obviously not talking the major players here. If I find an item at a few small online stores, I'll pick one. If that site won't let me finish the transaction I'll pick a different site and buy it there... I'll never remember the site I *almost* bought something at.
I really dont see any excuses at all to let your customers wait for more than 2 seconds on a broadband connection. If your pages load in 4+ seconds it is only a question of time (years) before 28% of your customers will abandon your site. So what are you going to do? Put wait statements in your code to prove that Akamai is wrong?
If the limit is not 4 seconds today, trust me it is out there somewhere, just waiting to catch up with the slow sites.
What are the arguments to not get the load times below 4 seconds?
looks like some people are not paying attention. ;)
|What are the arguments to not get the load times below 4 seconds? |
There are no arguments. If you read the thread you will see that no one is proposing that sites should not load quickly.
All I am saying is that striving to get less than two seconds is a waste of time for reasons stated in my previous post.
If people are willing to wait 5 seconds and you could offer a better user experience or better convey the reasons to purchase a product than your competitor by using that extra second, wouldn't you?
|What are the arguments to not get the load times below 4 seconds? |
On a well-built site the pages should load instantly, with perhaps a couple more seconds (max) for the images to come in. I mean instantly - so fast you can't time it.
Put it this way. Webmaster world is loading slowly right now, and it took about 15 seconds to load. I waited. It really matters if the content is worth the wait or not.
>> Put it this way. Webmaster world is loading slowly right now, and it took about 15 seconds to load. I waited. It really matters if the content is worth the wait or not.
well, not a good example; you know WebmasterWorld and the content. The idea is that instead of buying shoes from slow-shoe-site.com you go one of the dozen or so faster ones.
People will wait less than 4 second for the page to start loading. If they don't see something loading on their screen, they will leave. If they see the site is up, they will wait more than 10 sec for the page to load.
I think they reffer to 4 sec until the page start loading on the screen. If so, I agree with them. If I don't see the page start loading in less than 4 sec, I'll quit that site.
I don't buy the survey. Good luck to the 4 second worry warts.
Yes, and don't get RSI through having your mouse poised over the back button with your finger ready to strike. :)
If after 4 seconds you can see something loading then that is vastly different to a bloated site which remains blank.
Its all about usability, if a site is of clear open design, and doesn't waste time/bandwidth un-neccessarily, then people will enjoy using it and perhaps even buy more.
If instead your site looks like Yahoo's front page... then you really need to deal with your visual polution. Oh and don't use flash, this isn't a movie theater and we're not here to buy popcorn.
If you're a new or young site and need to gain a user base , page load time is very important. If you're older with an established base , not so important. Plus, the technology will catch up to you anyway over time. IMO
I cannot remember the times that my website would load within 4 seconds.
I 've had a ecommerce website for 4 years know , and lately my database is getting larger by the minute, thus increasing page load more and more.
I try to solve this with caching and optimizing queries..But, this is very hard to keep up with. Since i've insufficient funds to buy faster hardware and dedicated hosting, i have but one option and that is to settle with longer page loading times.
Actually when the page load isn't taking about 15- 30 seconds I am happy! I rely on the visitors to decide for themselves how much they will take. Most of them do know why it is taking so long( eg they are searching for a product and the output would take a long time)therefore the've peace with it.
So what can i learn from such a survey: If you have the bucks, try to buy the fastest hardware as you can , try to have several mirros, make the bst caching system, present your very stripped homepage with only a nice search form on an almost blank page. Scrape any content you use from other parties and put this on your own( fast) mirrored server! Hey you have just joined the big leage ;-), now write your own survey and say that the four second rule doesn't apply anymore because people aren't willing to wait longer than your own pageloading time.
I agree - 4 seconds for a full load is expecting a lot. However, if a site doesn't START to load visually after 5-6 secs, then it does start to get annoying (particularly with a cable modem!).
Just for giggles I tried to access Akami's site via google - here is the message I got (11/19/2006):
"You have reached this page by going to an Akamai Technologies registered domain. In 20 seconds, you will automatically be sent to www.akamai.com."
Thank you for your continuing interest in Akamai Technologies."
4 sec it's fine.
Are these 4 real seconds or 4 perceived seconds?
I have broadband, and sometimes it can take more than 4 seconds for the domain name to resolve. I wouldn't be looking at many sites, if any, at all if I didn't wait more than 4 seconds.
But, the question is, is someone sitting behind people with a stopwatch to determine how long they actually wait for a page to load, or is it just based on how long people say they wait?
Because, how many people have no grasp on time? When someone asks you if you have a second (or a minute), how often is it really a second or really 60 seconds that they're wanting from you?
The smaller your page size, the quicker it can load, all for the better. But, I would be distrustful of a survey that (a) is done by a company whose revenue derives from making people's sites faster (though, in my experience, web sites are frequently stalling when trying to fetch something from an Akamai server) and (b) is based on responses and not scientific evidence.
They may say 3-4 seconds, but what they might really mean is 6-8 seconds. Hard to know for sure unless someone is watching them with a stopwatch while they shop or surf online.
| This 80 message thread spans 3 pages: < < 80 ( 1 2  ) |