Why does Google Search still not support infinity scrolling? Ads?
Its pretty ridiculous to think that the most visited web pages still doesn't enable users to scroll infinitely through search results. Is it to preserve ad revenue? Image search has infinity scrolling.
32 comments
Why should Google support infinite scrolling? Because all the cool kids are doing it?
I hate infinite scrolling. I don't know why people think it's a good idea. If you like it, reply to this comment and tell me why, because I really, totally don't see the value in it.
I hate infinite scrolling. I don't know why people think it's a good idea. If you like it, reply to this comment and tell me why, because I really, totally don't see the value in it.
Google regularly tests features like these, and has a lot of data with which to make decisions, versus positing that it is simply "ridiculous" that it isn't enabled.
you can install some add-on to enable that function
The question was more of a curious one, I'd like to see what others think Google's rationale is for sticking to page by page.
Perhaps their usability testing shows most users don't like scrolling (or that those that do prefer image searches?)
That seems highly unlikely, and to my knowledge, they haven't even A/B tested any kind of endless scrolling on the web search.
The most plausible reason I've seen was (I believe) in 'don't make me think' -
The ideal search engine would return one result - exactly what you're looking for. If google isn't consistently return the relevant result in the top 10, it's already failed as a search engine.
Research has shown that(iirc) <10% of users go past the first page of results, and over half of clicks are on the top 5 entries.
he also states that an original reason was in order to optimize for speed rather than # of results, since people tend to modify keywords and research far more frequently than visiting the second page.
Disclaimer: I may be misattributing this, don't have the book with me.
he also states that an original reason was in order to optimize for speed rather than # of results, since people tend to modify keywords and research far more frequently than visiting the second page.
Disclaimer: I may be misattributing this, don't have the book with me.
Google runs thousands of A/B tests...there is a good chance you've just never seen one.
Regardless, there is published data about how people click on the top 3 links the most, and click dropoff is precipitous after the second page. They simply don't need to do infinite scrolling...most people will requery the engine before checking the second or third page.
Regardless, there is published data about how people click on the top 3 links the most, and click dropoff is precipitous after the second page. They simply don't need to do infinite scrolling...most people will requery the engine before checking the second or third page.
because it sucks!
Well, I don't remember the last time I had to navigate past Google's first page.
I only find what I'm looking for on the first page about 1/3 of the time myself. Even after some refining of terms.
Because they cannot. they only serve 1000 results,
http://statspotting.com/2012/12/an-interesting-fact-about-go...
http://statspotting.com/2012/12/an-interesting-fact-about-go...
That would provide 100 pages worth of scrolling. And scrolling through 1000 items would be a huge improvement from their current 10 results per page.
10? Go into the settings and put it at 30 or 50 or 100 if you want more than 10. It'll be far more reliable than infinite scrolling.
I've just done this. I never check the settings page so thanks for the advice!
Note: Google Instant gets turned off if you want more than 10 results per page.
Note: Google Instant gets turned off if you want more than 10 results per page.
What is funny is you can create a your own search engine running Google's API and you can dictate unlimited results of the 1st page (Bing offers the same option for their search API).
I created such a search engine myself, but instead of displaying the standard text results I displayed only the website favicons so as you scrolled down infinite favicons appear, click a favicon it takes you to the webpage.
I think nickporter hit the nail on the head about the first page of results being so important and as you mentioned the competition to be on Google's 1st page is good for Google financially. However, one complaint about my search engine was that since the results often repeated the same favicon users thought the results were subpar, never realizing the results were Google results all along. By limiting the results per page it is harder for people to see how often Google's search results repeat, which according to my feedback users associate with low quality results.
I created such a search engine myself, but instead of displaying the standard text results I displayed only the website favicons so as you scrolled down infinite favicons appear, click a favicon it takes you to the webpage.
I think nickporter hit the nail on the head about the first page of results being so important and as you mentioned the competition to be on Google's 1st page is good for Google financially. However, one complaint about my search engine was that since the results often repeated the same favicon users thought the results were subpar, never realizing the results were Google results all along. By limiting the results per page it is harder for people to see how often Google's search results repeat, which according to my feedback users associate with low quality results.
The repetition obviously applies for your search engine that displays solely favicons, but I don't think that there are many duplicate search results.
Here is a Google image of my search engine (I used Bing API at the time) displaying the results for "Obama": http://www.google.com/imgres?q=tomorrowbook.com&um=1&...
As you will immediately notice FB, Twitter, Whitehouse.gov, barackobama.com are repeated 3-4 times each, and this is only the 1st 30 results you could continue scrolling down infinitely and you are bound to see the repeat again and again.
My point was user feedback complained of repeated results on my search engine. Users generally thought my results were lower quality than Bing/Google despite being identical depending on the API I was using. Further, I believe a lot of this has to do with Google/Bing users being unable to see a large number of results on the 1st page.
As you will immediately notice FB, Twitter, Whitehouse.gov, barackobama.com are repeated 3-4 times each, and this is only the 1st 30 results you could continue scrolling down infinitely and you are bound to see the repeat again and again.
My point was user feedback complained of repeated results on my search engine. Users generally thought my results were lower quality than Bing/Google despite being identical depending on the API I was using. Further, I believe a lot of this has to do with Google/Bing users being unable to see a large number of results on the 1st page.
I think his point was that the actual search results refer to a particular page on each site, but your's only show a reference to the site as a whole. Thus, where you show 4 facebook icons (and so the results appear to be repeated), the actual results are to four different facebook pages (probably of four different peoples accounts) and so aren't acutally repeats.
You are exactly correct that the favicon links to all subpages. Example the FB favicon will take you to fb.com or fb.com/obama, scrolling over the favicon will show you a preview of the result page and the url before clicking.
But even linking to different subpages I think Google would prefer users not see multiple FB, Wikipedia, ect... results and that is just one rationale I offer for Google having an interest in limiting results/page (the question of the original submission).
But even linking to different subpages I think Google would prefer users not see multiple FB, Wikipedia, ect... results and that is just one rationale I offer for Google having an interest in limiting results/page (the question of the original submission).
I don't particularly like "infinity scrolling", as you call it. I like it less than "infinity 'Next page' buttons" as I call them now. Why would you think it is "ridiculous" that Google not offer this? I think that's an interesting/polarizing/dramatic way of putting it. I'm curious as to why you clearly believe this to be a clearly superior/better way than what it's been for so long.
That being said, it's interesting that I actually like it for the image page. I can't explain it - maybe it's just a "Well, for 15+ years I've used search this way" thing.
That being said, it's interesting that I actually like it for the image page. I can't explain it - maybe it's just a "Well, for 15+ years I've used search this way" thing.
Check out http://duckduckgo.com
I use Google because the search results a much better.
However, in terms of UX, I personally feel that scrolling though an endless stream of results is much nicer than having them split up in groups of ten. I find it much nice to be uninterrupted when I'm trying to quickly scan through a ton of search results.
I use Google because the search results a much better.
However, in terms of UX, I personally feel that scrolling though an endless stream of results is much nicer than having them split up in groups of ten. I find it much nice to be uninterrupted when I'm trying to quickly scan through a ton of search results.
Right, that's not the first "infinity scrolling" thing I've seen so it falls in with all the rest. My favorite option, I think, is to be given a drop down/combo box that asks me how many I want to see (and is populated with 3-5 options). I like having that separation. I think it's a more efficient way also.
Let's say that you and I both search for a topic. The result we both want is result #33. We both find it, read it, and close the browser. The next day when we come back to the computer, we need to find it again. I think it's much easier to remember "That was at the top of the third page" than "I had to scroll maybe 5 or 6 times".
Let's say that you and I both search for a topic. The result we both want is result #33. We both find it, read it, and close the browser. The next day when we come back to the computer, we need to find it again. I think it's much easier to remember "That was at the top of the third page" than "I had to scroll maybe 5 or 6 times".
It's a confusing user experience (the scroll bar no longer indicates the length of the content and the sudden updates when you reach the bottom are jarring). It's also an unnecessary added complexity for hardly any gain.
Aside from anything else, I'm under the impression infinite scrolling doesn't work particularly well in Safari...
...I'd love someone to point out how out of date I am though, this is a year-old presumption! :)
...I'd love someone to point out how out of date I am though, this is a year-old presumption! :)
it's unnecessary. if google can't give you your desired results within the first 10 results, they've failed. you should just retry your query instead of checking the next 10 results.
also, infinite scrolling is more for "discovering" without a specific result in mind whereas "searching" implies you have a result in mind and is either hit or miss.
also, infinite scrolling is more for "discovering" without a specific result in mind whereas "searching" implies you have a result in mind and is either hit or miss.
I wouldn't exactly say they've failed. Maybe I failed in that I wasn't accurate enough with my search terms, but I've found the results I was looking for on other pages countless times.
Though it would be interesting to find out how many people do only check the first page before retrying their search.
Though it would be interesting to find out how many people do only check the first page before retrying their search.
"Only 52% of Google users click on an organic search result found on page one"
According to this research http://searchengineland.com/organic-click-thru-rates-tumblin...
According to this research http://searchengineland.com/organic-click-thru-rates-tumblin...
if google can't give you your desired results
within the first 10 results, they've failed.
Perhaps it came out negative in a/b testing. Another possibility is that it could lead to significantly higher capacity costs such that an a/b test might need to be hugely positive in order to offset the higher cost. Just imagine all the users who don't click to page 2 but do scroll to the bottom of page 1 in order to see why this might be the case.