Furture ranking features

Hi everyone
What kind of features do you think will be important to the algo's in the furture?
I'll start it off -
Site favouability as in how long and how many people click and stay on the site this is part of the algo at the mo but I reckon it'll be a lot bigger factor over the next 5 years.

Any others?

That won't happen and here's why:

1) The search engines don't have access to the data they would need to use this as a ranking factor. They do not know how much traffic a web page gets, as it is not publicly available, and thus cannot use it to determine its rank. (For those of you who want to say, "But there is Google Analytics", that service is used only by a small percentage of websites and unless every web site decided to use it on every web page the data is far too incomplete to be used this way).

2) It would be a self-fulfilling prophecy if the search engines used its own SERPs as a means of determining it's search engines results. Obviously the number one ranked page for a search is going to get more traffic than a page not on the first page. If traffic was the indicator of where a page belonged there would be little or no way for a page to ever move up simply based on the fact that the pages ranked higher would be receiving more traffic from the search engine based on the mere fact that they are ranked higher.

3) Traffic volume can be manipulated. Spammers and black hats could easily write bots to artificially inflate their page views and thus their rankings. Plus you can purchase traffic from traffic providers or buy expired domains and redirect them to your site. It would just be too easy to do. (I can also see newbies hitting refresh for hours on end....)

4) Traffic is not an indicator of quality content. It is only an indicator of good marketing.

Their algorithms won't change much. What they'll be spending their time doing, and have already begun, is weeding out spam and manipulation of the SERPs.

Yeah I do agree about how easy it would be to spam. redirects hidden frames so you click on one page and in SE eyes you are viewing 5 pages at once. Automated scripts accessing the page 24/7/365

What about the Google toolbar there is no reason as to why that couldn't be used and it is more widely used as everyone/a majority of people have it installed.

But then again how easy is it to spam links. They will look for other alternatives.

I'm not saying it would be the only factor. It is part of the algo now though. An online test chose a rare keyword with little to none direct competition they had 78 volunteers click through to the site and then 4 separate links around the site each day stayed logged on for 10 mins all had the Google toolbar installed. Within 2 weeks it jumped to the top of the results. After 4 weeks they ended the test and it fell back down again after 5weeks.

Bare in mind this wasn't a competitive keyword.

Secondly the pagerank from last year included rankings based on traffic flow.

Click here for the proof

The link above is from the science of page rank at amazon check out 12.3. right at the top, 3 pages to the right.

Googles going to look for alternatives in the long run. Don't get me wrong links will always play a factor but I don't think they will be as major as they are today.

Look at on page factors now and in 2003.
edit:Ricips. now and 2004

Any other ideas?

What about trust rank?
Hilltop?
Topic-Sensitive PageRank?
temporal analysis?
and latent semantic indexing (LSI)?

Surly they will play more of an issue?
Come on people get creative but keep it topic related.

Good question! I think the engines will become cleverer at interpreting the meaning of queries, for example if someone in North London searches for Spurs kit they are probably looking for Tottenham Hotspur football kit, whereas someone in the mid-west USA is probably looking for cowboy clobber. Location and context based interpretation so that the same search could provide different results to two different people.

Yeah totally I mean all a Search engine is is an information collection and sorting tool. They strive for relevance so yeah spot on. I seen a suvey quiete recently that says that 33% of searchs at Google require 2 or more searchs to find the info from the first Search

Actually Ask.com has been doing this for years.

I do like ask but the problem with it is that every time I use it I only get what I want around 20% of the time. The UK TV ads are appalling but thats beside the point.

Ask is good but I just don't like the results it returns they are just irrelevant most of the time.

I typed in a terms the other day really specific but the results it returned was just like what? Went over to Google and found what I wanted straight away.

I almost always go to Google Yahoo or Live as a result.

It's a shame as I like Ask, overall it has some good ideas it's a shame it doesn't have a larger index.

Since web is getting more and more social I think it will be important whta kind of human experience offers any web site. Not just in real time communication but also as a reference resource

Some time back, links from wiki, .edu and .gov were considered highly valuable in search engines. But now it has been changed, because most of the wiki, edu sites and even gov sites are being targeted by spammers.
I have observed that Google is showing different search results based on locations, even if you are using .com search. That's quite interesting to see.

Yeah totally it also indirectly brings in traffic as a rating algo. As in if the site has lots to offer then the site will generate more traffic. higher traffic=positive change in the algo. positive change means higher rankings.

I read a theory somewhere, cant think where as I read so much , what if the SE struck a deal with publishers to bring offline content online? Whole books/documents that sort of thing but on a massive scale.

Before you dismiss the idea think of the possible upsides.
1. More relevant content. This is exactly what the SE strives for.
2. Spam sites will little content would drop like a stone as the bar would be higher to compete as of the influx of new unquie content on a huge scale.

Possible Downside
1. Sites that have adsense might not compete so they SE's would lose money?

Or would PPC just increase heavily giving them more money? Interesting idea. Lose adsense to increase revenue by making more people send money on PPC to compete with more relevant content. Just an idea.

What about better detection of spam sites?
Harsher penalties for breaking the rules?

The thing about wiki is that wiki now uses nofollow on all external links have since Oct. last year so this might be while links have been devalued.

Thing is I don't, I know people will disagree, think .ed.org got higher results due to those domains it was the reason behind them.

.gov and .org both tend to be
1.old
2.trusted, not the domain name, but they tend not to use balckhat.
3.Massive PR that some have.
4.authorutiy of some of the sites.
5.If they linked out they where related.
6.Have good content.
7.Massive backlinks.

Yeah I have to admit that has become a lot better over the last year. Locality factor will play a bigger result i rekon tieing in with what dominoman was saying.