What is the smallest thing you have done that has made a ranking difference?

[Mark_A]

I suppose for me, just playing with inclusion of some terms into a couple of pages has vastly improved our ranking for them..

What are your smallest changes - which lead to improved rankings?

[maximillianos]

We don't make any hasty changes when rankings shift. We tend to study the shifts over a few months at a time and then take action if necessary.

[austtr]

Linking out from my content to on-topic sites (authority or otherwise) had the most effect of any change I've ever done. It was a couple of years ago, but it's probably just as relevant today as it was back then.

I always shake my head in frustration when I get three way link requests from people who think that placing an outbound link on their "real" site is like giving away the key to a chastity belt.

And the sad thing is that these requests are usually from companies calling themselves SEO experts.

[Bentler]

Posing a query question on Facebook coincided with a page jumping from position #7 to #1.

[Planet13]

Posing a query question on Facebook...

Could you explain what this means to some of us who are facebook-challenged?

Thanks in advance.

[frank72]

@tedster: Well, the thread didnt say search engine ranking (SERP) so i thought to give another view of "ranking". In my case I added one space (in some cases two) before and after Google ads slots (<br style="clear:both;" /> ) and this increased my ctr. This made a difference and increased my earnings rankings.

This put me above a lot of my competitors which are stuck fine tuning their blogs to get the impossible... increase serp and get over millions of blog with the same topic.

[edited by: Robert_Charlton at 8:40 am (utc) on Dec 25, 2010]
[edit reason] fixed formatting [/edit]

[shigamoto]

A page on my site was referenced on Wikipedia, I did nothing but it made a huge ranking difference.

[Sgt_Kickaxe]

Having a domain thats 10+ years old.

It seems to work in tiers, I noticed gains after 5 years on one mostly static site with no new content in years. The site was a complete site offering guides but there just isn't more to write on the subject.

It also seems to be time registered with current owner, not just domain age.

[indyank]

It also seems to be time registered with current owner, not just domain age.

nice observation...echo the same :)

[httpwebwitch]

Incorporated a long list of local place names into the common page footer. Ranking improved imperceptibly for "big-city-name widgets", but jumped to #1 for dozens of "small-dinky-nearby-town widgets". Traffic from those longtailers within driving distance increased sales significantly for my client, who were willing to travel to big-city for their coveted services.

[NixRenewbie]

Making an old-fashioned hand-crafted site map page that included a snippet about the post being referenced.

The post being referenced was either a geo-located sales pitch that contained a short post about the location with our contact info and a link out to something informative (news report, census stats, landmark and so on) about the location. Else it was a product information page of a general nature. In that case it linked out to an authority site such as the EPA.

I was predictably hitting page one or two of the SERPs.(for a 4-term long tail: "keyword-one keyword-two city state" ... it wasn't magic, but it was putting us ahead of long-established businesses - right after paid listings - for the same terms. For a month old site, I think that's pretty good.)

[backdraft7]

Honestly, the smallest thing I've done is to simply be patient and wait for Google to sort out all their problems. On December 26, we experienced the biggest sales day all year, which was right on schedule with the past 10 years. This year I've learned that if a site converts well for the past 10 years, then in May 2010 suddenly stops, you can't blame the site.
99% of the problems we had this year were due to Google's "quality" changes.

[LostOne]

Avoiding

the easy way to the top

by traveling the road less traveled by the masses.

[techrealm]

By having the domain name in the DNS owners Record, just happens the domain name was also competitive keyword and the company and or owners name.

Content hasn't changed in years, no links to it and no ads otherwise. Value added, made life very comfortable...

[m0thman]

I'm not a "nothinger".. I'm a "somethinger".

Just write stuff, or publish some of the tons of stuff people send you and you're little acorn will grow into a big tree - literally!

Stagnant sites, or low content sites I think don't do so well, fresh new pages of interesting content do help a site get noticed... by visitors and visitors are the life blood of a website. Google I think is unique in understanding this relationship with their algorithms. (Please, no flaming from Bing, Yahoo and Facebook fans). Basically, it's about pleasing your visitors and providing them with the info and service they want. Check out Matt Cutt's blog and learn from it!

[mycrystalbridge]

article submission to ezine and other good ones with keywords anchor text has helped me to rank

[RP_Joe]

Tedster, is there a tool for finding Co-occurrence?

[aakk9999]

is there a tool for finding Co-occurrence?

For related words / synonims, apart from obvious words that may cross your mind, you can try any thesaurus, also google related searches and google wonder wheel.

[tedster]

Also note that co-occurrence goes far beyond synonyms. For example, "Bill Clinton" and "Monica Lewinsky" may well be co-occurring terms to a high degree of relevance, though not as high as a decade ago. Just so "cloud application" and "ajax refresh" or whatever.

There used to be a tool online, but it has long since changed. I prefer to do my own text-crunching anyway. This is not a function that I want to automate, and the public tools I tried in the past required us to BYOB (Bring Your Own Brain) to a large degree anyway.

So I don't pretend to make a pure science of finding co-occurring terms. I just look at what's ranking and take notes about how I might expand the text on my page, giving my copy full permission to stray from "keyword2, keyword2" rigidity.

[aakk9999]

LOL, I like the co-occurrence example.

I have just tried Google Wonderwheel for Bill Clinton, it does not have a direct connection from "Bill Clinton" to "Monica Levinsky", but the first node of "Bill Clinton scandal" then does show "Monica Lewinsky" connection.

[Robert Charlton]

I have just tried Google Wonderwheel for Bill Clinton, it does not have a direct connection from "Bill Clinton" to "Monica Levinsky"...

The Wonder Wheel is a great suggestion.

The way co-occurrence works in the context of this particular example, it's much more likely that a page about Monica Lewinsky will include a mention of Bill Clinton than it is that a page about Bill Clinton will mention Monica Lewinsky.

The Wonder Wheel behavior happens to be consistent with that expectation in this example. It does suggest Bill Clinton when you check it for a Monica Lewinsky search.

PS: I would be very careful to avoid overusing terms suggested by the Wonder Wheel or related searches.

[aakk9999]

PS: I would be very careful to avoid overusing terms suggested by the Wonder Wheel or related searches.

Agree with Robert - the bad idea would be to just add to the page content the first node phrases found on WonderWheel or found as related search phrases.

It has to be used more subtle - we often use wonderwheel to see in what direction 2nd, 3rd, 4th or even 5th node goes and in some cases you can get a complete "cycle" back to the original phrase (or back to "permuted" original phrase or its synonym). This can give an idea on how to expand body copy or even what new pages to add to the site with relevant anchors from the page being optimised.

[tedster]

the bad idea...

I know of one website that decided it would be clever to optimize a page for every phrase that shows up in "Related Searches" around a keyword where they already were ranking top five. All their established rankings dropped by three pages within the week.

[Planet13]

I know of one website that decided it would be clever to optimize a page for every phrase that shows up in "Related Searches" around a keyword where they already were ranking top five.

Do you mean that they took a single page that was already ranking, and tried to get that single page to also rank well for the phrases in the related searches, too?

Or did they try to create (and optimize) a NEW page for the phrases in the related searches and then linked from the original established pages?

[indyank]

Yes, that sounds like an over optimization penaltyfor trying to be too wonderful :)

Well, a good seo is one who identifies and does the optimal optimization and that is where he makes all the difference :)

[edited by: indyank at 4:30 am (utc) on Jan 17, 2011]

[Planet13]

PS: I would be very careful to avoid overusing terms suggested by the Wonder Wheel or related searches.

would you mind elaborating on that? For instance, could you give an example of what NOT to do?

Thanks in advance.

[tedster]

Those related searches were added to one already ranking page. They weren't trying to rank for all those related searches, it was their experiment in co-occurring phrases, which they incorrectly assumed were identical to Google's "related searches". However, I would think that smashing a large number of co-occurring phrases into one page could also trigger a penalty. The change would be a clear attempt at manipulating Google.

This happened a few years back. I think Google is even more astute at catching it today.

When I gave my example above, here's the process that unfolded. The page was at a directory level, one click from Home and on the main menu. I'd been trying to rank for the principle two word phrase for maybe two years, and it just wouldn't get off page four. Deep backlinks, internal linking tweaks, title element changes, H tags - all the usual SEO suspects.

I hadn't made any more attempts for several months and the time felt ripe for another experiment. So here's the way it went:

1. Study the top 30 results.

2. Note the common phrases of 2 to 5 words - this text analysis can be programmed, but you do end up with a lot of "phrases" that don't make sense.

3. Then the AH-HA moment: pages more deeply internal cover those topics too, but I wasn't using them on this top-level directory page.

So then I rewrote parts of the copy, and in three out of four cases I linked to those deeper pages with the co-occurring phrase in the anchor text. The URL jumped to page one, then began to climb. It's now been at #1 for almost a year with no more "fiddling".

One of the keys here, possibly THE key, was that the site had pages indexed that already covered that material. What my down-and-dirty co-occurrence study really did was spotlight something for me - the fact that I wasn't letting visitors know about those related areas.

I was letting the visitor know about many other areas, but apparently not the most important ones. Somehow when that ah-ha moment came, it was head-smacking obvious.

[Planet13]

However, I would think that smashing a large number of co-occurring phrases into one page could also trigger a penalty. The change would be a clear attempt at manipulating Google.

Do you have any rule of thumb (or gut feeling) when you really need to separate material into another page? I would imagine that wikipedia must have hundreds or thousands of individual pages that would have several co-occurring phrases per page. (Understanding that what google allows wikipedia to do may not be something that other sites are allowed to do.)

One of the keys here, possibly THE key, was that the site had pages indexed that already covered that material. What my down-and-dirty co-occurrence study really did was spotlight something for me - the fact that I wasn't letting visitors know about those related areas.

If one's site did not already have pages covering that content (i.e., pages of content covering the co-occurring phrases), would it be wise to link to external sources until internal content could be created, and then change the links to those internal pages?

Or would it be best to just wait until the internal content is created and then link to that internal content?

Maybe you would feel more comfortable putting it in a general time frame, such as, go ahead and link to external sources if it will take 3 months or longer to come up with your own content?

Thanks in advance.

[tedster]

Wish I had a formula - I don't. In this case my "gut" said four was enough. If the content had needed to be created, I might have even gone more gently. I wanted the copy on the page to essentially serve the same function as it always had.

Remember, my goal was to rank better just for a particular 2-word phrase, not all the co-occurring phrases. I was trying to find a clear signal that the page was relevant for my primary phrase.

I honestly felt that the algorithm was a bit defective, but that's not business. So I went after this particular factor which I thought was in the mix but I'd never intentionally tried to include -- and it seemed to work. From that point on, my web writing and that of anyone I supervise has become a LOT less restricted. And the writers love it, especially good writers with mostly print experience.

So this thread asks what is the "smallest thing you've ever done" and now my explanations made it sound like a big thing. It really was only less than an hour's work - but for a pong term project that had thousands of hours already invested.

----

Are there any other small actions that made a big difference out there? Come on folks, share!

[RP_Joe]

"What my down-and-dirty co-occurrence study really did was spotlight something for me - the fact that I wasn't letting visitors know about those related areas. "

This is a problem with sites that have lots of content. There may be 20 pages on topics they want to read but they don't know they are there.
The Wordpress categories is a step in the right direction but not a total solution.

Sometimes a site will have 300 pages or more but the visitor cannot tell the difference between it and a site with 50 pages. This is especially true of long narrow sites. Often there is not a way to show the visitor the breadth of the website.

[Prudence]

Blinked, then before I knew it, my site had lost 40 places for its main keyword.