One of my favorite Canadians in search (there's so many, don't make me choose!), Andrew Goodman, pointed me to a post by Jamie Roche of iMedia Connection - A ReDesign Worthy of Google De-listing. Mr Roche (who works with the supafly Jon Mendez if I'm not mistaken) had some good advice, sadly shrouded in long-past SEO mythology that deserves clearing up. Roche says there is "common knowledge" in the SEO sphere espoused by "an endless number of third parties," by which, I assume he means us. This includes:
- That the appearance and placement of key words or phrases on your page can increase ranking
- That the number and quality of sites that link to your site, especially when the link includes your key word or phrase, might be the most important factor
- That there is special magic that SEO firms know, including submitting the site or pages to indexes, as well as setting up meta-tags, image tags and other hidden stuff on the page to get the spiders to connect your page with your key words or phrases
- And perhaps the most ubiquitous belief is that, if your page ranks well, don't mess with it.
Say what?
Who are these mystical parties feeding you such obvious crud. I grant that numbers 1 and 2 are generally valid, accurate and should be presented by any SEO worth his or her salt. However, Mr. Roche, you've got a little explaining to do about #3.
The "special magic" that we know is similar to the "special magic" that you, as a conversion rate expert, know. Let me give you some examples:
-
Branding in the search results can bring returns in CTR
-
There are site architecture issues that can cause search ranking and inclusion problems
-
It's wise to follow the rules when crafting URLs
-
If you're not careful, the engines will ruin your smartly crafted copy
-
Search engines are very smart about interpreting links
-
We can improve your blog traffic
-
Lunar cycles and ocean currents have something to do with SEO afterall
-
When I run into tough questions, folks smarter than I have the answers
Needless to say, the hidden tags and top secret submissions are a thing of the far past - remember when conversion consulting was all about how to sell more chickens at the farmer's market? Same thing here.
The myth I most want to tackle, though, is this one - "if your page ranks well, don't mess with it." I'm fearful that Mr. Roche may actually be correct here - there seems to be a contigent of SEOs and clients of SEOs (or maybe just clients of mine) who believe that once a page is ranking, it's sacreligious to mess with it. This is a complete falsehood.
Today's search engine rankings are primarily centered around a few elements:
-
Strength of the domain (measured by links, trust, age, and possibly some usage factors)
-
Internal and external links to the page
-
Keyword in the title tag
Those are the big three. If you're hitting on all cylinders with these, you're 80-90% of the way to being as optimized as you can be. Thus, the changes a conversion rate expert might make, which are nearly universally on-page changes - where content goes, what headlines say, how graphics display, how embedded content or animations or javascript might function, are all 100% kosher to test to your heart's delight. We've done it dozens of times, and generally, when you increase conversion rates and make a better page for users, you gain rankings. The only place that conversion rate consultants may venture that could potentially hurt you is if they recommend that you make your content un-indexable (i.e. put in in Flash) or change your title tag (so that you're no longer targeting the keywords that bring you traffic. Barring those mistakes, I highly endorse working with successful conversion booster guys like Jamie & Jon.
Let's use a fun illustration to see what our friendly Google monster thinks of the situation:
Instead of worrying about preserving your rankings by never touching your "optimized" page, you should be attempting to improve that page as much as possible, to get the most you can out of your search visitors (for those who remember, this was my most overlooked strategy in SEO). You can hire a firm like Jamie's, or go out on your own and starting testing placement of conversion funnels, images, buttons, new text, headlines, the works!
I'm not suggesting that you can be reckless with changes to a page, but in general, we've seen complete overhauls of text content, images, embedded items and headlines without any trouble in the rankings. Preserve title tag targeting, link structure, some instances of your keyword on the page and general accessibility and you'll be safe from harm. What a happy note to sleep on.
p.s. No offense to Jamie in here; I'm actually very appreciative for the blog topic.
If your page ranks well don't mess with it???
Isn't page updates, and fresh content in general, a plus for rankings? Don't many blogs rank well because of this very important factor?
GoogleBot shouldn't look that confused about this subject...
I have got to agree with Rand on this one...
1. The BBC may never change the content on pages, but Wikipedia, Technorati, and every other 2.0 site out there that is beating your's for ranking for targeted keywords do. And they change almost daily. The content on the page, as long as it maintains relevance, is not going to affect the primary keyword being targeted.
2. That being said, shifting the content will affect secondary and tertiary terms. So, for example, if a client shows me a page that is already ranking for a competitive term, we start running through the options of non-competitive, well-searched, stemmed terms. So a page that already ranks for home mortgages should be regularly updated with non-competitive (often buzz) terms that will gain immediate rankings.
3. Changing and improving the content of the page encourages users to revisit the page. If you don't give them a reason to link to you the first time, maybe you will the second.
4. This is just speculation on my part, but Google NEEDS fresh content ever since considering historical data. In my honest opinion, the best, most valued content is that which is both trusted historically and regularly updated. PS: thanks for the shout out Rand. If we go by traffic ratings, I owe you probably 100 shout outs in exchange. :)
I think it comes down to comfort level.... the site finally gets to a point, so people get afraid of making any changes for fear of dropping back down.
Of course you can always copy your existing page and make tweaks to test. This could also be good for live visitors as it gives them a sense of freshness when they return to the site, even if the overall message is still essentially the same.
And with more and more sites running with a CMS, it would seem quite easy to make 3-4 page versions and publish/unpublish as necessary.
But it all comes down to comfort zone and risk. They may not have even hit number one yet. It becomes link a gameshow... do you go for the jackpot or do you cash out now and be happy where you are?
But unlike the gameshow, unless you do something completely stupid, you aren't ever out of the game, you get to keep trying for the jackpot.
Wow, talk about an outdated review of search marketing standards.
There are some who really don't mess with pages once they rank and have decent exposure for it and dont even continue to try and target high quality links?
As an optimizer, I am always trying to make things BETTER, including pages that may appear to be perfect, however I don't mess with too much at one time, or change anything I know won't make a difference.
I think it's wise to tweak content, titles and heading tags from time to time to keep up. After all, if a page stays the same forever, it's bound to fall off sometime (even if it doesnt fall from the first page, it may move down).
When reading that, I seriously thought it was something from the late 90s, not mid-late 00's.
In regards to number 3, Mr. Roche obviously hasn't read SEO's Collective Bargainning Agreement that doesn't grant us our official special magic until 2009, but I'm sure he was referring to that Black Special Magic Hat.
I'm actually doing this moving company right now. I just now put tracking on it (I'm ultra paranoid in not rocking the boat on this website, so I put it on a test server with "nofollow" commands on each of the pages, then put it on the real website server when I'm satisfied that I haven't ruined the code.) I'm going to let it track for a month before I make any major sales copy overhaul changes to it (more questions? maybe, definitely a list or so, more internal anchor links).
I like coming here, because being new to this, I'm glad that when go out on a limb that what I am doing is the right thing for my client. I'll admit, I did submit the website, but other than that, I've followed the instructions from this website.
Have to admit that I'm one of the persons who didn't mess with onpage factors when it hits top 5-10.
IMO the whole fresh content loving factor of our new robot friend is measured on a site level and not a page level, therefore not making it neccesary to update pages themselves, only to add pages to the site (which I would like to debate as well...).
But I have been so scared to loose rankings that I haven't touched some of my sites, while I know the conversion rate could be better. This article helps me push into the direction to finally make some changes, document them and analyze the results.
Thanks for another awesome post Rand, you really are on a roll lately!
I think the problem basically comes down to, people are afraid. Especially if they don't think the site/page should actually be ranking there. They take the view of, "If it ain't broke, don't fix it".
In any other field, this would be true, but I agree that here it just isn't.
For one, we're not in a static landscape. Other people may well be competing for that term. The algorithms may change. New filters may be put in, or old ones removed... You shouldn't be resting on your laurels.
Secondly, if you're winning a race, you don't say "Well, I'm one step ahead, that's it, I'll relax now". You keep going for it and build up as much of a lead as you can. Keep tinkering, refining and improving. Get as much of a lead over the other people in the SERP as you can. Don't just beat them, annihilate them.
And finally, tinker to improve conversions. Sure, it's nice to rank well, but I'd rather be pulling one tenth of the traffic, but getting a hundred times the conversions, by focusing more time on design, usability and sales process flow, rather than on rankings. In the end, what we're after is more sales. Sure, more traffic will tend to mean more sales, but so will a better tuned, more highly refined site. And that's something that's completely change-proof.
I don't necessarily agree with that statement....In business you always want to stay ONE step ahead of the competition. No need to be leaps and bounds ahead - you need to save that for later.
For example, Apple could have initially dropped a crazy widescreen/touchscreen bluetooth/wi-fi enabled ipod running on flash memory that would have destroyed the competition. But all they had to do was release a product that was a little better at being an mp3 player than other brands, then use their branding and marketing to pull them through. Now they've kept something up their sleeve and when Creative or Microsoft release a new product Apple still has a play to make.
I approach my SEO with the same philosophies.
I would have to agree with you on date modification. Date modification has to be one of the many reasons that wikipedia pages come up on the first page of some of the broadest keyword searches. Also, check the source on www.kaiserpermanente.org. Whoever does their SEO, puts the date modification in the first line of the code.
one more thing: i like the graphics in this blog post, especially the named pose of the google 'bot
I have a friend who makes tons of Adsense cash from some MP3 sites. He tells me they constantly re-evaluate keywords for every page. And they have thousands of pages.
They have people dedicated to connecting traditional media news with keyword research to remain highly ranked for terms that are yet to be tracked by services like Wordtracker.
His entire success is based on continuously optimised keywords and content changes based upon those keywords to get the search traffic before anyone else does.
With eCommerce you always have to be testing the page. If it ranks well, and you aren't converting, well what's the point really when you don't count eyeballs, you count boxes on the shipping dock?
I've recently become a CrazyEgg addict and have been testing religiously, moving stuff around each week to see what works and what doesn't.
I have also seen evidence of freshness being a huge factor in our traffic. I keep our home page updated pretty much daily with content, and can see traffic pointing to the new content within 4-5 days.
I agree 100% that you should be working to improve your content for conversions, seo, usability, etc regardless of how well a page is currently performing.
I have to question though whether making change solely for the sake of change is a good idea or gets an automatic freshness boost. Assuming that making an update gives a ranking boost then if you were to go in and change all 'the's on a page to 'an's you would see an improvement. I have a hard time believing that kind of change would make any difference.
If you're looking at a news search or a blog search where freshness makes sense then I agree you probably will see a boost for having newer content, but I'm not sure that would be the case with general search.
I admit I could be completely wrong about this. I haven't tested this at all, but again it's hard for me to reconcile change for change's sake as being an automatic improvement,
If you were to write an article today about Alexander the Great why would that article be any less relevant tomorrow because you don't make a change to it.
I think frequently updating will get bots to visit your page more frequently, but visiting your page more is not guarantee of improved ranking.
If someone has some real proof that frequent updates are automatically an improvement I'd be happy to see them and again I admit that my thoughts here are just me speculating. I'm have a hard time convincing myself that change is always an improvement.
There's no better way to jolt a funny analogy out of a guy than to get his dander up, it seems. Chickens at the farmer's market!? Gotta love it.
Huffy Googlebot is also a gem.
Stripping out all the pride of craft though, this debate can be boiled down to a sensible equation - a dilemma that any of us commonly face.
1. Let's say you really want to work on a page - put it through some sophisticated testing or whatever to maximize its conversion rate (even if it is not the only page in the user's visit or even the most important one - but let's say it gets a lot of visitors). OK, so let's say that page is the home page.
2. So potential scenarios include:
(a) Run your test, have SE ranks & referrals stay the same or improve; and wind up with a higher conversion rate and higher profit;
(b) Run the test but have the homepage change enough that you drop from #2 to #6 on search engine rankings on a range of core terms; wind up with a higher conversion rate; but the end result is about a wash;
(c) Run the test and for some reason (this seems unlikely, but it's a risk) the drop in search ranking is more severe, and traffic drops enormously; so your conversion rate increase isn't even measurable, or that useful; so you lose bucks;
(d) Run the test, don't get much of a conversion improvement, and also lose rank. Lose a bunch of bucks.
Mr. Roche is betting on (a) and (b), and telling us that he will continue to do so. He is warning people that (c) and (d) are in the realm of possibility. The extraneous comments about the state of the art in SEO are just that, extraneous, and don't affect the above scenarios or the probability of improving profit by conducting a good test, which I think is high but with some risk.
Great post, Rand.
Now to only convince some clients of this....usually the non-tech savy clients are horrified when you suggest changes to a page that ranks well.
Just show them "Huffy Googlebot" - that should do the trick.
That will do the trick for sure :)
I've heard rumours in the past that SE spiders look at the 'modified date' within the page header. The theory goes as follows: As we all know, the engines like to see fresh, updated information on websites. Thus, if they don't see this 'modified date' change over time, they start to believe that the content is going stale.
While not white hat, I know of one SEO company in the UK who actively use server side hacks to auto update this date when spiders visit, making it look like it's been modified since the last time the spider visited and they rank in the top 10 for 'search engine optimisation'. I'll refrain from naming said company just in case I offend anyone!
Now, whether this is actually true or not, I don't know. If it is, then it makes sense to try and continually optimise and improve the strength of the page (Note to self : This does not mean tinkering by simply re-saving the file)
Pretty sure Google can tell if the page has altered without looking at the headers.
Also do you ever see sites like the BBC altering stories after they have been published? They rank due to trust, not freshness.
I would have to agree with you on date modification. Date modification has to be one of the many reasons that wikipedia pages come up on the first page of some of the broadest keyword searches. Also, check the source on www.kaiserpermanente.org. Whoever does their SEO, puts the date modification in the first line of the code.
Our focus is mainly e-commerce of traditional products where in some cases you could buy the exact same thing over 100 years ago. Despite this we (try to) regularly go back and edit/make improvements to the pages even if there is nothing really "new" to say.
I don't have data for this but am convinced that it helps maintain or improve our rankings.
Not meaning to sound uppity, but that's like saying "This boat is floating because it's made of wood, not metal"
Metal boats float too. It's just something else is going on that you don't know about.
I suspect your changes have nothing to do with your rankings. Care to give us a domain name and some keywords to have a look?
I'm not saying our rankings are based on just making regular changes. I do however have the feeling that when we neglect parts of our site the rankings "wither" over time. I haven't done any tests to see if that's the case.
Testing this could have some interesting challenges. Do your rankings drop due to leaving things the same (no activity/neglect) or because someone else is doing better work to rank? A comparison might be difficult if for one test keyword you have an active competitor and for the other you don't.
I think if you're making changes to improve the page it's fine, but if you're just making changes for the sake of change keep in mind that any change can have a negative impact as much s it could a positive impact.
Updating the page so it always looks new will probably get googlebot to visit more regularly, but visiting more frequently is different than ranking.
I definitely agree here. When Googlebot comes to the site and sees a change, even a small one, it makes it look like there is work being done to the page. It isn't just some stagnant file sitting on a dormant domain.
If you don't 'tinker' to try and keep it fresh, you can bet your competitors will, and eventually they'll get a jump on you...
I should start by saying that Jon Mendez warned me not to write this article. As usual, I should have listened to his sage advice.
I am a believer in the skills of the search marketing experts. It is not easy to stay ahead of changes and confilicting information to deliver results to clients, and you do this every day, so my hat is off to you.
My point in number three was only to express, from a typical marketer's perspective, that natural search optimazation is seen as a complex, even esoteric art.
Also, I like Huffy Googlebot. It would make a good t-shirt.
Best,
Jamie
Jamie - I'm glad you wrote the piece, it's a good look into the value of conversion rate improvements and provided some much needed prodding for the SEO industry to start dispelling the myth of "don't touch my perfectly ranking page."
Jamie,
Thanks for posting a comment here. When I read your piece I had the same reaction Rand had "Say what?" I even thought that this had to be some other Jamie Roche and went to your bio page to check if it was the same person :). I'm glad you cleared up what you meant by #3. And I'm not sure how many people have the belief "that, if your page ranks well, don't mess with it," since the SERPs are not static and you are constantly being forced to compete against smart spammers and new competition on a daily basis.
Aw, my hat is off to you, Jamie, for reacting so well. I'm afraid to ever write anything public about SEO because Rand would probably find something wrong with it. I just direct everyone to SEOmoz.
Anyway, there are definitely things that SEOs know that other people don't, and a few years ago the things you mention in #3 probably did sum them up.
Now, those things have come and gone and there is new info to be found, if you're willing to take the time. I wouldn't say that it's top secret stuff, it just requires some diligence and study.
I must say that in all the SEO blogs and posts I've read, I've never heard anyone say to leave a well-ranking page alone. I've heard the opposite - that for any page to have some vitality, there needs to be "purposeful buzz of activity". This keeps the bots and the humans coming back again and again.