• http://www.seo-theory.com/ Michael Martinez

    Matt’s comment was directed at one specific person. Why do you continue to portray it as if he were speaking to everyone at large?

  • http://www.rustybrick.com/barry Barry Schwartz

    I have my reasons…

  • http://twitter.com/Stu_Draper Stuart Draper

    Oh come on. Anyone could get anything to rank for that term as that bogus string of letters probably doesn’t even exist more than 100 times over the entire web.

    NOTHING scientific about that test and it proves nothing.

  • http://taylortoussaint.blogspot.com/ Taylor Toussaint

    I believe that frequency of press releases for a particular domain and the content contained within each press release are very important in terms of how influential they are to a particular domain’s ranking.

  • http://searchengineland.com/ Danny Sullivan

    Matt’s made these comments about press releases for a long time, to various people and groups.

  • http://searchengineland.com/ Danny Sullivan

    It proves that with a few links, you can rank something or an unusual terms. Agreed, that’s not necessarily new. But it comes when there’s renewed attention on press release links and when Matt’s freshly said they don’t count. Clearly, they do.

  • Guest

    I believe the larger point that Matt was trying to make is that you can’t count on press releases to boost your rankings. You could be at position #1 and an hour later, be in position #30 for that same term as newer press releases with that keyword term are released. Sure, you can stick around for days or months optimized around “mumbledy frigosure parnisciplization” but for any real terms people are trying to rank for, a press release is good for a day of decent rankings, at most.

  • http://www.facebook.com/david.rekuc David Rekuc

    The test doesn’t do anything to prove how much value the links provide, but it does prove that they are in some way considered by Google’s algorithm. This isn’t an article claiming that you should use press releases as a form of SEO, its just simply stating: press releases like this are clearly factored into Google’s algorithm in some way.

    If we continue to track the site’s rank, we can also lend some insight to the time decay of these kinds of links. Provided the term doesn’t become polluted by publicity. Even if it is, it wouldn’t be difficult to re-create the test.

  • Chris Koszo

    Maybe just co-citation? There was mention of “Cutts” and “SEO”, and a link to boot.

  • http://www.rushlocal.com/ Rush Local

    leasreepressmm is a keyword so insignificant that this blog article dominates rankings right now for it when I do a search on my phone, however, yes, Cutts is still hangin’ in there. I see what you’re saying, I guess my largest concern is that there is no screen shot or mention about a test prior to posting these press releases. So at this point, Cutts could have been ranking for it before. Even if you did check it, without evidence that he wasn’t ranking before, this whole effort is FUBAR.

  • http://www.williamrufino.com.br/ William Rufino

    True thing.

  • clasione

    This proved you can focus too much on what matt says about every little thing.

  • http://searchengineland.com/ Danny Sullivan

    He wasn’t. I checked on this when the initial press release went up. Matt wasn’t ranking for this term. It only happened after the test began.

  • http://twitter.com/ItsHogg Jon

    To those saying the keyword is junk and therefore the test is invalid don’t seem to understand. A nonsense term was used to avoid the noise you’d get from using a competitive “real” term like “cheap phones”. The fact Cutts’ blog ranks is because the anchor text has passed from a PR to his site, therefore press releases have power to some extent.

  • prunderground

    Thanks Danny for clearing this up. Way too many SEO people without proper knowledge.

  • http://twitter.com/Aaron_Kocourek Aaron Kocourek

    We’ve conducted a LOT of test in regards to ‘do links in PRs help your ranking or not’ in a few of the tests we put out a group of PRs through all major services (it was unbelievably expensive). The end results was simply this: Yes, these links do help increase your search engine rankings, in a few of our large scale tests it made serious increases in rankings. BUT, the problem is that these links quickly disappeared cause most PRs are removed within 30-90 days. So the results were only temporary. When comparing the costs and the links that actually stuck showed us that the effort was NOT worth it. Focus on only putting out a PR when you have real news that will generate real PR, not just a garbage PR a week.

    No short cuts in link building anymore unless you only want short term results.

  • prunderground

    There are too many people who are not experts giving their opinions without testing the facts. I am the editor-in-chief of a social media release distribution company. I get many clients telling me they are being told by so called SEO experts, releases carry no SEO weight and are a waste of time. I constantly tell them if you use a keyword you wish to rank for along with a link, it Will Work!!

    I even have a section on our resource guide that talks about this. Here try this: I have a client who wanted to rank on Google News for the term “Urgent Care Center” do a search on Google and they appear as a Universal search result. Click on News and search it there and they are on top of Google News. Of course then we get the doubters who are relatively new to SEO (8 years or less experience) They will say: ” Oh come on that was posted today, of course you will appear” To that I say try this term on Google News: “Mobility Aids”

    I have not changed my SEO strategy much. newcomers talk so much about content is now king and content marketing. Content has always been king, content marketing has always been done and press releases have always worked.

    P.S. Careful with people advice without research. A press release for a hard to rank keyword can stick around for longer than just a few days even with stiff competition

  • RyanMJones

    Pretty much every theory in information retrieval talks about varying weights based on the size of the corpus. This is probably what happens with made up terms. Since there aren’t many pages with that term, the weights of the ranking factors change to better apply to the set of data.

    Example: If every page in the results set contains the same phrase, that phrase is treated differently than ones that only occur in some. It’s called tf-idf weighting if you want to get technical (http://www-nlp.stanford.edu/IR-book/html/htmledition/tf-idf-weighting-1.html)

    Anyway, I can see Google applying this type of scoring logic and theory to anchor text, meta data, etc. Thus, for made up words where there aren’t any other pages, the weights and factors won’t be the same as for other areas where there are lots of results.

    Make sense?

  • http://twitter.com/forefront1 Todd McDonald

    Why would Google place any true link value on most links achieved through press releases? The links end up on duplicate content pages with exact match anchor text on deep pages of sites that may or may not have much PageRank at all in them. That sounds exactly like what Penguin and Panda are about knocking out.

    Also, the example you gave of ranking in Google news is not great as the Google News results are built using different algorithmic ranking factors from standard results.

    A strong link from a legitimate site that writes a unique article about the news as a result of a press release can have tremendous value. It’s quite like that they standard $200 crap release will do much of anything over any significant length of time. It’s much more likely the boost would be temporary and comparable to a social signal.

  • RyanMJones

    I like to think of Google as more of a neural network type algorithm (especially panda and penguin) rather than a fixed formula. It just seems to fit more with the things we see.

  • http://twitter.com/forefront1 Todd McDonald


  • http://twitter.com/zachgriffith Zach Griffith

    If I were Matt, I’d feel slightly awkward right about now…

  • http://www.312digital.com/ Sean McGinnis

    I don’t understand the issue. What Matt said is not what was tested in this test, nor in the previous test.

    Take a look at exactly what Matt said:

    “Note: I wouldn’t expect links from press release web sites to benefit your rankings, however.”

    “But the actual content of the press release itself doesn’t directly affect a site. For example, on http://www.prweb.com/releases/2005/10/prweb296086.php those hyperlinks don’t help avatarfinancial.com (in Google).”

    “Matt clarified that the links in the press releases themselves don’t count for PageRank value”

    What I pull out of those statements is that links from the PR distribution sites themselves do not add value. That’s a VERY NARROW statement. Nowhere does he say that links from press releases republished whole cloth will not count.

    What was tested (and proven) in both these tests is that the overall value of a press release distribution strategy has value. When those original releases are republished by a media outlet, there is value there.

    In order to test the meat of what Matt said (according to my reading), one would have to publish a release on a press release site (say PR Web) and then have that release not get picked up or republished by any other publisher.

  • prunderground

    Danny do you think appearing on Google news really has any SEO value?

  • Guest

    You could rank for that with zero links.

  • http://www.facebook.com/mikekalil Mike Kalil

    “Editor-in-Chief at Social Media Release Distribution Company.”

    You must be an expert. :)

    Who is searching Google News for “urgent care center?” Your client is wasting money.

  • Joanna McDonald

    Do you have links to any blog posts or case studies from these tests? I am really curious to check out your results.

  • Graham Ware

    I agree 100%

  • prunderground

    Anyone who calls themselves an expert is full of it. I am no expert,. We are always learning even after 20+ years of SEO. You are definitely not a parent with a small child living in a low income area. I suggest you do your homework before criticizing. They actually get a lot of traffic with this search term. Please respect other peoples opinion and their work. there is no need to be sarcastic…

  • http://twitter.com/erocketSEO Dave Fowler

    Sean, agreed 100%. I’ve flagged this apparent misreading a few times since it was first reported, but still it persists. Maybe I’m missing something, but I’m sure Matt Cutts said links from press release websites don’t pass PageRank, not links in press releases per se. Therefore, if a press release gets picked up and published on a trusted site, it could still pass some value, IMHO.

  • David Newman

    I use a number of news wires to help increase the anchor text relevancy of a landing page. I find that new wire links are a great way to establish a rank for a desired keyword, however to sustain a high rank and actually pull some traffic I generally need to deploy article links from highly recognisable thematic sites. All good fun and we’re getting some great results. David Newman – Mercedes-Benz Hertfordshire L & L Automotive

  • http://www.facebook.com/mikekalil Mike Kalil

    I’m sure that search has a lot of volume. I just doubt that much of it comes from Google News searches.

  • http://searchengineland.com/ Danny Sullivan

    You miss the point. The page itself is ranking because of the anchor text in a completely different page pointing at it.

  • http://searchengineland.com/ Danny Sullivan

    This test involved a press release, from a press release service. The link used anchor text pointing at Matt’s blog. The blog started ranking for those terms. That’s an obvious ranking benefit, and from a press release. And if the page is ranking, it’s gaining PageRank from that link.

  • http://www.facebook.com/sixfiguremarketer Gary Huynh

    Just 1 day after this post, I checked the Google results for the term “leasreepressmm” and not surprisingly, this exact post is the #1 result. Matt Cutt’s site is now on page 2. Press releases can rank for very low competition terms because press release sites are authority sites. IMO, the best way to gain the most benefit from a press release is to target trending terms when you create the press release. These terms will gain a lot of searches quickly and face little to no competition so you can rank for a few days to a week and gain the most traffic before the number of searches diminish. You don’t need to focus on just press release sites for this approach as there are Web 2.0 sites which have better ranking power.

  • http://twitter.com/jaspalkalsi Jaspal Kalsi

    Definitely you can rank low or no competition keywords using other methods. But the fact that Matt’s blog ranks due to the link from a Press Release is good to indicate that press release links still do work … not that site owners should jump for it. I think we need to remember that G looks for specific signals and how much of what activity raises a flag is always going to be a mystery.

  • Abdul

    May be its just for Matt Cutts not for Us ??

  • David Newman

    I think it does for both co-citation and anchor text relevancy. I sent a press release on the Mercedes CLA release date, it had some major coverage in the google and yahoo news feeds, consequently i’ve seen some super top 3 ranks for related keywords…. As mentioned in my previous post, I then do some proper link building by deploying article links from thematic sites…. All good fun! :-)

  • http://seohour.com/ Nandita B

    Thanks Danny,

    Let’s see the sustainability of the ranking, before coming to a firm decision.

  • http://twitter.com/AlbertitoCarran Alberto Carranza

    Yes, Danny, but only works if you have an authority domain Like Matt :). I recommend to try this test in a crappy domain also

  • Grand

    It always makes me fun that people are still listening to what Matt says and even more fun is they believe, instead of testing.

  • http://www.seopros.org Terry Van Horne

    Depends what you call benefit… David Harry and I have seen the same thing, on more competitive SERPs only thing is it lasts a few days… this “test” (using that word in its loosest form) means something if you can search in say 6 months from now and that is still ranking. I would not be surprised if it’s gone in a few days.

  • http://www.seo-theory.com/ Michael Martinez

    Yes, Matt has also made other comments about not interlinking Websites, not reciprocating links, and not buying links — all of which have been shown to “work” and even to have limits that Google accepts. To take one of Matt’s comments completely out of context and misrepresent it as a generalization is irresponsible. You know most people won’t go read the thread and that they’ll take your article at face value. This is playing dirty. Why not share your reasons for doing so, or is transparency only a 1-way street in this discussion?

  • prunderground

    Well said David. I agree 100%

  • http://www.seo-theory.com/ Michael Martinez

    “Tests” using nonense terms NEVER prove anything. The search algorithms are not that simplistic. Matt has said “don’t expect press release links to help with your rankings”. The test cited above doesn’t address that point in the least. SEO testing generally sucks for two reasons: people don’t take in the complexity of the search algorithms and they don’t know how to control for multiple factors. Any test that uses a nonsense term like “leasreepressmm” fails on both counts because it creates an unnatural SERP.

  • Ved Tiwari

    Michael Martinez

    For sure this is neither a generic nor competitive term (leasreepressmm) has been used in Test.

    But its for sure that due to anchor text in press release, Matt blog appeared in search results.

    there could be not 100 websites having this word on Web, still why Matt
    Cutts blog appeared on #7 by keeping in mind that Matt blog does have
    any word like “leasreepressmm” on Blog at all.

    So signals are there that Google uses press releases links for ranking factor not that much but Do for sure…

  • http://www.seo-theory.com/ Michael Martinez

    First of all, Matt did NOT say that press release links won’t pass anchor text. So for Danny and Barry to carry on as if he did is simply outrageous. He DID say that the original poster in the Google Web forum discussion (and perhaps by extension everyone) should not expect the links to help with rankings. There is a huge flaw in the assumption that “passing anchor text” = “helps with rankings”. A LOT of other factors are taken into consideration and Danny Sullivan of all people knows that very well — who has made that point more often than him (outside the search engines)?

    This isn’t newsworthy. Matt basically said the same thing on his blog in 2006 when writing in response to a magazine article that implied press release links were responsible for the successful of a financial company. He didn’t say press release links won’t pass anchor text for anyone under any circumstances. It’s hard to find any other useful references since everyone has been repeating the nonsense that followed upon Matt’s comment in December and now this inexcusable test.

    If you want to (dis)prove something, first start with a reasonable hypothesis. You cannot disprove vague, ambiguous language unless you first show that it has a precise and explicit meaning. No one has done that.

  • http://twitter.com/optimizethis Keith Brown

    The competition for this term is non-existent. At the point where Google is only returning 8 results for a given query…that’s a stretch…

  • http://twitter.com/LauraGrace42 Laura Grace

    There is no algorithm out there identifying what is and isn’t a press release! Press releases most likely follow duplicate content rules. It isn’t that no links are counted from a press release; it is more accurate to say that no more than 1 back link will be counted for a press release. Most times this will be your own site, although if your release get picked up by the powers that be (and you aren’t tagging yourself as the publisher) than your press release could result in 1 link from another site.

    Making up a word means you have created a completely new keyword, which is not a normal situation. With a made up word, it is not unrealistic to think that a single strong back link could get Matt’s blog to rank for it.

  • http://twitter.com/LauraGrace42 Laura Grace

    Or it proves with a single link you could rank. I am convinced that when Matt says press releases don’t count he means that don’t count more than once. I don’t think there is any deception here. If he said “we treat press releases like duplicate content” the internet would implode because of so many SEOs taking that the WRONG way.