In the 1993 film (wow it’s over 20 years old already!), Groundhog Day, Bill Murray’s character is forced to re-live the same day over and over until he learns some important life lessons. Arguably Bill Murray’s best film but that’s a whole other discussion.
For SEOs with much experience, the reaction to Google’s recent Panda 4/Payday Loan updates in May 2014 does create a sense of Groundhog Day-like deja vu.
Typically, the reactions to big updates like this take two main forms:
1. Numerous new and old ‘experts’ quickly rush to deliver the definitive answers to what Google has changed e.g. Panda 4 targets X so change that on all your sites quickly or the Payday Loan update penalty is triggered by Y so better change that too.
On the whole, such expert responses tend to endlessly recycle the old cliches: create quality content (which apparently a computer can make a subjective ruling on across all niches), improve user engagement, don’t use overly curated duplicate content etc etc.
If my eyes could roll back further in my head, they would because all of the testing that I do here inside SEO Traffic Hacks (where around 40+ experiments are running every week) completely contradict those expert opinions.
Are those guys actually testing (and reporting observations on just one or two websites is not testing)? Or are they just repeating what they read other experts saying?
A neverending loop of misinformation is the depressing result.
Apart from the corporate-cliched, conference-driven worlds that these SEO experts inhabit where they don’t seem to either [a] have their own commercial online projects, [b] do substantial testing (how often do you see these expert opinion pieces prefaced with “Based on all my extensive current testing here…” Hmm never?) or [c] do anything but write about what other companies or people are doing online.
On the one hand, you can just filter out that recycled nonsense as harmless junk like a poor sitcom but unfortunately it does convince some SEOs and Webmasters that [a] Google’s algorithm is amazingly omnipotent, [b] SEO is doomed, and [c] we live in a world of diminished commercial online opportunities – nothing could be further from the truth.
Now certain changes SHOULD be made to your websites for conversion optimization but that’s a completely different issue.
2. The second predictable result is that a load of SEO service providers suddenly and magically provide (insert Google update name) Penalty Removal offers.
My own clearly defined (and proven) penalty recovery strategy is based on traffic recovery, not site recovery. Maintaining the consistent flow of converting targeted traffic is my main goal, not the rankings of any individual website that gets penalised.
On the whole, these recovery services don’t work but of course their exceptions will be held up as the rule (a very common marketing ploy in IM and SEO).
The Panda 4/Payday Loan Update ‘Snowglobe’: What’s Actually Going On?
Usually when Google supposedly updates their search algorithm, the search results go topsy-turvy for a while (weeks) but the underlying SEO practice doesn’t change that much.
In fact, Panda 4’s updated algorithm currently (at time of writing, June 25 2014) has this site at #1 for ‘buy viagra’ (does it look high quality to you? And I haven’t even started on their spammy backlink profile, supposedly targeted by the Payday Loan update roughly around the same time as Panda 4):
In fact, any time you want to see how the Google rhetoric matches reality, go and Google a few superblackhat target terms like ‘buy viagra’ and ‘buy cialis’ (also several hacked sites on Page 1 for ‘buy cialis’ including the current #1, #2, #7 and #8):
Anyways, across literally hundreds of test sites, affiliate sites and client sites that I track (I currently do initial tracking on about 5000 keywords and then many more through deeper inspection with SEMRush), these are the main observations so far from Panda 4 and the Payday Loan updates:
1. Most of my sites’ rankings remained stable (90%+) including many test sites that have absolutely no content whatsoever (so much for Panda 4’s insistence on quality content!) – something I did a lot of before Panda 4, after Panda 4 for new test sites, I have minimal content on new test sites where I am evading Google’s Random Ranking Factor (RRF).
2. The sites that went down – the Panda/Payday losers – had EXACTLY the same SEO work done on them as the SEO winners from these updates – same PBN type, same PR range (usually PR2-PR4), same contextual nature, same number of links, same linkbuilding velocity. EVERYTHING identical.
It’s the same pattern reported elsewhere by webmasters: no single clear identifying factor to separate Panda 4/Payday winners from losers. Some webmasters report that sites of theirs with ‘thin content’ were slapped but other webmasters report the opposite.
One caveat on that is that I notice that empty sites are much harder to rank now so I tend to have thin sites with relevant YouTube videos when finding fast ranking winner sites. They rank well and quickly.
So it looks like another Google ‘shaking up of the snowglobe’ to keep SEOs off balance rather than massive actual changes made – and it was probably more driven by Google’s growing army of Indian site reviewers than massive actual software revisions.
However, I don’t presently have any sites that have curated, duplicate content so I can’t comment on the impact of Panda 4 on sites like that.
Nothing works 100% of the time in SEO and instead we are looking for generally consistent trends that we can profit from with a reasonable level of success – all based on a lot of current testing.
In short, don’t listen to anyone in SEO unless they are doing plenty of testing and ideally you should be doing your own testing too.
Like the old saying about look at what people do and not what they say, I am only interested in what Google (the search engine) does and not what Google (the company or the online journal SEO experts) says it can do.
And there’s no getting away from this dilemma: to Organically rank for decent commercial terms requires artificial linkbuilding in almost all cases but that artificial linkbuilding does carry some risk. The question then is how we strategically manage that risk and quickly respond to rankings setbacks.
That’s a topic I’ll come back to in another post.