Panda … holding your nerve

I have now got over the shock of the October 13th hit … in fact I have spent a lot of time on Webmaster Forum and decided that the hit to me was no worse than to many people … I’m still cross that I was hit, but probably more cross at myself than at anyone else .. well, expect for one website which got a Panda boost and yet I know it has not been updated in two years and currently claims that there is only 1 show on in London which it also claims, incorrectly, ends this month …. useful? I think not

So, what have I been doing about it?

Well, I decided that we were potentially approaching the end game so there was no point in holding back. This time I have gone for a wholesale 301 redirection of UKTW what’s on listings to SeatChoice. Some may think this foolish as UKTW had seen a slow decline where SeatChoice had hit a brick wall, perhaps I should have made the call to move everything the other way? We shall see. I needed to remove the last vestiges of duplication within my own stable so that’s the call I made … from now, UKTW is responsible for the archive and for theatre users (notices, links, etc) and SeatChoice is the What’s On and ticketing service.

The move didn’t go as we ll as I expected in some ways. I was always told that 301 carried all the “link juice” with it so I expected a decline in UKTW to be matched by a growth in SeatChoice .. I got the decline but precious little of the growth. So back to the forum with one further plea for help .. this time I was picked up by someone really helpful, someone who, it would seem, is fairly local and actually cared whether I got it right or not … very, very helpful in that he undid a few misconceptions for me and pointed to some reasonably obvious issues that I couldn’t spot … when you work 8 hours a day on the same site you sometimes fail to see the obvious!

Basically though, I am coming to see that there’s a toolkit that you can use to get just some of the basics right, or at least help you to understand things that might be getting in the way. For me this has come down to

• http://gtmetrix.com .. make sure that your website delivers quickly as its an important part of user satisfaction
• http://www.copyscape.com .. are you copying someone else? is someone copying you? Do you simply have the same content as other people? Well worth checking …
• http://validator.w3.org … can Google make sense of your html
• Zenu – a link checking tool – no point in sending Google broken links
• Webmaster tools … check and recheck the error reports in here and make sure you fix them .. if google is reporting errors then its not getting the information you think it is!
• Google Analytics … make sure a drop in traffic is not caused by keyword shifts .. for us it can be, an important show ends and so does searching for it … check also where your traffic comes from but very importantly, how long it stays, whether it returns and how much it looks at!
• Webmaster Forum .. read, read and read some more, compare the advice to your own situation and see if it applies (be honest!), ask direct, simple questions yourself and take note of the replies
• AdSense .. if you use it check that it really is earning you, remove any Adsense blocks that are not worth it and remember, as AdSense gets better at delivering targetted customers you will lose more and more traffic
• Google .. sounds daft but search for your site and see how it looks on the result page … is it a “more desirable click” than any other result? Does your title and description draw the user in and encourage them to click? If you appear in the results and no one clicks on you Google will mark you down, if you appear in the results and people love to click on you then Google will mark you up … user behaviour is a part of the Google Machinery

Finally, make some changes, fix what’s wrong BUT don’t tinker for the sake of it … Google takes time to notice changes and also only runs the Panda every now and then (4-6 week cycle?) so don’t expect a change to work overnight – the only thing that will work overnight is a sudden influx of traffic because you got such a great piece of content that people started linking to you … the first article I put up about The River Line (with Charlie and Lydia Rose Bewley) got picked up by the Bewley fans – traffic went through the roof in a couple of hours and stayed high for days … it fell off, of course, and few (if any) of the fans bought tickets, but the links and retweets will be good for me longer term.

And rest …. be patient, if you perpetually tinker you will have no idea what it was that fixed your problem .. and if you do have a clue what worked you can then make it part of ongoing site philosophy … but remember, the days of stability are probably gone, you will need to keep an eye on this in the future

Panda Pain and Anguish

Ahh well, looks like I spoke too soon.

October 14th saw another run of Panda, as far as anyone can tell, and there’s a lot of screaming and running around like headless chickens happening of the Webmaster Forum … seems like it hit a number of people hard

Like us

Yep, after a few months of slow growth we were kicked in the teeth … dropped 35% on one site overnight, dropped probably 50% across all sites … not sure how much longer I can keep this up

Google Swings and Roundabouts

Well, as suspected, its a bit of swings and roundabouts when it comes to search engine traffic and changes recently.

In fairness, I’d say that Panda is not currently an issue for us, but in making the Panda smile we have (a) lost our position for certain keywords and (b) made our site convert fewer visitors!

So, now for the next round of tuning aimed at doing these two things

1) regain the ground lost for certain important key phrases

2) improve our conversion rate

Whilst not bring back the Panda frown. Tricky one, made even trickier when you consider that our updates, Google’s scan and the Panda runs all happen asynchronously … and we have monthly and annual changes in traffic anyway. This means that seeing if something has worked is a near impossible thing .. it used to be a little easier s the world stayed roughly the same for long enough to see things stabilise, now this no longer seems to be true.

The shock today was to lose our position in Google for “dirty dancing tickets” and “tickets for dirty dancing” where we have always (for a few years) been in the top half of the first page, not even the Panda upset that. Today, all of a sudden, we are not in the first 10 PAGES of results! This fact was made even more curious when I noticed that in Yahoo we went from nowhere to number 7 and in Bing we went from nowhere to number 1!

 

 

A more mellow Panda?

Looks like the Panda got a bit of a make-over, or, as Google always put it, was subject to the 500 or so algorithmic changes that they make each year. A recent post or two has made it clear that they have been monitoring the reactions on the Webmaster Forum (where I can be found lurking from time to time) and, despite the fact that most of the Bionic Posters seem to think the problem is 99% the fault of webmasters, Google has clearly made some changes. Good for them.

Of course, it is hard to know what actual effect the Panda changes had this time as my sites have undergone quite massive rebuilds, redesigns and checks since the last time the Panda strolled by but it would seem that we might be on its good side now … depends if its just run or just about to run!

Mind you, though traffic has grown very recently (fingers crossed it keep growing) its still much less than it was and, worse still, the changes in traffic and in the website content that I have had to create to keep the Panda’s interest mean  that this traffic is not anywhere near so lucrative …. Still, if I can get good traffic because its a good site I should, hopefully, be able to monetise that again … we shall see

Waiting for the Panda

Ok, lots of changes made and now to wait …

Changes made to the website might affect the SERPs reasonably quickly … the Googlebot spider is forever trawling across websites looking for new and changed materials so new items on the site can be indexed pretty quickly … by which I mean the pages will be shoved into the Google database

The Panda, however, does not run all the time. Every now and then (cycle to be determined, somewhere in the 6-8 week frame) Panda comes along and goes through the Google database weeding out what it sees as low quality sites … having done this, it goes to sleep again until the next time.

So in theory you could see a sawtooth waveform on your traffic, slow build up between Panda runs with large drops immediately after a Panda run – if this is the case then you have not yet solved the Panda problem because the best scenario is that you go UP after a Panda run and your competitors (or at least the spammers in your market segment) are the ones going down.

But to know this, you have to wait …..

Meanwhile, our good news is that our Reconsideration Request has been handled and no manual penalties were found on our site …

The Great Content Gathers Links Myth

I have been answering some more questions on Webmaster Forum, and asking a few … I keep seeing people say that links to your site are important (perhaps less than they were) and that if you create great content people will link to it – no, they wont!

Actually, having great content doesn’t mean you will get good links and lots of shares – I know Google thinks it does but its not universally true, it depends on the topic area – if people wont need to come back often then they’re unlikely to link and they wont share unless they want to keep up with your news and think they’re friends might want to keep in touch with you. I shall not be sharing links to recently used funeral services, great sites for those with exotic diseases or indeed a million and one other things. I search for, find and use hundreds of sites each week, the vast majority of them are informative and useful AT THE TIME … I have been doing this pretty well every working day since 1994 … I have actually created public links to, perhaps, a couple of dozen sites in that time … and compared to all the users I know, I’m prolific!

Panda fightback stage 2

Time for an update. Back in April, when Panda hit non-US English language sites our traffic fell around 86% – hard to give an exact figure as we were just having our best months ever. Over the next few months we saw a very small recovery followed by another drop and then we hit August, traditionally our worst month and, true to form, traffic started declining again.

Our traffic has always fluctuated, that is the nature of tickets/listings websites … it depends on what’s happening, a celeb appearing here, a new concert there, a new tour announced – all these things change our world naturally, spotting underlying trends is therefore pretty tricky.

During all that time we were not idle of course, as you may have read in our other posts. We are now a lot cleaner as far as SEO and robots are concerned, this may be somewhat more confusing for users but that’s the price they have to pay I’m afraid – no point in having the best organised site in the world if no one finds it because Google doesn’t rate it.

I have continued to take part on Google Webmaster Forum, am now a Level 2 (not very impressive really) and have received my first “Best Answer” – an answer based on things I have learned here. The Post.

Basically we have been making sure that we find and fix problems such as

• Unnecessary secondary domains that will appear as duplicates (remove and 301 them)
• Possible url mismatches that can give the same content (rel=canonical them)
• Really lightweight pages (noindex them)
• Content duplication (move content, combine content or rel=canonical)
• External links to our site from toxic neighbours (ask them to remove links, confess to Google as part of a reconsideration request)
• Poor meta descriptions that will fail to get users to click-through

We are still looking at things such as a large scale url re-write project to fix some url issues – meantime we have created our own url shortener to help in the management of archive links

Its a slow process but it seems to be working … there’s a long way to go till we get back to previous traffic and conversion levels but we seem to be moving in the right direction …

The other big change, however is that we try now to not get excited about traffic gains, nor depressed at traffic loses …. traffic has a habit of coming and going organically anyway and changes we make could take weeks to have an effect on our position … no need to overact!

Some Thoughts on the Panda Myth

Panda, the filter Google run apparently around 6-8 weekly whose job is to sift the wheat from the chaff in their huge database, is sadly not a myth … but some at least of the “quality measures” it appears to use are … in my very humble opinion.

“Every site must be chock full or all original, meaningful articles that add or else the site is of no value”.

That is to assume that all the world’s knowledge, and all the world’s serice requirements, are supported by articles.

Well, in fairness, knowledge is better when supported by articles but if I need to know what a owrd means I will look at the brief summary in a dictionary rather than reading twenty peer-reviewed thesies on the topic … note the word ‘brief’. Actually, for most people, for most topics, that’s what we want, the easily digested summary with links to deeper knowledge should we chose to take them …. I do not want some external algorithm (and I have worked with algorithms for many years) to decide on my behalf that I will not be satisfied without receiving the whole truth … even if I’m not qualified to understand it.

In the case of services this is even more likely that Panda is making a mistake. When I was at university, way back in the 1970′s, I was studying what was then called Computer and Communications Engineering – but in term 1 we had to do maths to bring us all up to speed. I remember one very long, very detailed, very obtuse lecture in which a professor proudly spent a long time labouring the proof that 0=nothing and any number +1 is bigger than the original number. Now, I can see that for pure mathameticians this is a fundamental proof that is required before higher level theories can exist. For an engineer working in the electronics world (at the time) of +/-10% components it all seemed a bit much.

I now presume, however, that if you ask Google for an on-line calculator page it will only be content to show you ones that prove all of the underyling maths of each button in major, original, articles before giving you any answers … but wait, it doesn’t it gives you this http://www.metacalc.com/ a fabulous on-line calculator with NOTHING else around it except google ads!!

The calculator is just one example. In the real world we also do not operate in a Panda-improved way. The most complex thing you are ever likely to won is a computer .. most computers are chosen in electrical stores where they are differentiated by a label with 6 bullet points and a price …. no big articles here. Most are chose by price, colour, brand and … well, in my case, keyboard feel! When you get the computer home what do you fine, a separate box for the manuals, no, just a quick start guide a small document (your computer) and a large document (your computer’s warranty) … that’s it. We do not get, nor do we expect, large amounts of articles and supporting documentation … and we do not get them because people do not want to read them .. so why is Panda obsessed with giving us sites full of something which every piece of evidence says we do not want?!

I suspect that Google people are a bit like me … they love manuals, I use them as bedtime reading, nothing I like better than browsing a good thick manual and learning what’s behind the buttons .. but I am an engineer, I am a bit OCD about information  and I am considered a freak by many

The second most complex thing you will ever own is likely to be your car … my cars manual is not a think document (it would be thinner if it wasn’t in 8 languages) and most of it is about the radio, sat nav and aircon. Another example where the thicker of the two manuals is the warranty!

I could go on …

I wont …

The Generic Search Engine Myth

I’m just wondering whether Google’s goal of being able to answer all questions for all people is really either achievable or desirable. I mean I know its desirable to Google but what about the rest of us?

Recent updates to Google’s searches have made it clear that sites providing search or linking are “no longer acceptable”. The declared aim, very laudable it is too, is to remove all of those horrible thin link sites from the results to give cleaner, clearer responses. Indeed, a common phrase now is “why would I want to link to you if you just link elsewhere, why not just go straight ‘elsewhere’ myself” ….

But the question in my head is, can a generic search engine with limited semantic capability really answer topic specific questions? If the answer is “yes” then I suggest that imdb, Wikipedia and a whole range of other sites need to pack up and go home as all of their internal links and all of their data expertise will soon be replaced with simple google search links

The simple answer, of course, is that Google can never answer all the questions because they don’t know what is inside the users head when they asked it …. I suspect that there are more people searching “panda death” at the moment than ever before but is this concern for our endangered wildlife or worry about Google’s own algorithm updates? Google can not know that, indeed the same person may well mean both things in separate searches on the same day! Of course, the user can expect to get somewhat off-beam results if they use only a few keywords in their search but I would argue that that may well be what we want to be able to do.

If I were standing in a library and I was looking for “giraffe” I have several options ….probably the most likely are

1) Pick up Encyclopedia Britannica (volume “ganges to hyena” or whatever) and get some basic information

2) Pick up a book on animals and get more specific information

3) Select a nice hefty book on African Mammals and get quite detailed information

4) Pick up a book on Wild Animal Veterinary Practices and … well, you get the idea

In search engine terms, (1) is like looking things up on Google, the editor of the encyclopedia didn’t know what context I was looking for (what the heck is a giraffe?, where do giraffes live?, how long do giraffes live? …) so will have filtered all the world’s giraffe knowledge based on what most people are likely to find interesting – just as Google does.

What this “library analogy” shows though is that, depending on what I want, I will select a “search context” before doing the actual search – I do not expect generic tools to do a specific job, I expect them to do the generic job well … as Google increasingly does. But, and its a huge Panda Butt (!) if I want more specific information then I will use a more specific book … in this argument for ‘book’ read ‘search tool’ …

So to answer situation (4) above I need a Veterinary site with a wildlife section where I can search or use the expertly gathered and annotated links that the site author has put together to help me go to http://www.fixing-your-giraffe.com whilst avoiding http://www.giraffe-quackery.com…; However, a general Veterinary site with those characteristics would fall foul of Google’s view of a “good site” as they don’t themselves have the Giraffe articles on!

In other words, it is very important to me to have a search engine (Google) which will actually find other search engines and indeed even high quality link sites!

Google  has mistakenly assumed that if I type a question into it then the ONLY good answer is the final answer … this is simply not true … if  Google tells me who knows the right answer then that is also a good outcome … unless you are Google and feel that the whole world of human knowledge should be yours to filter as you see fit.

Google has recognised some of this, of course, as they now geo-code results so that you can find things in specific areas and they also have a QDF function that changes how they view the “age” of an article – for some things they think that oldest is best and for others they think that freshest is best – what you think is neither here nor there

By the way, if you type “search engine” into Google then Google only comes fourth

Panda – Step 1

Step 1 of our attempts to recover from Panda was undertaken a little while ago, just after it hit the non-US English language searches with the consequent drop in over 85% of our traffic and our virtual disappearance from the SERPs (Search Engine Results Pages).

Our first action was to try and get over the shock, to realise that there was no point in bleating or howling at the moon and to start thinking about a long term strategy to recover.

Initially we thought this might be a ban. Why? Well, our SEO company had failed to warn us about the Panda under the bed so it came as a surprise. Lesson 1, SEO companies may not be all they claim to be (more of that later!). To check all this we tried the following

• Did a Google search for site:seatchoice.com to see if we were still in the results … we were
• Did a check on some of our key-phrases … things we used to be on page 1 for … you have to be persistent here, we eventually found ourselves on pages >25 … not where you want to be but we were at least still in the results!
• Did a check on a range of obvious penalty issues; site speed, malware (see Google’s Webmaster Tools), stupid htaccess errors, 404 errors, code errors …. none of these seemed to apply
• Put in a reconsideration request – unusually this was answered (a) quickly and (b) with the helpful phrase “no problems were found with your site”
• Decided it wasn’t a penalty ….

And breathe ….. Next step was to search and read … a lot. Doing a basic search on “loss of google position” revealed a flurry of posts from around the English Speaking World, it was from this that we learned about “Panda” which meant we could read up on it, assess that this was a very likely cause of our problem, work out that this was not going to be a quick fix and inform our SEO company (yep, we had to tell them!). I cannot emphasies too strongly how important this background reading is, really, everything you need to know is already out there if you take the time to look – and if you’ve just been Panda slapped then you do have the time, nothing will be happening to your site’s position until you read, understand and fix things!

Given that the SEO Company couldn’t come up with a co-herent strategy for recovery and seemed like it might have been at least part of the cause of the problem (bad links from bad articles placed in bad neighbourhoods) we fired them. As mentioned before, we got ourselves a more professional (expensive)  SEO Company but, after a few months, they too went as their strategy seemed to be a secret known only to themselves. It was then that I came across the advice “The best thing to do with your SEO Company is push them under a bus” – well worth considering. Oh, and while we’re on the subject … all those emails you get promising Google Page 1, bin them, now!

Having armed myself with a good deal of background material (save links to good articles, you may need to re-read them later once you understand a bit more about the situation and your site) I went hunting more specific guidance on the Google Webmaster Forum. This forum is manned largely by volunteers but with the occasional Google Employee lurking there too. It is a place to raise questions and receive the combined wisdom of good people who give up their time for free. A few quick pieces of advice if you’re new to it

• Read for a while before posting, both to check that you don’t have a standard question and also to get a feel for what the current topics and responses are
• Try to ask a clear question – put in enough information to help those helping you but not so much that they’re bored by the time they’ve finished reading it
• Don’t argue with people, they are giving you advice, either take it or leave it but respect their opinions, they probably actually do have more experience than you
• Be a bit humble about your site – they will see it for the first time and so are like users, you know, those mythical visitors who earn your keep?! If they say “on first impressions this is rubbish” don’t dismiss them as idiots, they are not, what they mean to say is “on first impressions this is rubbish”! This is valuable feedback and should make you look at your site again!
• Personally, I find it easier to ask a few specific questions rather than one huge “why is my traffic crap” one … read a few posts from other people, they think their sites are wonderful and have been unfairly treated .. so do you … why should an outsider agree?
• READ THE ANSWERS .. if someone takes the time to look at your site (and some people will spend a good half hour researching duplication, backlink history, code anomalies etc for you!) then the very least you can do is spend a few minutes reading and trying to understand what they say!

My first couple of questions on this forum had me told a few things that smarted

• Site was too slow …. turns out from Webmaster Tools it was slower than 70% of other sites
• Couldn’t see the point of the site … well ok, its a particularly British issue to have a mass/mess of suppliers and prices for theatre but the fault remains mine for not making the aim of my site clear as a bell on page 1!
• Just a thin affiliate site – well, the definition of this is that your site adds no value (content or service) to a set of affiliate links … this hurt, but it was the perception of new users (those on the forum) so it was probably the perception of Google … I needed to work on that
• Thin content … could be true in some areas, wasn’t sure what I could do about that given the size of the problem, but I put it at the top of my issues list
• Duplication – so many sites with similar content, content coming up under multiple urls, my own sites duplicating content

This gave me my first round of changes, and some of these were hard to get my head around as they went against things I had been doing, and been sure of, for 16 years … but basically

• Speed: I minified the js/css, removed some junk, increased cache times and got it to be faster than 52% of sites … I also now monitor this criteria regularly
• What’s the point: Increased visibility/explanation of “the point” .. still working on making it clearer!
• Affiliate site: Made sure that affiliate links were noindex,nofollow, tried to make sure that they were not seen as part of the main site
• Thin content: added a new blog on-site (as opposed to this one which is separate), added in user reviews, moved editorial from a secondary site to this one, increased ‘created’ content for otherwise thin pages (in our case this would be shows for which we have no available synopsis)
• Duplication: Really hard, closed down some sites with duplicate content and 301 redirected them to the main site, made sure I used rel=canonical links to help Google to understand which pages were which, re-checked sitemap, re-design pages so that not all pages for a show/tour carry all of the information in the same format

I also worked on Google analytics to make sure it recognised our iframe and ajax codes and went through and checked all our meta titles and descriptions ….

Phase 1 complete, took a few weeks but then all indications I had had were that the recovery from Panda could take months and all changes could take weeks and weeks to actually have an effect.

So after that lot the time had come to make sure I had good monitoring habits and I went looking for any deeper causes and issues that could be addressed … oh, and meantime, I kept the company going on our reduced income!