MIX! ATP curated by TV on the Radio

I can't be there this year. SADFACE . God bless you ATP and all who sail in ye, I miss you. Australia needs a Minehead! In the meantime, I'm making do with this mix from over on Soundcloud. Loving the Read more

Flying Scooter Man Gif: Friday Fiend

It's Friday. I don't need a reason to post these Read more

Goodbye Daily Updates..... or IS IT?!

It was good while I lasted. I really REALLy tried to post something everyday but of course work and life got in the way again. I blame house hunting in Sydney. But that's another story, and we'll get to Read more

Animal Latte Art: Daily Updates No. 11

Well today was a pretty good day for coffee I'd say. Popped upstairs in my very own office building to have a meeting over coffee and look what greets us at the little coffee shop upstairs? Who knew such Read more

Analytics

Optimise vs Optimize Part Deux

UPDATE

Google were clearly having their usual playaround, just joshing wit’ ya, haha funny funny joke around yesterday – and the SERPs for “search engine optimisation” are back to their usual selve with no auto-correct from optimisation to optimization.

However Malcolm Coles has done some more research into Google and its desire to change the way we spell and has made some really interesting points. It’s not auto-correct as such but Google is certainly nominating itself as a rather over eager teacher – telling you when you’ve spelt something ‘wrong’ and what it is you most probably really searching for – WHETHER (weather) YOU LIKE IT OR NOT.

Another example I stumbled across is Humour/Humourous. Aside from auto-correct, I’d also be interested to know more about how the algorithm works on ‘did you mean’ suggestions and whether this then relates to search volume, leading to possible auto correction.
So for example e.g. Humour has no “did you mean” suggestion and only shows correct results for the UK-spelt Humour with a u.

humour serps

Meanwhile Humourous does have a “did you mean: humorous” suggestion with the top two results from that serp being shown.

humourous serps

Interestingly Insights for Search shows that like Macolm’s Stationery/Stationary example, there has been a crossover in search volume – with the volume of searches for ‘humour’ dropping while the volume of searches for ‘humor’ has increased.

YET – there is not even a ‘did you mean’ let alone any form of auto-correction for that keyword! (Apologies for the poor quality of my images – few technical issues…)

I’m definitely planning on monitoring a few of these examples more closely going forward – there is obviously a logical reason behind auto-correcting poor spelling to some extent. However Google is taking is to absolute extreme in some cases, and seemingly not touching it in others.

And again for a user’s perspective – is it really preferred to by default auto-correct rather than give us the option of clicking on a ‘did you mean’ button?
Maybe not auto-correcting would actually also help teach us to spell more correctly in the first place if we are erring of the righteous path – AND allow those that spelt it correctly in the first place the option of not shouting irritably at the computer for auto-correcting yet again.

Meanwhile from an SEO perspective, being constantly aware of the possibility of auto correction may take a serious toll in keyword research. If I run a website of ‘humourous photos’ in Britain and I optimise for it – I don’t want to find myself in 3rd position (which is technically 1st) just because Google sees fit to show me the first two results for ‘humorous photos’ above JUST IN CASE.
Similarly – should I then be focusing my efforts on my ‘humorous photos’ just because I know I will be top for both humourous and humorous thanks to “did you mean”? Seems a tad unfair to me.

Posted on by AnnabelH in Analytics, SEO 2 Comments

ProSEO Seminar- These are a few of my favourite things

Oh crap I’ve done it again. Started a blog post with the best of intentions, got too busy (ok and distracted) to finish it and here I am 3 mins from leaving my house and retreating to 17 days in a tiny kampong in the depths of Malaysia with my mum and her 10 brothers and sisters. So I apologise for the incompleteness but I figure – better post now than possibly never. Consider this a ‘Part One’. Enjoy!

To  kick off , my apologies – that blogpost title was not an intentional evil plan to get Sound of Music kittens and schnitzel rhymes into anybody’s heads… although I’ve successfully managed to infect myself. Oh dear. Moving on…

ProSEO Training Seminar logo

ProSEO Training Seminar logo

I think it’s safe to say that all who attended the (official name): “Across the Pond” Distilled/SEOmoz Expert
Seminar came away significantly more educated, and generally excited about working in our beloved industry. Too often talks I’ve been to have sounded great, are really interesting and thought-provoking – but are also difficult to align back with the real world.

For me, the key to its success was the amount of genuinely actionable insights that were shared and helpful tips that can be immediately implemented and taken on board.

So… in no particular order and I apologise for the ramshackle nature of it – here’s my short version favourite key micro instantly applicable gems:

Quick and easy Analytics Customisations (that could change your life?)

Thanks to @WillCritchlow and the very first session of the seminar – Advanced Analytics.  Here’s 3 analytics customisations he gave that I found massively helpful, and have already set about impelementing for some clients.

NOKey – Avoiding that 3rd party shopping cart referral the easy way

Great simple solution to a really common problem for ecommerce sites. If your site uses Paypal or equivalent to take payments, then setCampNOKey can be used to retrieve the campaign no-override value from the URL. By default, this is not set, meaning that in tracking – the most recent impression will be the campaign that is credited with the conversion.

If you can send the visitor back to a page named, for example – www.abc.com?noo=1 and set the variable to setCampNOKey (“noo”) – this will then disable tracking from the referrer (e.g. Paypal) and credit it back to the original source (e.g. Google).

Supersetvar – Removing ‘last-click’ counts issues

We all know that your average product buyer is keen to shop around before making their mind up to actually purchase. The difference between how a visitor initially finds you, and the final source of that last click can be massive. Google Analytics, like most analytics packages in fact, attributes the source of that traffic using the ‘last click’ approach – which can be hugely misleading.
The solution to this is to use the _setVar funtion in GA. Or, as Will suggests for multi-touch tracking -  you can use the  Supersetvar hack created by Lunametrics, which will allow you to track multiple traffic sources rather than the single one which is set as default by GA.

The standard _setVar function would require checking for a previous visit cookie, then if there isn’t one – recording the referring url. By using the Supersetvar function, you can set a variable for each different time the same person visits your site, through different sources.

It also removes the need for you to check for visit cookies. Instead this simply appends information to the user-defined values each time (and doesn’t overwrite it like the standard _setVar function would).

You can then see this data in the user-defined field and use segmented reports to compare conversions where the buyer originally came to the site through an organic keyword.

Organise your SEO

Next on my list of favourites comes from @RichardBaxter‘s presentation on ‘Getting SEO Done Against the Odds’. This was all about structure, planning and organisation and it was hugely enlightening.

Below is the diagram Richard used to demonstrate a basic method to put in plan a complete SEO plan for any given project. This is not only beneficial in getting the work done – including planning for seasonal changes, it also demonstrates accountability to any client/boss and helps to set clear objectives and success criteria.

I will without doubt aim to use a model similar to this for all projects going forward – a simple yet very effective visual aid.

SEO-plan-large

Image credit: SEOGadget

Penalties and Filters

This presentation by @coplandmj created a lot of chat over the next couple days about ‘toxic backlinks’, and the same topic was brought up again by @DaveNaylor later in the seminar.

Premise being  that whilst we all know that good links are great, and serve to help us. Bad links from spammy sites can also harm us. So much so, that in some cases the act alone of removing them could actually boost rankings too.

So the moral of the story is – take them seriously. Investigate your site’s backlinks thoroughly – check ones that all come from the same dodgy looking domain. And bother to make the effort to see if you can get rid of them, it could be as worthwhile as building other good ones.

Posted on by AnnabelH in Analytics, Integrated search marketing, SEO, Social Media, Twitter 1 Comment

URL encoding and analytics tagging

A very quick little post as this struck me as so very basic and simple when it cropped up in an analytics tagging issue for one of my client’s recently. The theory was simple, the client was sending out a standard newsletter email and wanted to tag up the various sections of the email to be able track their impact on site traffic and visitor behaviour.

Check. Pat self on back for FINALLY getting the client on board when it comes to the absolute necessity of email tagging, and the simplicity of it – especially when using Google analytics.

Let’s take the example of this url: www.annabelhodges.com

Say I was sending out an email newsletter informing my blog readers of my panda giveaway, every person to fill in a questionnaire on my site will be entered to win a free panda.

The newsletter header would be tagged up something like this: www.annabelhodges.com?utm_source=newsletter&utm_medium=email&utm_name=pandagiveaway&utm_content=header

This would be fine sent directly through most of the major email marketing software companies, but what if like my client – I had just started to use a new in-house CMS that required a redirect via that CMS’ backend first before then redirecting again to the actual link url could look something more like:

www.annabelhodges.com?MailOutID=0&Type=3&Email=&Redirect=www.annabelhodges.com?utm_source=newsletter&utm_medium=email&utm_name=pandagiveaway&utm_content=header

This would mean that everything after that first query string ‘?’ would simply be ignored as it would all be associated back to the first redirect url rather than as a second entity in its own right with its own query string.

How to deal with this? Simple. Encode your url!

And it really is that simple – rather than using the ‘&’ as part of your second query string, encode it using %24 so that it continues to be read and passed through.

So your final url will look something like:

www.annabelhodges.com?utm_source=newsletter%26utm_medium
=email%26utm_name=pandagiveaway%26utm_content=header

In the case of analytics, the %26 is realistically the only change that you are likely to need. However for reference’s sake, here is a list of basic url encoding characters (borrowed from here):

; %3B
? %3F
/ %2F
: %3A
# %23
& %26
= %3D
+ %2B
$ %24
, %2C
<space> %20 or +
% %25
< %3C
> %3E
~ %7E
% %25
Posted on by AnnabelH in Analytics 1 Comment