Local keyword variations – thoughts from trip to London

Down in London over the last few days meeting clients, existing and new. It’s good to have face-to-face time as despite our good use of tech from skype to extranets, a face to face does provide an opportunity to pick up nuances and subtleties that help us do our job better. Being in the room is hugely useful.

An extension of this came up with regards to keyword research. This area is really the salt mines of SEO – figuring our audience behaviour and demand. Getting it right is crucial and very hard work. In two meetings we had some interesting results for local differences in searcher behaviour.

Local keyphrases differences – and how to capture the advantage

In the first, there were some very interesting differences in phrases as you traveled across the UK. A north and south divide was very evident. The result is that in order to increase visibility for a company with a huge inventory there is an interesting opportunity in terms of optimisation.

We had realised that the creation of bespoke titles and descriptions was going to be impractical in terms of both resource and budget for the client. A pragmatic opportunity was required… cue concatenation schema. This is  the creation of title and description (but not exclusively) syntax variations which are assigned to content across the site. Within the syntax, keyphrases including product attributes, names and calls to action are assigned.  These are mixed up across the site based on fields which for this client are part of a consumed xml feed.  After our meeting we also added a further refinement i.e. the change of keyphrases based on location and  local search demand.  It’ll be interesting to see how much additional traction this element adds in terms of driving relevant, valuable traffic.

Local keywords aren’t always clearly logical

The other local keyphrase issue came from a meeting with a client in a completely different sector. The issue here was that in taking four neighbouring counties around London and looking at search volumes for the client’s services, one of the counties had  no  reported search volume while the other three had  ok levels of demand. What caused this difference? Do people in Buckinghamshire not require this client’s particular service? is the market sewn up by a competing brand?  was our tool acting up or perhaps the folks in Bucks consider themselves more London than Surrey, Middlesex and Hertsfordshire and use “London” in their searches? It’s not something we’ve got to the bottom of yet.

Takeout: Doing keyword research is hard work if you do it properly but it doesn’t end at the drill down. Having local knowledge and being able to apply it to a solution can ensure that you get that extra bit from your keyword strategy.


Tandem gets great results for leading marketing agency

One of Scotland’s leading agencies commissioned Tandem Digital to work on their site visibility. When we won the business they were nowhere to be seen on Google for their services in the Scottish market or the UK.  In the last few months they have launched their new site based on Tandem’s recommendations and over the last two months they are now hitting first page visibility for some pretty competitive keywords.

Searcher Behaviour – where is the traffic?

Search behaviour – increasing impatience of search engine users

49% of search users change their keywordphrase before checking out the second page. Does this mean that searchers are growing increasingly impatient with the results served by search engines. Absolutely, there’s no doubt that in the “think-it-get-it” environment of the web, users expect to have the right information at their fingertips.

Additionally if you appear after page 3 in the SERPs for a phrase, you can expect only 2% of users to go that far down the list. A whopping 68% of searchers don’t venture past the first page.

TAKEOUT – Your listing on the search engine returned page(SERP) not only has to be in the top listings on the top page, it needs to be as compelling as possible. Engage the searcher with content which meets their needs.  It goes back to understanding searcher demand and levels of intent. Too many keyword research projects focus solely on looking at volume and search share. This misses out half of the equation and has optimisers placing effort in the wrong keywords – lowering ROI. By understanding how searchers behave, how they communicate their needs and relfecting this in your content and messaging, you balance traffic volume with relevancy.

Search isn’t about throwing as much at the wall and hoping some of it will stick. The search industry is full of people being busy achieving nothing. By applying communication principles to search, commercial and communication objectives have a better chance of being met.

Vertical search vs blended search choices

How are searchers choosing to search?

Blended search, also known as Universal search, has shaken up the search industry over the last year. This is the serving of different types of content in a search engine returned page (SERP).

Vertical search is the ability for searchers to look in content-type channels in a search engine e.g. news or images.

According to a study by Jupiter Research, searchers are still a little shy of vertical search, but when different content types are presented in the SERPs users are much more likely to click on these results.

Searchers still unsure of vertical search

The response for searchers using blended search results differ, perhaps indicating that either users are more comfortable with the main results being served to them than going down a channel which they perceive to offer less choice.

Blended Search responses


Blended search offers search marketers additional means of getting traffic. While searchers are still reserved about going down vertical search channels, this may change in the future as users get less patient with search results. (68% of users now don’t search past page 1!!)

However what is clear is the success of blended search – still in it’s infancy but it seems that searchers are liking the choice it provides.

Yahoo and MSN back in talks – Google looks on

MSN and Yahoo! are back in talks.  You’ll remember that MSN approached Yahoo! with a $40 billion evaluation which was later sweetened by $7.5 billion. On refusal Carl ICahn started getting vocal with his 4% of Yahoo! and a threat to depose 10 of Yahoo’s directors.

Google are in the sidelines watching as the number 2 and 3 in the market (leaving Yandex out of European figures for the moment) disuss becoming a possible number 2 with a third of search market share. something not to be sneezed at with the value of search valued at about $40 billion per annum and set to double by 2010.

It’s certainly all to play for. The test of Google adsense being used in a percentage of Yahoo’s paid search ads last month has attracted the anti trust boys so the previous avenue of a commercial agreement which would see Yahoo as inventory for Google sales looks like its being put on the back burner.

Where will it all end? Without Yahoo, MSN will struggle to grow or even to maintain marketshare with their search offering being so limited in comparison to the products offered by the other two goliaths. Will MSN pursue the social media route instead with it’s proposed talks with Facebook.

Exciting stuff.

SEO not Spam – it’s official

Matt Cutts has come out and said it.  Search engine optimisation is not spam. The white hat stuff that is. I’m tempted to download his presentation at the web2.0 conference, “What Google knows about spam” onto my iPod to play to those caught somewhere in 2000 who think that all online marketing is spam.

Nice presentation which covers some of the basics of protecting a site from web spam – worth a look.

Yahoo glue – search getting sticky

Yahoo India is the test bed for a new iteration of Yahoo! search. Yahoo! Glue serves a page of search results but also adds in links to popular pages across the web which Yahoo! thinks might have interesting content on the same subject.

It seems that the sites searched by Yahoo! change dependent on the type of content the searcher is looking for.

Test Searches

As Yahoo Glue is being tested in India, I played around with Indian based queries. Here’s a couple;

A search for “Ashe Boshle” – India’s premier diva of Bollywood pop returns results from Yahoo Search (India partition), Wikipedia, LastFM, YouTube, Yahoo Image search.

A search for “Bangalore”, the city where Yahoo! India is based returns Yahoo Search (India partition), HolidayIQ, Flickr, Yahoo! India Maps, Yahoo India news and sponsored links in both the Yahoo! search area and an additional area for sponsored links. Yahoo! Glue also provides related pages.

In other searches I’ve seen results from google blog search and howstuffworks pages. There’s also a contact email for sites which want to appear on that Glue SERP.

Yahoo Glue – search portal

It’s an interesting development. Yahoo! which has differentiated itself from Google through providing content portal interfaces to it’s users as well as search is now providing search users with basically a search portal. So while Google moves forward with universal and blended search in their SERPs, Yahoo! Glue separates out the results. The user can choose to follow presented content based on their perception of brand loyalty.

The danger is the creation of an implicit oligarchy of internet content – sites which get the Yahoo glue endorsement as a popular site. On the other hand, with Yahoo! looking at costs of content, it does provide useful information to searchers at a fraction of the cost of traditional Yahoo portals.

Wonder if the next step is being able to customise the popular sites based on the searcher’s profile.

Have a play at http://www.yahoo.in.

GoCompare whopped by Google’s bad link big stick

Really good post on the Hitwise blog following up on GoCompare’s natural search visibility for the term “car insurance” plunge after being hammered by Google’s inward link moral authority. Great graph on the change in proportion on natural “free” traffic to paid traffic from PPC. That’s gotta hurt!

Insurance, as we digital marketers know, is one of the most competitive industries online, (should show you the proportion of insurance spam I get on this blog!) with profits being made generally on the 2nd or 3rd year of a policy, affiliate commissions have been squeezed over the last few years. Can any player in the insurance arena take the risk of getting into Google’s blacklist?

With natural search continuing to be provide the backbone of digital marketing activity, playing fast and loose with the Big G and it’s seriousness about linking is to be avoided.

Agencies and Paid Links

Links from a bad neighbourhood is a sure way to destroy a site’s ranking. Getting paid links (I have no idea if this is what happened to GoCompare) from 3rd party offshore organisations has been used by many SEO agencies. Highly profitable for the agency (i’ve heard of examples of at least 50% pure cream on 6 month campaigns with agency work limited to sending over a few anchor text variations to some link farm sweatshop in Asia.) and in the past it worked with these campaigns getting great results with linking shoring up some pretty pedestrian SEO campaigns.

Allelulia – that’s over!

Now we’re into a much more interesting times. The risks of bad neighbourhood liking is a reality which in-house marketers are picking up on. (Cue epiphany moments, shafts of sunlight and heavenly choruses).

Linking will always remain a challenge but, to state the bleedin’ obvious, the creation of valuable content (not hit and run link baiting) and utlising web2.0, and indeed 3.0, platforms responsibly will sort the men out from the boyos!

Bet Gocompare think so.

Yahoo!has more than one choice

Times Warner (owners of AOL) and Yahoo! are in discussion regarding a possible deal which would allow Yahoo! to gain more shareholder control while increasing Time Warner’s market penetration. The Wall Street Journal goes into the ins and outs but basically AOL would be rolled into Yahoo! and Time Warner would stump up some cash giving it 20% of the merged entity and allowing Yahoo! to buy back some of its shares. Sopunds like a sweet deal for both. Especiall considering that Yahoo! was the most visited site in the US last month.

News Corp is now also back in the frame but in a triumvirate solution with Microsoft rather than on it’s own as in previous talks.

Google’s AdSense test roll out on 3% of Yahoo!’s US searches is a great way for Google and Yahoo! to measure the potential revenue generation from a Time Warner (AOL) Yahoo merger…

Yahoo! looks like it’s got more choices – wouldn’t you love to be a fly on the wall in Microsoft Chief Executive Steve Ballmer’s office ?

Interestinger and interestinger.

Universal search and CTR

Universal search is starting to shake up search marketing. The recent posting on the Comscore blog makes the interesting point that while CTR for paid search might decline, referrals from Google on the whole could increase i.e. Universal search has the potential to drive more valuable traffic to sites.

At first sight this looks like a very non-commercial move from Google that might have a negative impact on thier paid search revenues (shareholders will no doubt have another wibble). But Google’s success is built on it’s importance as a resource to consumers/searchers. By blending the results, Google is servicing the needs of their audience first, and as sure as a cat’s a hairy beast search marketers will follow. SEO will become increasingly vital.

Put this in the context of CTR for paid search now averaging about 0.2% and video CTR on sites like Youtube averaging between 1-2%. There is every chance that Universal Search will provide more relevant traffic to sites and with around 17% of all searches returning blended results search marketers need to rise to the challenge.

Watch out for the scramble to market as the reality of this filters down the marketing hierarchy.