Category Archives: SEM Info

SEM Info

Google Quality Score

In case you were wondering exactly how complicated Google’s quality score really is, take a look at this article on Optimize and Prophesize.

Beware: there’s different scores for rankings, min bid and search vs. content. Throw in a landing page score that doesn’t seem to work effectively at measuring conversions and there you have the gist of it..

Now all I need to do is go figure out what multivariate content delivery is.

PPC – Out of the Frying Pan?

With the specter of clickfraud showing no signs of abating over the coming months, the eloquent Jeffrey Rohrs has put together the sausage manifesto. Based on the principle that people who enjoy eating sausage should not watch it being made, the plucky Rohrs says that there’s a tendency for paid search marketers who like paid search to not put too much effort into seeing where clicks come from.

In short, the manifesto calls for the emphasis to be on the search engines to be more transparent about where ads come from, and shift the burden off the marketer.

As with all manifestos, it gets a bit whiney after a while, but the essential point is a good one – search engines can tell us more about where our clicks are coming from. And as click fraud becomes the big story around paid search, it seems to create a large degree of skepticism in non-search marketing execs.

Keyword permutations

Welcome to 2007. And if the projections are to be believed, an increase in paid search spend. Which in turn means an increase in competition.

Timely then, that this neat little permutation tool popped up courtesy of Dan Thies of the SEO Research Labs. Essentially, you split up the term into its constituent parts and the tool throws back all the possible combinations. Simple, and apparently effective.

See a demo, or give the tool a go here:

marketing to algorithms

it’s funny – you play around with an amorphous idea for a while and then you come across an article that eloquently and succinctly sums up that idea. the idea in question being marketing to algorithms.

the idea is simple – as more and more of our commercial life is lived online, we use filters and searches to to help fulfill our needs. and we leave a measurable digital trail behind us.

for instance, we may search the Opodo database for the cheapest flight, or use recommendations in iTunes to help decide our next download. for marketers, this offers two distinct possibilities:

  1. to find out as much as we can about how the algorithms work and use that information to give our products and services the greatest placement. ie. make sure the flight we are promoting comes up first in Opodo.
  2. to tap into (monitor) the trail and find out exactly what it is people are looking for. that might be to use yahoo’s keyword selection tool to pick the best term to describe our ‘flights to san francisco’ to ensure that we get the best coverage.

the second point seems to be the hardest to communicate internally, but the easiest to put into practice: that marketers can know more about their audience than ever before and use this in shaping their products and the way they communicate them. ie. bear in mind the top search terms when writing copy – these are the trigger words that will link our products/services to potential users’ intent.

the first point, the studying of algorithms has been the mainstay of fields such as SEO for a long time now. although to maintain the integrity of the information they provide, most website publishers won’t let marketers know how their algorithms work. what do you do? resort to black-box science experiments – changing possible criteria and looking at what comes out. for SEO’s this could be changing title tags on a page, and watching the effect on rankings. where this gets interesting is when you look at the algorithm as a proxy for the masses that lay behind it. the black box will always have one common objective – to provide the users with the most useful information. what i’m suggesting is that we use the digital trail to find out as much about our audience as possible, but then use traditional techniques (like well-written customer-focused copy, as opposed to over-optimized SEO-heavy copy) to attract the crowds, and hence the algorithm.

in the SEO world, this is called employing ‘white hat’ techniques, ie. spending less time looking at every possible factor in the search engine’s algorithm, and more looking at what the search engine’s visitors might need. after all, the black box is just the proxy.

getting to grips with Adwords quality score

so, how exactly do you figure out what’s the best way to maximise your adwords campaign?

this article from se roundtable gets into nitty-gritty.

main points:

  • ctr is important but not the be-all and end-all
  • the ctr is normalized for ad position: if you are lower down the rankings you won’t be penalized for a low ctr
  • think of the ad as the bridge to relevance: the link between the keyword and the landing page – both of which are important for the q score