February 2008 Archives

Genius bit of sub-editing at Marketing Week which has brightened my day no end!

Even the photo caption is rich in innuendo.  Or is this just my grubby mind?

Will Cooper at NMA has a story about a new working group at Google looking to develop a more complete picture of how the entire online journey impacts on sales.

Clearly this is a hot area right now - I was at a buy.at event last week, and even people within affiliate marketing are looking for alternative metrics to CPA.

Getting a more accurate picture of how everything works together is very much in Google's interests as it gives them a powerful argument to cross-sell customers into their content network. 

The Guardian and Times both report that the trade union Unite has launched a 'cyber attack' on Marks & Spencer.

It seems that Unite has managed to target searches on 'Marks and Spencer' and 'M&S' to deliver a message complaining that the retailer has encouraged discrimination amongst workers in the meat packing industry.

It feels like a PR coup rather than a genuine 'cyber-attack': I haven't been able to see Unite's ads for any M&S related term and targeting brand terms would be a clear infringement of M&S's trademark protection on Google.

Still, it's an interesting story - and brands without effective trademark protection would certainly be vulnerable to this kind of action in the future. 

ThoughtGadgets gets top marks for the headline 'Google results falling fast. Here's why.' 

But their analysis - that search volumes and therefore ppc revenues are consistently falling - is surely wrong.

Google Trends data does show the volume of searches for certain generic terms dropping over time (relative to total search volume).  But that masks two other trends:

• More savvy consumers searching for three and four word search phrases
  
• A rise in navigational search which results in more searches for brands

So here's the Google Trends data for 'music download' - which shows a consistent decline (and peaks each Christmas.

And here is the trend data for the top branded term in this sector, 'itunes', which shows a clear rise, with even bigger peaks each Christmas - well done Apple!

Of course, from a revenue point of view this might not be such great news for Google - searches on brands with brand protection in place will deliver a much lower CPC than key generic terms. 

 

Update: just noticed that Hitwise has some data showing that traffic going from Google into shopping sites has actually increased over the past 3 years.  So whatever the source of Google's woes, it isn't a drop in search volumes.

I work for an integrated digital agency and for us it is common sense that all digital channels are inter-related and have an impact on each other.

But calculating the impact of say display advertising on search and making sensible optimization decisions is quite complex.  It's also a pointless exercise if clients are obsessing over metrics like CPA as measured by the 'last click wins' method.  Basically this means that the last click before purchase deserves 100% of the credit for that sale.

So today's announcement from Microsoft of the beta launch of  Engagement Mapping is interesting.  It's the first big new thing to come out of Atlas since their purchase by Microsoft, and looks very much like an enhancement of the conversion attribution modelling stuff that Atlas have been talking about for a while.

Essentially Engagement Mapping will look backwards from a conversion and see what ads the consumer has been exposed to before purchasing.  "Engagement ROI evaluates and assigns measurable value to a consumer’s interaction with ads, giving advertisers and publishers a more complete picture of online behavior."

Measuring all this stuff is certainly a step in the right direction, but the real gold is in working out a model that attributes a value to everything.  The last click at present is more likely than not through a search on brand terms: we need to know what prompted that search so we can push budgets at the right channels.

What kind of factors might influence the level of engagement you get from an online ad?  Stuff like size of ad unit, level of interactivity, amount of clutter on a page from other ads, degree of targeting behind a placement.  Some of this data is already sitting in an adserver, but it's tricky to pull it out in a meaningful way.

The current obsession with 'last click wins' clearly favours search and affiliate marketing - but it also directs budgets away from online media that probably play an important role in the decision making process.  That includes good quality blogs and forums, and of course the online properties of 'old media' companies - newspapers and magazines particularly.

So a small step towards re-balancing budgets - and hoorah for that! 

 

 

Over the weekend, SEOFastStart gave out some information about a private Adwords beta program. A few Adwords advertisers in the US have been invited to participate in a beta test of 'Automatic Matching':

"I'm excited to tell you that you have been selected to participate in a beta for our new Automatic matching feature which will be starting on February 28th.

Automatic matching automatically extends your campaign's reach by using surplus budget to serve your ads on relevant search queries that are not already triggered by your keyword lists. By analyzing the structure and content of your website and Adwords campaigns, we deliver more impressions and clicks while maintaining your current CTRs and CPCs.

For example, If you sold Adidas shoes on your website, Automatic matching would automatically crawl your landing page and target your campaigns to queries such as: "shoes" "adidas" "athletic", etc., and less obvious ones such as "slippers" that our system has determined will benefit you and likely lead to a conversion on your site.

Be assured that automatic matching will try to never exceed your budget. If you're already meeting your daily budgets, automatic matching will have a minimal effect on your account."

So it's a *bit* like broad match, except that Google's algorithm is magically choosing the keywords that it will broad match against by looking at your website as well as the current keywords you are targeting.

Most bloggers seem to be seeing this as yet another opportunity for Google to rake in more cash by helping advertisers to spend up to their maximum budgets each month. But this is explicitly denied by Google Adword's unofficial spokesman 'Adwords Advisor' who comments that Automatic Matching "…is not intended 'exhaust the budget' - rather it is only meant to deliver additional traffic where performance metrics such as CTRs and CPCs stack up well against the adgroups current CTR and CPC. If there is no additional relevant traffic to direct to the advertisers campaigns, automatic matching will not spend additional money."

This is just a beta test and may never be rolled out. However it does suggest a certain pragmatism within the Googleplex around the previous holy grail of relevance. Back in the real world, a media buy that placed adverts for airlines alongside editorial on sailing would be simple common sense: for Adwords, you risk low CTRs, high CPCs and ultimately having your keywords disqualified.

I can see a real advantage for clients operating in sectors where CPCs are relatively high – like insurance and loans. But if these advertisers start being automatically placed in 'less relevant' areas, that can only push up bid prices in those areas.

Of course the real benefit will be for agencies whose goal is simply to blast through a client's budget each month with little thought for effectiveness. Automatic Matching gives them a great way to guarantee that budgets are spent – and all with a minimum of human intervention. So perhaps that will be the final scorecard: good for Google, good for lazy agencies, bad for the clients at lazy agencies.

I was speaking at Engage for Search today, and after my session had the chance to sit back and enjoy the other speakers.

The session I enjoyed most was by Nilhan Jayasinghe, who is Head of Search at iCrossing in the UK. 

I liked the way that he seems happy to pose questions without having all the answers up his sleeve.  Which is reasonable, because frankly some questions are a bit tricky.

So he was talking about how Google infers relevance from what the published community is talking about and linking to.  But the published community isn't always right!

A good example is the Google search results page for 'table' - see screengrab below.

When the man in the street searches for 'table' he is thinking of a wooden object upon which he can rest his pint of beer.  But when the online community thinks about a table, more often than not they are thinking about an HTML table. 

The result - in Google's algorithmic world - is a strange situation where the natural results and paid results are about totally different subjects.

So furniture stores have two choices.  They can wait for Google to 'fix' this situation - or they can facilitate their own conversations and start building equity around their search results that way.

iCrossing are working with MoreThan on a content site - More Than Living, which is one possible solution to this.  I'm not 100% convinced, but it's certainly an interesting experiment - and it's always fantastic to see SEO companies creating useful and interesting content rather than stuffing fake blogs with keywords!

Great post - as usual! - from Avinash Kaushik on the impact of radio ads on web traffic.

I'm doing a session at the IAB's Engage for Search tomorrow on the integration of search marketing with other on and offline media - this data might have arrived just too late for me to rip it off, sorry, add it to my presentation.

Anyway, there are some great slides on the potential way you could measure web traffic against radio advertising impacts.  Well OK, they aren't real slides - they are just a bit too neat for that.  But they do show a plausible example of how radio ads could impact on total site traffic and on branded search.

It's exactly what we showed through a piece of research we worked on with Nielsen Netratings.  We asked consumers what action they would take as a result of seeing a banner ad: 58% said that they would search either for the brand or for a related generic search term.

Essentially search has become an all-purpose glue between the on and offline worlds, and between different online channels.  No one needs to write down or remember a URL - you now trust that Google et al can quickly find you the site you are looking for (or a better one).

What are the lessons from this?

Well in the real world, very few people are tying together seemingly unrelated metrics like radio ratings and branded searches.  But if they did, they might start to see some very interesting insights.