Guardian Readers and Weak Ties

Nothing to do with their taste in clothes, but lots to do with their ability to influence other people and spread word of mouth. These are the findings of new research from, of course, The Guardian.

It's actually a serious piece of work, taking on board existing theories about influencers and the process of word of mouth and building in new qual and quant evidence to bring the whole thing to life. It was conducted on behalf of The Guardian by Crowd DNA (led by Andy Crysell formerly of Ramp Industry) and is nicely packaged - although the design and art direction could be a bit more more exciting. Check it out here

It sits very much on the Gladwell end of the Gladwell vs Watts debate over influencers, in other words some people are much more influential than others and can be identified as such (it's easy, they read The Guardian!)

You can certainly see the logic that people with bigger social networks (made up of more 'loose ties', rather than simply strong ones to close friends and family) are likely to be more vocal and potentially more influential. But there is still the sense that word of mouth is very sector-specific, ie an influencer for nappy brands probably has little influence with regard to mobile phones.

Still, it is good to see media owners doing this kind of research in times like this - although I bet it was commissioned pre-downturn.

this blog is having two weeks off

Time Poor, Data Rich

I’ve been pulling together a piece to help empower planners at my agency to roll up their sleeves and get stuck in to doing their own insight gathering and analytics, rather than brief an analyst to do it for them.

When you line everything up it’s amazing what a modern day planner/researcher can get if they know where to look. I thought I’d share some of these tools here in the hope that someone out there might be able to add in some of their own.

Google’s Adplanner can tell you about site visitors, time spent and demographics (Google are a bit hazy on methodology, but they’re a clever bunch so let’s just play along). It can also allow you to create a bespoke audience based on sites they have visited, keywords used in search, etc, and then look at what other sites they visit (note, you need to sign in to Google to use it).

The daddy of web insight tools Google’s Insight for Search has been around a while. It’s a breeze to use and allows you to look at search trends for multiple terms by region and time. Top tip, combine it with their Adwords Keyword Tool to get actual search volumes.

Wordle has brought some much needed pizzazz to tag clouds. They just look so pretty. Paste in text from any doc, or try pasting in Tweets around a keyword (use this to find the tweets) to instantly visualise how a brand is being discussed.

Tag Galaxy must win the prize for visualisation. It crawls Flickr and grabs images tagged with your keyword before illustrating the results on a spinning interactive globe. Great for demonstrating the visual cohesion, or lack of, for a brand.

Brilliant new tools are also emerging from the likes of Facebook, Twitter, YouTube and other digital giants who are sitting on a ton of data. In time these will be as sophisticated and easy to use as the Google suite.

The most amazing thing though, is that all this is available to anyone, anywhere for absolutely nothing, zero, zilch.

How Much Is Enough?

The previous post about online audience measurement prompted a discussion in the office about the mechanics of measurement (heads up, bit of a nerdy post this one - the discussion moved swiftly to more pressing matters concerning Graham Onions and Steve Harmison).

Anyone who has worked closely with ratings data, be it TV, radio or print knows that the mechanisms for recording viewing/listening/reading can be complex. More importantly the definitions used are often no more than arbitrary. For those that are not familiar with these definitions these are the ones for BARB, Rajar and the NRS:

• TV viewing is defined as 'being in the room with the TV switched on'

• Radio listening is defined as 'listening for 5 minutes or more during a fifteen minute period'

• Newspaper reading is defined as 'reading or looking at for at least 2 minutes'

Beneath these definitions are more complex rules and regulations that also play a role in the process of measurement, eg the TSA areas for radio (that define where a station broadcasts), sort cards for print (that build the short list for the readership questions), and so on. But these definitions also have one other factor built in, namely persistence. You have to listen to a radio station for 5 minutes in order to be recorded as a listener. Two minutes is the rule for reading a newspaper. For TV it’s 30 seconds in a clock minute (consecutive). These are designed to weedle out those who aren’t really of commercial value to an advertiser.

So what is the persistence level for online? 5 seconds? 20 seconds? As far as I can tell there isn’t one (if anyone knows differently let me know). Imagine the scenario: you are sitting at home in front of your laptop and you click on a link. No sooner has the home page loaded than you notice your trousers are on fire. You shut down immediately and call 999. But you’re still counted as a unique user. Is that right?

The purpose of media measurement is to evaluate the likelihood that an individual is exposed to advertising. On this basis established media are getting a raw deal. There is no way the online world will give up audience by introducing a persistence level, so maybe established media should think about relaxing theirs.

Despatch from the frontline of the online measurement war

There is increasing hostility between the internet audience measurement rivals Comscore and Nielsen. As a Comscore subscriber Sigint received an email from their top brass to ‘set the record straight’ following a press release from Nielsen boasting about the size of their own measurement panel.

Clearly on the offensive, Nielsen had claimed to have the largest and most representative online measurement panel in the world. Comscore, now out-flanked, hit back hard saying that Nielsen’s claim was ‘factually untrue’. In other words a lie (this is about as good as it gets in the world of media research believe me). We eagerly await Nielsen’s next move.

But does any of this really matter outside of the PR departments within each camp? Sample size is an issue for any type of research but for gold standard currency data it’s generally accepted that while it would be nice to accurately report everything, in reality you don’t need to. The vast majority of ‘currency’ data users (who also fund the measurement systems) are only interested in the audiences to the more established channels and titles – the bottom of the long tail is simply classified as ‘other’. It makes it hard for new entrants, but no subscriber is going to double or triple their contribution just to measure the tiniest niches in the market.

In the US Comscore claim to report on 70,000 sites monthly, in comparison to the measly 30,000 that Nielsen now covers. But do we really need to know the audience profile to internet site number 70,000? Probably not. What we are really seeing are tactical exchanges of fire in the continuing battle for domination of the online audience measurement world. Meanwhile subscribers hold out for a ceasefire and maybe even the unthinkable, a single, better and cheaper measurement provider.

Insight by Algorithm

With simple digital analytics you can quickly and easily paint a rich picture of your customer. Let’s say you’re Ford and you want to know about visitors to pages on your site featuring the S-Max MPV. A quick run might turn up the following:

Visitors to the S-Max pages also spend time on car sites such as What Car and Car magazine - maybe they are looking for some independent advice. They also over-index on sites such as AutoTrader - perhaps they want to see what prices are like for used Fords. Intriguingly they also score highly for food related sites - perhaps they are modern cosmopolitan types who love cooking as well as cars.

They usually arrived at the site from Google, but downstream they often went to social networks after leaving the Ford site - to talk to their friends about cars perhaps.

The analysis also shows that they are most likely to be ‘new homemakers’ - so maybe looking for a car to suit a young and growing family. You can also see that ‘new homemakers’ are likely to visit interior design sites such as Mydeco and Fired Earth.

So, a strategy built around young families with an interest in food, design and interiors, utilising social networks, looks like a great idea.

However, the following could also be true:

It’s true that visitors to the S-Max pages did also visit What Car and AutoTrader, but they couldn’t really give a stuff about food websites, that was their partner looking for recipes. People usually share the same PC at home and analytics can’t accurately tell different users apart. The downstream traffic to social networks wasn’t the potential S-Max buyers either, it was their 16 year old using Facebook.

And they are not what you would describe as ‘new homemakers’. They have two teenage children. Their interior design tips come from the Argos catalogue. But they do live near to a new housing estate and the geodemographics that link to the analytics assign all addresses in a postcode the same category.

In reality of course most of this noise comes out in the wash. But it does highlight the increasing amount of trust we place in algorithms to decipher the amount of media data we have to play with.

Social media blurs the line between research and marketing

The majority of the media that people consume are still what we call traditional. They are one-way, ‘few to many’ channels whose creation and distribution requires complex infrastructure. They are also channels for marketing of course, and so it’s important to know who the audience is and how many of them are out there. As a result there are industries to create and distribute the media content, other industries to pump that content for commercial value and, sitting in the middle, the media research industry counting the winners and losers.

And it’s research that takes the high ground. It independently audits audience sizes and profiles and even when called to measure commercial effectiveness, research is usually seen as objective and beyond reproach. It’s an industry determined to keep its hands clean when faced with the dirty discipline of marketing.

So called social media are very different to their traditional forebears. They are open channels where media content can be created by anyone and passed around freely. Brands are increasingly in the business of creating their own content rather than just advertising around someone else’s (Nike is a great example). Social media are also comparable in scale to traditional media, at least in terms of reach if not time spent, so they are increasingly taken seriously by brands looking to connect with audiences in new ways.

In this new environment a new type of media research is emerging, one where the space between ‘clean’ research and ‘dirty’ marketing all but disappears. I’m talking more and more to these new breed of agencies about hybrid projects that start with research and end with marketing. A typical brief might flow like this:

- Identify key opinion formers/influencers within specific social media communities

- Recruit them to help refine strategy and develop content

- Turn these influencers into advocates to help mobilise communities and generate maximum word of mouth/mouse

To do this requires an agency that knows how to move between observation and participation. An agency that knows how to report demand but also how to stimulate it. This is completely alien to an industry built around objectivity and the need to neutralise a research effect. But looking at where things are heading there is plenty of business to be had for an agency that is prepared to get its hands dirty.