Social media measurement: the experts at work

The issue of how to measure the value of PR activity is contentious at the best of times, with the debate around AVE (advertising value equivalent) raging as fervently today as it did ten or 20 years ago. But a new subject for debate has reared its head, with the rise of social media. Suddenly, the question of measurement becomes far more complicated.
Like it or loathe it, AVE is at least a fairly well-understood way of measuring traditional print or broadcast coverage. But this does not translate to social media.

There is no AVE for a Tweet, a Facebook fan page, a mention on a blog or the number of times a video clip has been viewed on YouTube. Leaving  aside the thorny issue of whether AVE is a suitable measurement of PR activity in the first place, if a client wants to know what its social media coverage is actually worth, how can PR professionals respond? ‘It's a hot potato,' admits Will Scougal, head of digital content at Mischief. ‘There is no official or industry-accepted way of measuring, in terms of financial return, how much a piece of social media coverage is worth.'

Measurement solution

Obviously, the ideal solution is a way of measuring social media coverage that the PR industry can embrace and clients can understand. However, those in the know do not see this happening.

‘Owing to the nature of the space and the rapid change that is part and parcel of it, I do not believe we will see a robust enough solution to suit everyone's needs any time soon,' says Paul Armstrong, social media director at Kindred. ‘Social media should be a long-term business solution and any measurement solution needs to reflect that - the right tool for the right job.'

Use of tools

Scougal, Armstrong and others in the field say the first rule of measuring social media coverage must be: what does the client actually want to achieve?

‘It is not about measuring "something", it is about measuring the right thing, to link back to the client's objectives,' agrees Ged Carroll, director of digital strategies at Ruder Finn. ‘If you are looking to drive sales, you would look at how many people have posted or responded to posts on a Facebook group, for example, and how many people have linked through to the website.'

For this, Carroll advises using tools such as GoogleAnalytics,  Omniture and tracking URL mentions with Radian 6.
There is a plethora of tools, both free and paid-for, available to aid measurement and evaluation of social media coverage. PRWeek has asked two social media experts to explain which tools they would use to measure and evaluate two very different types of social media campaign and their responses appear here and overleaf.

However, Carroll warns that tools can only do so much. A computer cannot measure sentiment accurately. ‘Machines do not understand humour or irony,' he warns, citing as an example some review.

Social media Case Study 1: The celebrity

Expert  Pam Lyddon, CEO, Bright Star Digital  (pictured, right)

The brief  The client, a singer with a new album out, wants to monitor online chatter about her following the news that she is to separate from her footballer husband. Describe how you would monitor, measure and evaluate coverage of this celebrity on social networking websites.

Which sites would you monitor?

Facebook, Twitter, Bebo, YouTube, newspaper and magazine sites, forums and blogs.

What tools would you use to measure the coverage?

Google alerts for real-time alerts from across the web universe; Twitter feeds  for trends across Twitter; Socialmention.com, a free social media search platform that aggregates user-generated content from across the web into a single stream of information; Broadreader for comments and forums across the web universe; Radian 6 to measure and analyse the social conversation in the web universe; Nielsen Buzz metrics to measure online buzz and what people are talking about.

How would you analyse the coverage?

I would monitor the artist using various social media tools on a daily basis to see what people are talking about, before we started a campaign with them. I would want to see how they are perceived in the online world.

We used Cheryl Cole as an example and monitored her for a week. The majority of her coverage in the social media sphere was about her personal life and appeared to be extremely positive.

People are very focused on her personal life and wellbeing, the fact that she pulled out of the Jonathan Ross show and a scheduled appearance on Radio One, and they are also focused on her style.

Her album has been out for a while, so people are talking mainly about her new single and the video with her alleged dancer beau.

As she is a recording artist, you would expect to see more comment on her skills as a singer – this is not the case with Cheryl.

Advice Celebrities can work social media very well if they are strategic. They can easily manager their reputations and control news stories by speaking directly to fans via official pages and sites.

Online media are fantastic for engaging with fans, connecting with them, building up a following and rewarding them by posting up new content for fans on official websites, live webcasts and exclusive photos.

The fans will feel closer to the artist and use the power of word of mouth to build the artist’s reputation – just look at the recent Lady Gaga and Beyoncé music video. This broke world records, because of the use of social media.

How would you measure the value of social media coverage?

There is concern in the industry about not having an AVE for content and social media.

However, as an industry, we are moving away from traditional media and we need to recognise this and educate staff and clients. ComScore is looking into providing a currency and Procter & Gamble is looking at the number of mentions of a brand across online media.

In this case, I would suggest we monitor the chatter and cross-reference with the download sales of the album, single and any brands with which the celebrity is associated. It would also be worth monitoring traffic at the artist’s official page. that appeared on Amazon of Katie Price and Peter Andre's album of duets. ‘Some people were writing things like "after hearing this record, my blind cat regained its sight",' says Carroll. ‘It took Amazon a while to sort through those reviews because machines did not get the sentiment.' On the face of it, these were rave reviews, but they were clearly posted with a heavy dose of irony.

The overarching challenge remains establishing a credible way of measuring social media coverage, but the new digital world has also raised questions over more traditional forms of media measurement.

‘Social media, and digital media in general, offer PROs a more precise look at who is reading their handiwork,' says Armstrong. ‘Be it through a bit.ly link or a more robust, sentiment-based monitoring solution, PROs are getting a more realistic view of how many people read their work than in previous years. I am sure most are more than a little surprised at the numbers they are seeing.' With measurement this precise, the days of asserting, ‘well, our article was in the FT,  so that will have been read by 200,000 people' are over. But the benefits are clear. ‘PROs have never before been in such a good position to say "because of X, we saw Y", and track it,' says Armstrong. ‘Thanks to ever-increasingly sophisticated measurement and monitoring tools, the days of ABCs, pass-on rates and multipliers are numbered, and rightly so, because basing decisions and reports to clients on actual clicks, rather than how many copies were printed that day, has to be the future for which we are all hoping.

Social media Case Study 2: Political communications

The expert  Ivan Ristic, director, Diffusion (pictured, right)

The brief  A major political party that wants to be elected into government in the upcoming general election has been using blogs, a website, Twitter and Facebook to try to engage potential voters in conversations.

What sites would you monitor for chatter and what tools would you use to monitor them?

The first step would be to choose a big-beast monitoring solution such as Radian 6, which covers standard blogs and forums but can also monitor Twitter, YouTube and Flickr.

You would then look to supplement this with platform-specific tools. For Twitter, tools such as TweetDeck and Twhirl allow you to organise tweets based on specific search terms, such as the party leader's name, and identify emerging trends and topics.

I would set up automated alerts from a specialist video search engine such as Blinkx, which trawls for video content from a wide range of sites and platforms, not just YouTube.

Advice Huge amounts of social media content would be identified on a daily basis during an election campaign and it would be very easy to be blinded by the sheer volume of data. To be using social media monitoring effectively, there would need to be clear roles and responsibilities within the communication team and a clear process for quickly escalating potentially damaging mentions that would require an immediate response.

I would ensure that all key personnel across all departments within the political party used tools such as Netvibes and Inboxlistening.com to create their own personal dashboards and set these as their web homepage. Dashboards can be set to automatically pull in content from platforms such as Google and Twitter in response to very specific keywords relevant to that individual's role and responsibility.

How would you measure the coverage?

Effective monitoring needs to be a combination of powerful automated software that can identify online mentions and powerful human analysis, which can determine sentiment and context.

There are, unfortunately, no short-cuts, and lofty claims made by providers on the accuracy of their ‘auto-sentiment' features should be treated with a pinch of salt. Politics, like many topics, will be discussed online with a huge degree of nuance, subtlety and sarcasm, so there is no substitute for real-life consultants who can expertly analyse the tone and implications of a comment and be in a position to act immediately if required.

Social media by their very nature are long-term and iterative, but it is naive to think you can move away completely from a traditional ‘campaign' mentality'. For a political party, there will be clear campaign peaks such as local, regional and general elections where social media will need to deliver real results over a short period of time.

However, the success of those campaigns will be dependent on longer-term success factors and impacts from social media, such as the number of connections you have made on Facebook or Twitter, the strength of your relationships with key bloggers or how many email addresses you have been able to collect.

The measurement criteria for social media activity need to operate over a variety of timescales, but from an understanding that you need to invest every day to create as strong a social media foundation as possible from which to launch an online campaign. 

This article was first published on prweek.com