This is an age of mass production. In the mass production of materials a broad technique has been developed and applied to their distribution. In this age, too, there must be a technique for the mass distribution of ideas. (Heath 2004 p. 78)¹¹

Insights

For years advertising agencies have had research teams that have helped prove the case as to why a campaign should be run and provided evidence of effectiveness. Part of the reason why this hadn't been a large-scale phenomenon in PR companies is one of scale and budget. In more recent times, insights have become the starting point of campaigns as they provide knowledge of the consumer and the campaign becomes a conversation opener.

The changes online put some more cost-effective tools in the hands of organizations.

Google Trends provides a visual indicator of level of search interest over time, shows hot spots of interest in a given country and allows comparisons against competitors to be run. It is a very good proxy to understand relative brand awareness and interest and how it has developed or ebbed over time. You can even get a rough idea of whether news stories have stimulated brand interest.

This works well for consumer brands, but is less beneficial for business-to-business orientated brands where the search volume and hence quality of the insights obtained is likely to be lower.

Both Google Trends and Google's Keyword Tool are different from most of the other tools available because the data is much more democratic.

Most tools look at what is being said in online conversations which, despite the social web, is still only a small sub-set of existing and prospective consumers for a product or service.

Whilst he was at Yahoo! Bradley Horowitz¹⁴ came up with a model of the relationship between consumption of social content and conversations using audience segmentation based around three segments: creators, synthesizers and consumers.

• Creators were 1% of the user base – these were the people that wrote blog posts or started threads in forums.
• Synthesizers were 10% of the user population – these were people who may retweet content or like a post.
• All of the user population would consume at least some content some of the time rather than be an originator.

Google Trends¹⁵ gives some insight into the last content consumer segment. This model generally holds true in terms of its proportions; the only exception would be social properties of luxury brands with a large aspirational fanbase like Burberry, which, by 2012, had a Facebook Page with over 14.5 million likes.¹⁶

There are a number of tools that help get an impression of online conversations. The three free ones I would recommend are Addict-o-matic,¹⁷ Social Mention,¹⁸ and WhosTalkin.¹⁹

Addict-o-matic provides a snapshot of blogs, Twitter posts, YouTube videos, Flickr and the latest news. It is useful to find out what is happening in near real time. Facebook and forums are two major gaps in the content snapshot that it provides.

Social Mention looks at similar channels to Addict-o-matic, but provides a bit more depth in terms of the content that it surfaces rather than just a snapshot. It provides metrics on how often the brand has been mentioned and the identity of the people who have discussed it most often recently. Probably the biggest advantage of Social Mention over Addict-o-matic is the ability to download this information as a CSV file which can be opened in a spreadsheet.

WhosTalkin manages to consolidate similar content but includes public Facebook posts and some content from forums; this is presented as a stream and has no analysis.

There are a number of paid-for tools that provide a greater depth of search for insights, different ways of segmenting the conversation and a greater level of analysis. Econsultancy reviews 15 vendors alone in its buyer's guide.²⁰ I have used and recommend Sysomos MAP²¹ because its fee is transparent and predictable, it trawls a large amount of online data and is able to provide access to data more than 30 days old.

Most of the major suppliers will provide a free trial and comprehensive demonstrations to allow you to see the power of their products. Test-drive a few to see what works for you.

Influence

Finding influencers for a PR campaign used to be relatively easy; there were a number of database services that listed out publications and their editorial teams. Their information was usually updated once a year or so, so the contacts were sometimes out of date, but it was easy to obtain replacement information.

Influence is now a bit different; influence can be more contextual although major media brands are still referenced and trusted. There are three broad measures that tend to be considered:

• Popularity – how many followers, Facebook likes or readers that a content creator has on their Twitter account, Facebook Page or website.
• Propagation – how much content is shared or retweeted.
• Mentions – how much a person or brand is mentioned by others.

Cha et al. (2010)²² found in their study of Twitter audiences that propagation was the most important attribute to influence.

There are a number of tools that look at measures of influence. Here are some of the ones that tend to rely more on propagation rather than popularity as a measure.

Edelman's TweetLevel²³ is a particularly good way of surfacing Twitter accounts that are influential around a subject over time, or assessing how relatively influential an account is with propagation being a major factor. Tweet­Level is unique in terms of the way Edelman has been transparent with its algorithms' composition. Twitalyzer²⁴ provides more of a rounded view on a given Twitter account. The service is available both as a seven-day free trial and a paid-for version.

Sysomos MAP turns out similar results to TweetLevel for the authority of tweets, with the additional advantage that the results can be downloaded as a file that can be opened in a spreadsheet program.

For blogs, the best analogue of propagation would be backlinks from other sites. Backlinks are the basis for a lot of search engine optimization (SEO) work; consequently there are a wide range of good quality tools out there to use.

Google has a specific command

link:insertyourdomainname

But, like most things that are free, it is limited in its utility. The results provided aren't exhaustive and the information can't be pulled into a spreadsheet easily.

A second Google command that provides a more holistic view is

link:insertyourdomainname

info:yourdomainname

This provides access to:

• Google's cache of the domain.
• A list of sites that Google considers similar to the domain mentioned in the link command.
• Where the domain links out too, which can provide a better guide to the blog's community than an author-curated blog roll.
• A means of containing the domain name as a term on page, to capture non-hyperlinked mentions.

Moz Open Site Explorer²⁵ provides a paid-for solution that is ideal for analyzing backlink data. Factors that are important include:

• The authority of the site domains.
• The variety of the site domains linking back.

Pulling this data into a spreadsheet allows these factors to be looked at more easily.

Whilst the current Facebook obsession is with likes, a better measure of propagation is shares of posts as a metric. Facebook's newsfeed algorithm governs how many people see those shares in their news feed, but despite this, shares are still the best analogue for propagation.

Messaging

Messaging is something that PRs generally work on with key decision-makers within an organization to reflect an image internally. As salespeople have long known, mirroring a recipient's language is a key way of building rapport.

With this in mind, Google's Keyword Tool²⁶ allows PRs to put some science into the process. By looking at local monthly search volume for key phrases within the messaging, Google can provide data to back up choices or suggest higher volume alternatives.

This won't prevent some PR faux-pas phrases like “leading end-to-end solutions provider”, or “best-of-breed solutions” appearing in the messages created but does put the PR on a stronger footing to provide a data-based rationale for choices made.