PR Analytics: Using Public Relations Data Analytics for Effective PR Campaigns

One of the favorite quotes used by PR evaluation specialists is now well over 100 years old: ‘Half the money I spend on advertising is wasted; the trouble is I don’t know which half.’ It comes from John Wanamaker, a merchant and forefather of marketing and the problem he talks about is just as true today; a huge amount of activity is wasted. Grown-up PR analytics ends that problem.

Here at Sonus, we have quite possibly gone beyond any other tech PR agency in developing our own Intellectual Property and methodologies for measuring the impact of PR, so we have learned a thing or two about how PR analytics can guide you to more effective PR campaigns.

How has PR analytics changed?

When I started PR, over 20 years ago, the closest we got to PR analysis was to measure the number of ‘column inches’ that an article achieved in a publication and work out what that same amount of space in a publication would cost if it was bought as advertising – the Advertising Value Equivalent (AVE).

After that monetary value was worked out, a magic number was used to multiply that number, because “PR is more influential than advertising”, as all the PR Account Directors used to parrot. That magic number varied between each of the agencies I worked at (2 at one, 4 at another, 7 at another).

The final number was always bigger than the fee paid to the PR agency, thus ‘proving’ the value the PR agency provided.

Very clearly, this is unscientific in the extreme.

Thankfully AVE has, for well over a decade, been ridiculed. Despite that, a relatively recent PR Week study reported that over a third of PR agencies not only were using it, but were still willing to tell an industry journal that they still used it, which somewhat worrying.

Equally worrying is that Internet searches for “PR measurement” or “PR evaluation” haven’t increased in recent years, suggesting many agencies still don’t take it seriously.

A new scourge in Public Relations analytics

For those that have moved beyond AVE… the major metric used to analyze the effectiveness of campaigns has become ‘reach’.

Reach takes the number of visitors to a website in one month and applies that number to a single article. Again, it’s meaningless.

To take the raw number of, say, 1 billion, when an article appears for a client on the BBC website, is at best, meaningless and at worst, fraudulent. Because, very obviously, a billion people did not read that one article that featured a client, while ignoring the thousands of other articles on the BBC website.  Even calling it ‘potential reach’ does not improve the usefulness of the number by much.

I once worked on a campaign (before Sonus, of course) for the Giro d’Italia bike race that suggested more people than live on Earth had read articles about the race. Enough said.

A more evolved PR analytics tool

We do, of course, want to understand how many people read about our clients in a particular time period. That’s a key part of understanding the effectiveness of awareness campaigns for clients. But there must be a more evolved and nuanced methodology for determining that.

It is of course harder to determine how many people were likely to read a particular article, which is why most PR agencies don’t even think to try to work it out. Even the commercially available PR analytics systems aren’t particularly helpful in addressing this, which is partially why we created our own tools and systems. It has been hard work, and it has taken years, because we were starting from scratch, but it has been worth it.

Here is the starting point of our algorithm. As with ‘reach’ it starts with getting the figure for the monthly web traffic from a reliable source such as Similarweb. From there, you also need the data point of the number of articles appearing on a website in that month. And you need the number of social shares for that article versus the average number of social shares for that site. When you have all the above, you have the building blocks needed to algorithmically create a very realistic estimate of the number of times that an article was viewed.

Quite simply, we call this ‘estimated article views’.

Target audience analytics

We go beyond just a final number of estimated article views, because, frankly, it’s rare for a client to have a single target audience. There could be sales prospects in multiple sectors or across multiple horizontal job roles, in addition to potential recruits, investors, partners and other stakeholders. Each of those audiences has a very different set of media. So, lumping all those different target audiences together and coming up with a single number doesn’t elucidate much.

So, we need to dig down to understand how many times people from each target audience are likely to be reading articles about you, compared to your rivals.

That starts with gaining a clear picture of who your clients’ target audiences are, then using data science to determine which sites they visit. Apart from laying the foundation for a professional, well-targeted PR campaign, it also gives you the means to really understand the world from the perspective of your target audiences – which companies they’re reading more about, which messages they’re receiving and more.

From awareness to action

While awareness covers most of the expectations that clients have of PR agencies, it is possible to measure beyond this, through to actions taken by your stakeholders. This is a manageable feat now that nearly every business metric that matters can be tracked digitally. While we can connect via API to any data source that has an API, the most common data source we connect to, to visualize results, is Google Analytics, the de facto standard web analytics system.

By overlaying data about the outputs from a PR campaign (typically articles though it could be anything measurable) over data from Google Analytics, we can see clear correlations between our outputs and real-world results. Again, those results could be anything measurable via Google Analytics, including visits to a particular page or web form submissions or even sales, depending on what is important to the client.

A quick word about correlations. Although correlations are not the same as proof, they are often as close to proof as is possible to achieve in the digital world. With correlations, by measuring the spike in results and comparing the results to a previous time period, you are able to determine with a high degree of scientific rigor whether the spike in results was caused by a PR output or not.

What’s next for Public Relations data analytics?

At Sonus, we are pretty sure that we’ve conclusively answered the age-old mystery of understanding and visualizing results from a PR campaign. Inevitably, we will continue to optimize what we already have to make the data and visualizations clearer and more enlightening.

Our biggest challenge now is to bring along clients and prospects into making Public Relations analysis using reliable data a central part of their modus operandi. Our challenge is no longer the science, which we’ve pretty much nailed. Out challenge now is emotion and inertia. When you have been receiving the same report featuring ‘reach’ or AVE, or possibly worse, ‘gut feel’, for months or years, it takes bravery and confidence to admit the weaknesses of previous approaches and embrace something new.

It is time to end the vanity metrics, and start using more evolved metrics, not only to present visually appealing reports to executive teams, but to start making better decisions, based on solid data.

One final note, we don’t charge additional fees for our evaluation technology or methodology – we recognize that it’s a fundamental aspect of executing PR campaigns strategically – and we can never go back to the dark days of vanity metrics.