Review: Proving effective PR with big data

Who interprets the data?

According to Morgan, there is a heavy reliance in the PR industry on external media monitoring companies to analyse data.

She said the “real magic” happens when the data gathered from media monitoring companies is combined with other data such as internal reports, as assessing it takes time and understanding as “you need to get to know your data to be able to do it well.”

Many organisations are “still trying to interpret the wrong data,” explained Williams, which comes from a tendency to look at all of the clippings and coverage, which she said is irrelevant if it is not engaging or has no impact on the audience. 

Williams explained it has been demonstrated with client data at Metricomm that 80% of the results come from just 20% of media coverage obtained – and this is what is having an impact or changing audience behaviour. 

“I think that is the biggest problem with interpretation, we have to start with the right data,” said Williams.

“My advice for anyone leading a PR team would be to be brave and get hold of some data that actually gives a different perspective on the outcomes and the impact of the PR campaign,” she added.

Effectiveness of PR

PR is not very “linear” and there is no “silver bullet,” said Morgan, who explained it is different for each campaign and is dependent on what the objectives are and “where the brand is at on its journey and what it is trying to achieve.”

There are tools that can be utilised to prove PR is making a contribution to business results if you look at the data, added Williams. 

On proving the effectiveness of PR, she argued that “it is constantly evolving and I think we have to keep educating the industry about the ways in which it is possible.”

Not having a sufficient measurement function can be a “deal-breaker” for potential clients and it is a core requirement in every pitch as “it’s all about creativity and execution but they need to know you can deliver that,” said Morgan.

For Williams, the industry needs to find something that is more “meaningful” in terms of impact on the audience and it needs to start demanding its PR teams steer away from focusing solely on volume of coverage. 

It is about looking at the impact PR has on the engaged audience instead of counting the output or wrong metrics about impressions and reach, she added. 

Both panellists discussed the use of econometric analysis for clients but “econometric modelling” can be viewed as a barrier with costs and budgets taken into consideration, explained Morgan.

“People will buy in quite quickly,” once you start on a journey towards better measurement that is aligned with your objectives, she added.

Metricomm data analysis

At Metricomm, Williams explained the company has developed a method – by applying an algorithm based on academic research – to calculate the probability the audience will engage and be prompted to take action as a result of coverage they have seen or read. 

She emphasised that it is not just about “counting things” but about working out what it means in terms of behavioural impact. 

Introducing Metricomm’s data approach and analysis, Williams explained a report which aimed to use big data analysis to demonstrate similar results for the car market. 

There is a “very strong statistical relationship” between the size of the engaged audience created by online media coverage and search for the car brands, she said.

She explained with lower priced car brands, a larger audience is expected, whereas higher priced brands may generate a lot of search from online media coverage but with fewer resulting purchases.  

Williams pointed to Metricomm’s conclusions from this analysis, arguing that there must be a clear relationship between the engaged audience from online media coverage and advertising, which it hopes to explore in more detail. 

She added that it suggests day-to-day PR coverage for brands is also “amplifying the effect of advertising.”

She explained the connection between the size of the engaged audience from online media coverage and search as it “allows us to show us that media coverage is actually having a demonstrable effect on the top of the sales funnel.”

She demonstrated there is a measurable link “between the engaged audience, the people who read PR generated content, and the sales that are achieved for any given brand.”

There were a few key points PRs could note from the research, she said. Despite pointing to one of the car brands experiencing negative PR, it did not appear to impact on sales for the brand or have a long-term impact, she explained.

Discussing observations about the car brands, she said one brand achieved very similar levels of sales from a similar amount of media coverage in comparison to rival brands from a lower level of media coverage.

She said this provokes a number of interesting questions, such as the effectiveness of the brand advertising or “long-term residual reputation” accumulated from its PR activity. 

To summarise, she said the firm’s analysis illustrates the impact of online media is “increasingly short-term” whereas in print and broadcast it is potentially a different story.

Metricomm plan to put out a paper within the next few months around the issue. Its online report is available in full here.