Get in touch with our team
Feature image for 20.04.2018


8 min read

How to Find PR Gems in Your Data Sets

This article was updated on: 07.02.2022

Data has always been a great source of potential PR and link worthy content. While data isn’t viable as the source for all of your link campaigns, it is still a very worthwhile part of your PR toolkit.

We’ve run some highly successful campaigns off the back of data, from huge data sets analysing the state of business across the UK which achieved placement in places like Forbes, Financial Times and Yahoo News, through to much smaller data pieces where simple visualisation of third party data achieved some valuable new links for a growing client’s site. Here, I’ll share some of my tips for gathering and analysing data for PR.

Step 1: Know what you want your data to say

Whether you are creating your own data or using that of a third party (check out for some awesome data sources), it’s important to at least have an idea of what you want your data to say before you even get started.

What do I mean by this? Essentially, you need to have your story in mind at the very beginning. What headline are you hoping to create? What hook will you be selling into the journalist/webmaster?

We worked with Company Check on their Business Census campaign, which was a data driven piece of content based on a survey of the site’s users. Before we even started gathering data or thought about sending out the survey questions, we first asked ourselves what stories we wanted to generate from that data, based on what was most likely to achieve coverage in our target publications.

In this case, we wanted to get into popular business press, and our research showed that anything which reviewed performance of business and looked ahead at potential challenges always went down well. In one year, Brexit was a popular business topic, so we also wanted to generate a story around that angle.

As such, as our survey questions were written so that, no matter how people answered, there was a story. For example, to generate a story on the biggest challenges facing British businesses in the UK, we provided multiple choice options, knowing that whichever option was most popular, our headline would be “X set to be biggest challenge to UK businesses in 2018”. For the Brexit angle, we asked “will Brexit negatively affect your business”, so either a ‘yes’ or ‘no’ answer would create a compelling news hook.

When we created the Enterprise Investment Scheme map with another client, we knew we wanted to show how funding was affecting businesses in different locations so that our headlines would be along the lines of “[City name] tops UK’s most invested cities”, so we looked for data that gave us that.

Whenever you’re planning a data based campaign, consider what you want story(ies) you want that data to tell before all else. This should, of course, be based on a solid understanding of your target publications/websites and what stories are most likely to be successful with them.

Step 2: Consider multipliers for more angles

Whenever we invest in a data based campaign, we accept that there’s going to be quite a bit of work involved in the analysis of that data and in the creation of stories accordingly, so we know it’s going to be a relatively large investment in terms of time.

For that reason, it’s always worth looking for as many additional angles to supplement your core data stories. This mitigates the risk, because it means you have far more opportunities to sell your content in, and a broader range of targets to whom it will likely appeal.

I like to use the following ‘multipliers’ fairly commonly on my campaigns:

  • Location
  • Age range
  • Sector/industry
  • Size of business

Of course, you will adapt those multipliers for the type of campaign you’re running, but essentially what they do is provide additional overlays to give your main stories a more niche appeal.

To use the Company Check angle again, as well as asking the respondents questions about their business challenges and so on, we had them tell us where their business was located, how many people worked for them, how long they had been in business for, and what industry they worked in. This meant every answer we had to each of our core questions could then be analysed a second, third, fourth… time to generate even more stories – e.g. we could now break down the biggest challenges faced by location, creating location specific stories relevant to the local press, and local trends relevant to nationals.

Step 3: Know how to ‘spin’ a story

Quite often, the word ‘spin’ is used to negative effect, but I don’t think it has to be. What it basically means is using your skills to communicate the story in the most effective way.

Let’s say your data shows that 15% of people spend more than £50 per month on skincare products, for example. “15% of people spends more than £50” doesn’t sound that exciting, because 15% as a proportion of 100% actually sounds really small.

Now, 100 divided by 15 is 6.666 – so let’s try saying “1 in 6 people spends more than £50 per month on skincare products”. OK, now we’re getting closer – one in six sounds like a much larger proportion, even though it’s exactly the same. It’s just saying it in a more compelling way.

So let’s have a look at that “£50 per month”. There are 12 months in a year, right? So 50 x 12 is £600 – now we’ve got “1 in 6 people to spend £600 this year alone on skincare products” – and bam! That’s a story! Now we’ve got Cosmo and Harper’s Bazaar and Good Housekeeping knocking on our door because we’ve got a compelling news story they want to share!

Here’s another example. We have some data that shows the number of hours people spend cleaning their homes per week:

0 hours/ I don’t do it myself/ I have a cleaner 45
1 – 4 hours 396
5 – 10 hours 265
11 – 15 hours 79
16+ hours 57

These numbers don’t really tell us much, so let’s make them into percentages instead (percentages are always more effective in story headlines that actual numbers, because they can be applied to larger populations):

0 hours/ I don’t do it myself/ I have a cleaner 45 5.34%
1 – 4 hours 396 47.03%
5 – 10 hours 265 31.47%
11 – 15 hours 79 9.38%
16+ hours 57 6.77%

The largest percentage here is the 1-4 hours bracket. “Half of us spend up to 4 hours per week cleaning” just doesn’t sound that exciting. Hmm.

OK, so 31% (nearly a third) spend up to 10 hours – that’s a bit more interesting, but still not that shocking. I’m not going to talk about that statistic in the pub with my friends. But what if we aggregated all the people spending more than 5 hours – so the bottom three rows? That also equals 47% – so nearly half. This could be more interesting – so “Half of us spend more than 5 hours per week cleaning” is a bit better.

Again though, let’s think about how we can ‘spin’ this, without affecting its truth. 5 hours per week, over 52 weeks per year is 260 hours. Now let’s go even bigger. Google tells me the average life expectancy in the UK is 80. Now, we’re not going to be cleaning our houses at this rate from the day we’re born, but bear with… what if we did aggregate that data for the average life span? 260 hours per year, multiplied by 80 years, is 20,800 hours in 80 years. Divide that by 24 to find the number of days, and it’s 866. Divide that by 365, and it’s 2.37.

So now, our story is “Nearly half of us spend over 2 years of our life cleaning!” – much more interesting, right?

Bonus tip: if using data for PR, use your data

This might sound obvious, but the purpose of using data for PR is to allow that data to show you the news story. There’s no point in gathering data if you’re then just going to run ahead with one idea that you probably could have had without it.

I like to thoroughly dissect my data, and I genuinely find it quite exciting, because I never know exactly what I’m going to find. I know what I’m expecting the data to show me, but it’s only when you really dig into the analysis that you can spot the really exciting opportunities, as well as the plethora of smaller opportunities that will become the bread and butter of your link building while you go for gold with those bigger features.

The way I do this is to spend time analysing every data set. So if it’s a survey, I will pull out each question individually and run the analysis by aggregating numbers and creating percentages, noting the stories I find as I go. I’ll also consider every permutation available to me through my multipliers. In the case of the Business Census, this meant 12 original questions resulted in 47 news ‘headlines’, each with its own target publication and bottom line results of more than 100 unique placements across the web. We’re working on a campaign right now where a very short survey has resulted in 14 unique story angles, and we’re excited to promote all of them.

Data is still very much a valid PR tool. Are you using it in your campaigns? I’d love to hear how you make more of your data – let me know in the comments below. In the meantime, also check out Paddy Moogan’s latest Whiteboard Friday for more useful sustainable link building tips.