Data Journalism a Practical Guide to Off Page SEO

So we have Ross Tavendale speaking at the Glasgow SEO Meetup event, and his topic was Data Journalism, a Practical guide to off page SEO along with some other tips about Content Ideation and various other bits and bobs including why a budget of £50k was spent and got no links as a result. You can watch the video above or read the transcribed version below.

Transcribed Version

Right, thank you very much Craig, that is fantastic. There’s nothing. So actually, that way we can protect the network, kind of thing.

I have only known Craig for a little period of time and every day our relationship grows stronger and the hatred also grows stronger.

So my name is Ross Tanvendale, I am the managing director of a company called Type A Media and like Kate says we’re, it’s a company based down in London Bridge. I’m actually originally from Glasgow and actually went to this university and dropped out of it about, 9-ish years ago. So it feels kind of weird to not to be in here a little bit tanked up listening to like Journeys ‘Don’t Stop Believing’ at this point but, today what we’re going to be talking about is data journalism.

So, a lot of the guys, can I just get a bit of an understanding of the room? Did anyone work for an agency here? A couple of them… Any hands? Business owners? All right, so it’s a reasonable mix, so this talk is going to be kind of in the middle, with regards to, what we’re going to be looking at. So like, day to day we build a lot of, kind of, high-end lines from places like The Guardian, the BBC, et cetera.

But it is not just a case of reaching out to them and asking for the length or maybe the content. Again Matt should have been through all of that with you…

Is there a clicker?

All right, I will just sort of stand here. There is not a ledge here, just ledge the hips.

It’s nice to be back in Glasgow.

Tools for Link Building outreach?

All right guys, so today what I am going to be talking about is how to actually make the content that gets into The Guardian and The Telegraph and the BBC, if you’re interested in the outreach parts or how to physically pitch people at skill, there’s a company called Pitchbox which are a kind of outreach CRM. I launched their blog for them with this particular post, it is a big long kind of 4000-word thing so, if you are interested or if you have got a team that is interested in that kind of skill each, then please do have a look.

All right so, without further ado, I am going to tell you a little bit of story. So, I want to take you back, it was 2012, we had just won our first client. A big, massive client, in fact, they ran an airline, and for us, that was a really, really big deal.

It was a brand new domain name and they were moving from one to another and the brief was we had to build links into this new domain so we can increase the rankings.

That wasn’t the interesting bit. The interesting bit was they were going to give me fifty grand per month to do that. Now I was a 22-year old account manager, so I’d been working in Edinburgh at the time, and so what’s the first thing you’re going to do when some guy gives you fifty grand to build links for him?

Technically, in a strategy session, that’s the first thing that we do. And as you’re gonna go through this, you’re like, “how did I even begin to spend fifty grand on links?” Keep buying them from Craig, but that would get me what, five thousand pounds deep, and then he would run out, perhaps?

How to spend big SEO budgets quickly?

So on the side of the agency, how do you spend the money really quickly?

Talk to a creative team, of course, and they tend to do something called an ‘ideation’ session.

Ideation Session?

Which essentially, what it looks like is people sitting on a bunch of beanbags and coming up with ideas. And, the lead creative guy which we’ll call the Commercial Director comes up with these ideas using their subjective opinion; no data, no logic, literally just their opinion.

To be fair to them, some ideas were actually pretty good. Someone sort of came up with this idea for an apple watch concept. So if you went to, let’s say, the Eiffel tower, it would look at your heart rate, and it would guess, depending on your biomarkers, how exciting that particular place was. It was kind of link trip adviser meets biomarkers and like, I could do some PR on that- It’s pretty cool.

Typically there’s always some old guy in the room who’s gonna talk about key articles or something he’s just read in the press that day.

So after a couple of days of what they call creative ideation, this is what we came up with.

So this particular content perhaps took about twenty thousand pounds of that fifty thousand pounds in about three weeks. Does anybody like to guess what we’ve done?

Infographic! Points to that man, exactly. It was a very nice infographic… Or at least our Commercial Director was inclined to think it was a very nice infographic, some would say phenomenal.

So we essentially, we went and pitched it to the client, and they used their subjective opinion to pick one of three designs and one of three concepts. Like, “yeah, we’ll take that one.”

The interesting thing about the client, their job is to maintain a profit and loss account, not to deal with creative, not to deal with marketing, not to deal with brands, but they were the ones calling the shot on the ideas the agency was pushing, which felt a little bit strange.

Alright. Amazing things started to happen when we pushed this fifty grand infographic out.

How many links can you get from an Infographic?

A show of hands, how many people think we got a hundred links?

fuck you guys, we’re not that bad.

Um, fifty links? Great. Less than 10?

We actually got zero links. All right. You’ve got two links, actually. We got one from my personal blog, and one from my client’s personal blog, as we’d spent all of the money on the gig and absolutely nothing on the outreach.

There’s a couple of other reasons why it failed. It wasn’t objective, there’s literally a bunch of guys in there making some decisions to take to some more guys and then to make some more decisions and then went

“Yeah, fine, we’ll just go with that.”

There are zero frameworks, so the decision-making was literally just how we felt that particular day, and it was through the lens of the people in the room, so this particular one was about travel in Saudi Arabia. There was no one who’d ever been to Saudi Arabia in their life the room so we can understand why that didn’t work.

Another one is, they had no concept flop risk. So technically, if you look at this graph, Flop Risk is inadvertently correlated to Algorithmic Risk. So for example, if you’re building VPN’s, the Algorithmic Risk is typically much higher in that if they’re discovered, then the risk of that being penalized is relatively high. However, if I’m building links on the BBC with Telegraph, I yet to see someone get featured in the press and be penalized. However, the flop risk associated with doing PR is considerably higher; they may like it, they may not like it. We need to think about how we mitigate the risk of doing that PR if some clients make some money on it, they actually get the result.



We have a methodology called RAPTOR. They can see people that are probably quite happy who’d be using these little daft segments and stuff like that, everyone else was just secular lengths. RAPTOR essentially stands for Research Angles Pitch. That’s the first part we’re gonna have a look at and I’m gonna take you through how we actually do all this up.

So we have something called a RAP sheet, so technically if you come up with an idea, we need you to write a bit of it down and tell me why it’s a good idea and who’s going to physically pick it up. You’d be very surprised how quickly your ideas just go in the bin, and you can’t come up with one reason why it’s gonna do well, and some people are gonna take the actual work.

The RAP sheet allows you to do three angles and three headlines, which means you get nine total pictures. When we’re talking about de-risking a campaign and reducing flop-risk, it’s about 11% in total you need to hit. Usually, with a point to point outreach, you will email someone and you will get a link, or you will not. But with this style of PR typically need to hit 11% of the time, which is a much better odd. So, if the client’s putting a bunch of money in, the likely hood of you getting it is much much higher.

We have something called, ideation which, again, the agency people are like, “yes! We love ideation.” And everyone else is a little bit nervous when I say that. Essentially, what we do is start with the audience first.

Facebook Ads

I take it we’re all on Facebook? If you’re doing any ads on Facebook, I’d encourage you to go to Facebook Audience Insights first and foremost and start putting your brand name in, or your keywords in if your page isn’t big enough. And that’s gonna start spitting out some interesting bits about your audience.

So your next word, ‘casino’, is there any men in the room that play casino games? Are you engaged? Oh, you’re married. So when we asked Facebook Insights, men who play casino games, typically are 93% more likely to be engaged than the rest of the Facebook population. However, they are 58% less likely to be married, which we thought was a really weird outlying piece of data that we could start using as part of the outreach.

Another thing about people who enjoy casino games, they’re really into investment banking and they love a cheeky Nandos because that was one of the things that came important. They smell like a teenager because they wear lynx quite a bit and they also shop in JD Sports. If any of you play casino games it’s fucking hilarious how much it will be psyching you right now.

They’re a safe kind of edgy, because they like bands like Tinchy Stryder and we think a public figure is the dance band Diversity. Fantastic. So, we can get an idea of who these people are. ‘The Legend of Scott’ is one of the major pages they follow so they think they’re a pretty big deal.

Are you starting to get a picture of who this person is and the type of content they might enjoy? So when we start building that framework and start making that content, this is the exact same thing the newspapers are doing, so all we’re doing is matching these two things together.

Ideation Sessions

So, in these ideation sessions, we make these assumptive statements so I want to know this person is this thing, this person is not this thing. And if you look a little bit closer on this we’re making statements like this is a female, 30-40, into Hobbycraft, value-shopper from this bit of the world. And through that lens, when we’re coming up with ideas we can actually start to work out if the content will work for them or will it not work for them. And, we’re pitching journalists, if the content that journalist’s making does not fit the original person, we know not to pitch them.

All right, so here’s the framework. So, if you were to look on the Daily Mail website right now, you can see that every single headline uses this framework. So we have a client called Unilad, last year they were the biggest Facebook page in the world, they’re the most engaged Facebook page in the world. 80 to 100 million clicks each per month, and because of that we can have a look at that level of data and we can start seeing some patterns emerging.

And the patterns we see is, with this free lump of content, we can see people targeting people’s identity, something utilitarian, or something that plays on social information.

So, what does that actually mean? Identity is essentially a piece of content which describes you better than you can describe yourself.

So, here’s our classic example of that. Anyone who’s left-handed in the room? It’s hard out there, for a lefty. Being left-handed is an identity factor. Got it picked up in Harper’s Bizarre. Real easy. Why you shouldn’t ever… mess with gingers.

Any gingers in the room? In the corner. Terrifying. And it says, more ginger men are having sex thanks to Ed Sheeran. We use the identity piece as well as the social information piece. Social information piece, we pick on ginger people at school and therefore we don’t think they’re gonna have as much sex therefore we break, the fact that you think that, we break your social information and we hit against it. We hit two in that particular one.

Another one is people with ginger hair have genetic superpowers according to scientists. Again, using identity and breaking social information.

All right, so next up its utility. So, utility is something that’s physically useful to someone. So, if you’ve ever tried to get a mortgage online, you’ll see mortgage calculators, credit card calculators, things of that nature.


We can do it really simply with content… anyone ever tried to lose weight? Keep trying… no, I’m joking. What do loads of calories look like in fast food? Really nice. We have done it for alcohol as well. Again, we get someone who’s, say a dermatologist, and we get them to write the alcohol that’s best and worst for your skin, package that up nicely for a particular journalist that wants it and, before you know it, we have the links.

In that, we have these kinds of bubbles you live in on social media where if you click left revenue you’ll get more left revenues, if you click right revenues you’ll get more right revenues. And you just live in your bubble of consuming the same thing over and over again. So, this particular style of PR, where you’re trying to confirm social information for the person or we’re trying to break it.

Does London have the most miserable people according to stats?

Londoners are found to be the most miserable people in the UK. So, let’s break this down for a second, so what is one thing we know about London? The house prices are very expensive and there’s a bunch of money floating around there. So, if they’re the most miserable, that confirms our social information. Who here’s been to London and had a horrible time? Anyone? Really? So, the moment you step onto the tube you’re like, “these people are horrible.” So, that actually confirms social information. But that will only work with people who are in Yorkshire. This is a Yorkshire based magazine, so that’s why that works. So, we pitch those sort of stories, to the North.

Allergies are on the rise globally. This is an example of just using data which sits around to make something interesting. The way to reframe that, the Daily Mail published one that ‘snowflake millennials are hypochondriacs – study reveals’. Because, when we cut this by the base type, someone with mobile phones and someone under 25. We can package that up and give it to the Daily Mail with the line which they actually enjoy using and they actually publish it for us.

All right, so, the whole point of this methodology is the content is finding the audience and not the other way around. So, typically you’ll meet something and then you’ll push it. What we’re actually doing is getting the audience involved in the actual push itself, so the fact that it’s about ginger people, or left-handed people, or men, or women, or whatever it is, that actually finds them, not the other way around.

The framework is really really basic, so the next time you go to make a piece of content, run it through this and see if it makes sense to you. Or literally, just open the Daily Mail app and see if they’re hitting gender, or age, or location. Location’s a great one, we all kinda bind together as like Scottish people for example. That’s an identity factor that we all share, so the moment there’s any type of Scottish news, we all kinda rally around it.

When you’re making a headline, they need to have a level of Drill down-ability. So, what I mean by that is, if I’m going to Nationals, I want it to be big, and broad, and talking about lots of different data points, and I want it to drill down into a niche  publication, and I want it to hit a blog and then I want it to hit a hobby or something. So, the dat you’re putting together, as soon as you’re putting together the ultimate, we need to have this level so that I can just re-frame it depending on the person I’m actually talking to.

Where to get data?

In terms of the data sources, there’s 3 in total: Private data from Hyperledger and CRM, Public data and from government databases, and potential data from surveys and pro commentaries. When someone comes up to you in the street with a survey, what’s your immediate response? It’s thank you, but fuck off, I’m busy, please don’t talk to me. So, that’s actually translated into media. Every single time that we’ve tried to do something like a Google Survey, or something like that, to get links or to get coverage, only tabloids will cover it. You’ll literally never get… well unless the survey is phenomenal with hundreds of thousands of respondents, that I would not try and push that into any major ones.

All right, so what do you do with the private data? Is anyone here from East Kilbride? I’m really happy that I can actually ask that question, someone might actually be from East Kilbride… Anyone?

Speaker 3:
I work there.

You work in East Kilbride? And you’ve made it out alive, congratulations sir.

So, we had a recruitment client and we pulled all their data and their analytics and we have seen that the most amount of jobs for young tradesmen was in… East Kilbride. But, we think about the social information we have in our head about East Kilbride and any old industrial town, we think that there’s not a lot of opportunities there, you’ve got to move into a bigger city. So, we’re breaking social information and playing on the identity of people from East Kilbride.

That lovely woman in the bottom corner is the MP for East Kilbride. One thing I’d revise, if you’ve got a local MP, try and strengthen it. If you’ve got good news, go to the local MP, get them to actually comment on it. She actually raised it in parliament, she was like, ‘This amazing study… blah blah blah.’ And we used that in a substitute pressure to keep getting likes and keep this story moving and keep this story going.

Mash it up with your own data sources. So, we have a client who’s a casino and they have a lead sports team and we have all their salary data. So, there are kids who are playing FIFA, at home, who are getting paid more than 75% of the English Premiership. So, literally, there’s a guy who plays FIFA who earns more than Ronaldo.

Not a joke. And that’s quite interesting in itself, right? But then, you split it by gender and it’s the exact same thing we’re putting this pout, all the gender pay gap information came out with the government, and then the headline becomes, ‘100% of all female football players of every existence, gets paid less than their teammates son, sitting in his bedroom playing FIFA.’ Which isn’t 100% the case, but that particular story, and that wrapper hits identity, because it hits gender, and it hits social information, because we all kinda know that to be true.

All right, public data. I love this one. Anyone ever flown Ryanair? Keep your hand up if you’re happy with the experience. One. All right, fair enough, men representing. Typically, it’s a bad experience. So, Ryanair cancelled half a million flights, I think this time last year. My Grandmother was due to come down and see me from Glasgow, coming down to London. They cancelled her flight, and I was just thinking of her just doing the whole control F thing. She’s no control F lass, she can’t find her flight. The way Ryanair told everyone was a published, half-a-million flight record in pdf. So you’re scrolling. She didn’t have a clue, she was scrolling, she was not coming down.

So, we took their data and turned into a tool so you could put your flight number in press the button, tell if it’s cancelled or not and then get a refund. Put in a clients site, found some journalists who don’t like Ryanair… there’s a lot of them, and then we pushed it to them and it went mega-viral, beyond belief.

Slice your data differently, so the government released all this gender pay stuff and the way they released it was a flat spreadsheet. So, they released 100,000 data records from the biggest companies in the UK, and to actually drill into it and make anything meaningful from it was brutal unless you’re a data scientist. Thankfully, we have endless data scientists so we started slicing it differently, we allowed it so we could compare universities- so we could look at all the top universities, compare them. NHS trusts, that was interesting because people identify as people who like the NHS, if you’re a nurse you identify as a nurse. They then cut it by a region in NHS trusts, so the list just keeps going and going and going because we need to have big data sets which have drill down-ability.

Where to find public data?

How do you find your public data? For anyone who’s ever done any guest-posting stuff, you’ll typically do in URL, write for us, and then a keyword and then do that at scale to find lots of guest-posting opportunities. You do the exact same thing to find free data resources, in particular from universities and the government. If you use .ac.uk or .gov.uk it’ll link down every single pdf the government have and if you put a keyword next to it, it’ll pull down the keyword targeted data that you want. Or, to do yourselves a massive favour, sometimes people Google set search where you can actually put in any keyword and it’ll return the data set that you want, related to that particular keyword.

So, they’ve used potential data. Has anyone ever done a freedom of information request? Oh, there we go… for PR purposes, or because you’re a bit dodgy or what?

Just for a laugh. He scoped out freedom of information request… for a laugh. Fantastic.

So, we were looking at all the different programs up and down the UK and we asked St John’s Ambulance for a freedom of information request, they gave us all the heart attacks that have happened at the grounds up and down the years so we could work out the most exciting football ground. Cut it at different angles or whatever.

And they gave it all to us, the real genius of getting a freedom of information request is when you package up the press release, just forward on the email from the government with your press release on top of it and the email chain is there for the journalist to see. So it’s instantly validated from data that didn’t actually exist before, that came from the government but you’ve put together perfectly for them.

This is my favourite one, and also it’s a two-week turnaround, so they need to deliver it to you in two weeks.


Google Surveys for Collecting Data

Google surveys, typically we don’t do a lot of this, this is kinda the last gasp if we’ve not got any decent data. You can do a Google survey for a couple of hundred pounds, get 10,000 or so people to answer questions and you get a bunch of demographic data. The reason why we like this is because it cuts by region, so if you’re doing it in the USA, you’re gonna get all 52 states and then you’re gonna get all the regions and then you’re gonna get all the cities and you get all the towns inside the cities, there are about 1800 total publications you can then pitch absolutely perfect data. Just for that journalist. So, when it comes to scaling this allows you to do that relatively easily.

So that’s everything for me, you’re never gonna have success with this approach but you will get repeatable and offendable results and hopefully, you’ll be able to build once you have decent links.

Thank you very much.


seo profile image

Craig Campbell

I am a Glasgow based SEO expert who has been doing SEO for 18 years.

  • social media icon
  • social media icon
  • social media icon

Online Courses