#13 Controversy of Coverage Guarantees with Paul Stollery

Steve:
Lou is back from maternity leave. Hello, Lou.

Louise:
Hi, Steve.

Steve:
Which means we’re back for a third season of the Digital PR Podcast. It’s fair to say we’re extremely excited about this series. We’ll leave you to be the judges of whether it’s bigger and better than ever. But what we can guarantee is a fantastic lineup of wonderful guests and talking points, such as the rise of TikTok search, the role of brand in digital PR, and how agencies and practitioners can be thinking about and using generative AI in their day to day. It’s a bumper schedule, so let’s dive straight into season three of the Propellernet digital PR podcast. Can you ever guarantee coverage in PR? It’s a big question and one that has got quite heated responses on LinkedIn and the like over the last few months. You’d once be directed to the advertorial department if you brought it up, but recently there have been more agencies that are happily offering it. One such agency is the brilliant hard numbers. And today we’re speaking to co-founder Paul Stollery, all about guarantees and measurement in PR.

Louise:
Hi, Paul. Thank you so much for joining us on the podcast. Would you be able to give the listeners a little bit of an overview of you and your career to date, please?

Paul:
Yeah, of course. Firstly, thanks ever so much for having me on. I’m a huge fan of the podcast and was really pleased to be invited. So I started my career in SEO, in the early days of SEO when that acronym was sort of new and exciting. I moved into PR about 10, 10 or 11 years ago. I was drawn to the slightly more creative and campaign-driven nature of the discipline. At the time I think SEO mainly in the guise of Digital PR has really caught up a lot in that time, but certainly 10, 11 years ago, there’s a bit more scope for creative and campaigns and things like that in media relations. Worked in PR for a couple of years, then took a couple of years working freelance when I set up a youth marketing agency called Hype Collective, and ran that for a couple of years, sold that, and then started Hard Numbers in 2020. Tech PR agency operates in the overlap between PR and sales, big focus on measurement, and that’s what I’m doing today.

Steve:
You’ve had an amazing career path so far. Thanks for taking us and the listeners through that. I want to take you back to 2020, which you just mentioned when you first set up Hard Numbers. What a time to launch a new agency. Obviously a lot going on that year. Can you tell us about that experience? What was that like? Hugely traumatic.

Paul:
Now it was actually a really interesting time. Let me step back to just before all of that went down. So I’ve been chatting to Daryl, who I’d worked with, Daryl Sperry, I’d worked with him at Hotwire for a couple of years. And we spotted what we thought was a gap in the PR agency market, and that was really an agency that was, yes, PR, media relations first and foremost, but one that was really comfortable with things like sales, measurements, demonstrable return on investment. So we’ve been talking about setting up an agency from late 2019, initially over a couple of beers, then January 2020 is when we started getting serious, you know, talking about paperwork terms, all that sort of stuff. And then I believe it was the end of, it was either the end of February or the end of March, he handed in his notice on his very stable, very well-paying job at Hotwire. And then it was literally sort of a week or maybe 10 days later that COVID arrived on the UK shores. And it was sort of the week where I think all brands realised, and agencies as well, that it wasn’t just going to be another sort of new story that was happening somewhere else. It was really going to be something that turned the country and everything else upside down. But we’d already kind of started, so we thought, you know, we kind of pushed on. I think it was partially out of optimism, partially out of stubbornness, and we’re still here today. So yeah.

Steve:
An amazing, amazing success story. Like you have the utmost respect of me, Lou, and I think everyone else in the industry for like pressing ahead with that and creating such a brilliant agency out of it because it was such a tough year. I mean, Lou, you’d returned to Propellernet that year after a brief hiatus and me and you were trying to make sense of what was going on. So the fact that you’d spotted this gap in the market and persevered with it, absolutely astonishing. Brilliant.

Louise:
Yeah, what a time. Why was it so important for you to start this agency that had quite a different proposition to your kind of peers?

Paul:
Lots of reasons, but they broadly boil down to two buckets. Firstly, one looking at us, our commercial model. Secondly, one looking at our audience. Firstly, you know, on looking inwards, there are 4,000 other businesses in the UK that are registered where PR is the primary thing they do. I think it’s around 700 or 800 of them are VAT registered, so therefore they probably do over £100,000 a year in terms of revenue. So probably sort of, you know, 700, 800 quote-unquote agencies out there, which means the market is incredibly saturated on choice. We wanted to make sure that we stood out. We wanted to make sure that there was something where when somebody asked a question, do you know an agency that can do this? We’re either top of mind or top three. And that’s really the key proposition is you need to do something that where that question is going to be asked, so therefore it’s broad enough, but you need to be narrow enough where you become top of mind. It’s almost a contradiction in those two things. And having worked with Daryl for a couple of years, I worked at the Creative Hotwire. Daryl was in a commercial role, so worked a lot of pitches together. We asked all the prospects that we spoke to, First question, why are you spending money on PR? And in tech especially, especially earlier stage, you know, series A, series B, the answer is almost always some form of growth, leads, sales, more ad cover for my sales team in pitch meetings. And it was funny, a lot of PR agencies are really uncomfortable with that area, so we saw that as a real opportunity. So that’s the first one, looking inwards, more looking at our audience, more looking at our sort of prospects and our clients. The reason we chose to, yes, focus on sales and commercial growth, but then also measurement within that, is fundamentally, you know, the client doesn’t care how long a byline took to write. They just care that it’s written and it’s out there and it’s driving impact. So we decided part of our proposition was going to be moving a little bit further along the sort of measurement framework food chain. Rather than talking about hours, you know, you work with an agency, five, eight grand a month, you get X number of hours, we’ll talk about outputs instead.

Louise:
It’s often something that people talk about as, I don’t want to say the dream, that’s probably like the idea of selling the output rather than the time, but it’s the confidence to go out and feel like it’s a, I guess, well, it’s going against the grain slightly and feeling like, you know, you’re going to have a successful business and it’s going to make sense financially as well as for the clients and being appealing to them.

Steve:
How exactly does it work with Hard Numbers? For complete transparency at Propellernet, you know, we’re used to and still do to a certain extent work with, you know, linked targets and things like that. But what you’re offering is slightly different. You’re selling the output and kind of, you know, a guarantee of what you’re going to provide. So how exactly does it work? What metrics are included within that?

Paul:
We will look at a client’s news pipeline. We’ll then look at the work that we’re proposing for them. We’ll tell them how much coverage we think we can generate, often within sort of a defined media list or, you know, define that in some way. And then we’ll write into our contracts. If we fail to hit that target on earned media coverage, we’ll work for free until we do.

Steve:
There’s obviously a huge benefit to that, a huge benefit for the client and for you guys as a selling point. Because, yeah, honest truth, if I was a company that was looking for PR, I would be really, really attracted to that because it’s like, you know, you’re sort of taking out the risk, I suppose, a bit for them. What, though, are the kind of the main pitfalls of that? Do you ever get into a situation, and I’m hoping you don’t mind us asking, but do you ever get into a situation where you are working a lot of the time for free? Because we know in PR not everything works all of the time, no matter how skilled or brilliant you are.

Paul:
We’ve never actually had to work for free. And the way we’ve done that is by always hitting our KPIs. So they’re written in a rolling, it’s a rolling quarterly basis. So we’ll install a culture at the agency that, yes, we have our agreed targets with our clients. We’ll aim to hit them, exceed them every month. Every now and then we’ll drop a little bit short, but the key thing is if on a quarterly basis we’re at or above them, the guarantee isn’t triggered. I think we’ve worked with, I don’t know off the top of my head, maybe 80 or 100 clients since launching and we’ve never had to engage that mechanism of the contract. We don’t end up working for free very often. I think one of the perceived pitfalls of offering some sort of an earned media guarantee is this notion that it’s impossible to predict coverage. And so there’s a grain of truth to that. Yes, it is impossible to take a single release and say it’s definitely going to be in a single publication. It’s earned media. There is an element of a lack of control to it. But it’s perfectly possible to do it with a full pipeline, with a news pipeline. If you’re working across a press office, for instance, and there’s multiple tactical levers you’re pulling in order to hit your results, you know, or you should know, roughly how many times a byline is going to be picked up based on how strong the angle is and the topics that it covers. You should know whether a campaign is good for the nationals or whether you’re going to struggle to take it out of the trades. You should know whether that campaign has a regional element to it and which regions it’s relevant to. You should know fundamentally whether a news release is newsworthy and whether it’s newsworthy to the degree of, you know, two or three hits in the trades, or whether it’s, you know, gonna blow the doors off. And when you average all of that out, you should be able to say roughly how much coverage that’ll generate in key titles. And the idea that it’s impossible to predict is complete fallacy. I think it’s actually one that’s really damaging for our industry. I think a lot of PR, agencies walk around saying to their clients, you know, in media coverage, they almost talk about it like it’s some sort of, you know, dark arts, magic, no one can predict what it is. But in reality, they’re really simple, you know, really simple based on knowledge, based on what’s going to get covered in the media, really simple metrics you can put in, where you can say, right, well, if we do this, we’ll get roughly this. And it’s obviously up to agencies whether they then choose to put that in a contract. But I think the core idea that’s impossible to predict is one that I think is really damaging in the boardroom.

Louise:
When you’re having your conversations with your clients around the target or the guarantee that you’re going to agree on, do you have a different idea to what your client has? Is there a negotiation really, like where they might be saying like, oh, I think we could up that number and you’re like, no, realistically, I think we should have it at this number and you’re kind of finding that happy medium where obviously you’re not just putting in an easy target, but you’re doing something which you think you can achieve. Like, do you find that it’s a kind of negotiation to find that happy spot for both of you?

Paul:
More often than not, the comment we’ll get is it’s great that you’re actually putting numbers in there, no one else has told us how much we’re getting. So we’re almost coming from quite a low level in terms of expectations. Sometimes it does. Sometimes people turn around and say, no, you know, six piece of coverage a month isn’t a good ROI. We need 25 pieces of coverage. That’s a really good thing because you want to know that before you start working with a client, not after, because after two things are going to happen. Firstly, the client’s going to spend money on a PR retainer and they’re going to think they’re getting one thing, whereas actually they’re going to get something else. Secondly, you’re probably going to get quite a difficult relationship. You’re probably going to invest a lot of resource into building that relationship, getting that account off the ground. Only when six weeks goes by and the first sort of monthly report goes over, the client is suddenly unhappy. So normally we, you know, putting those numbers in, we say what’s realistic and, you know, normally it’s a positive conversation. But we’d always prefer to have a difficult conversation about that at the start rather than one, three, six months in.

Steve:
I mean that makes perfect sense and I think is a really good lesson for listeners whether they’re starting out their career or you know more senior like myself and Lou just to remind yourself of really always have those difficult conversations. I love the way you think about it as well Paul, the coverage guarantee because you’re absolutely right in that we will pitch for business and we will say to clients this is kind of roughly what we might expect or where we would expect the story to land. You’re right in that it’s completely up to the agency whether they then put that in a contract. You have chosen to and that makes it really nice and clear and that’s part of your business model. So I think you’re right. It’s much easier for PRs to go. Oh, yeah, it can’t be guaranteed. It’s, you know, as you say, almost treating it like a bit like a dark art because it benefits them if anything does go wrong. Have you ever rethought the guarantee or the way you do things or the way you offer or what you offer your clients based on recent changes in the media landscape? So, you know, less journalists, more redundancies. It seems like it’s, you know, getting harder and harder to get kind of cut through and coverage out there. Or is that just not something you’ve encountered or really thought about?

Paul:
Trends in, you know, number of media titles, column entries available, journalists to pitch to, the trend of that shrinking I think makes the guarantee more important than ever. It is really hard to generate earned media coverage and it’s getting harder, but we bill our clients five, maybe ten grand a month and our position is always, you know, we should be able to do hard things in return for that. I suspect as it gets harder, and I think it will continue to get harder, I think that’s one of the biggest threats facing the PR industry, we’ll see a greater demand for things like guarantees because the harder it gets, the more PR agencies are going to promise to do things and then fail to deliver. And every client who has been motivated by our guarantee, they all have one thing in common, and that’s that they’ve hired a PR agency in the past who said we’ll do X, Y, Z in the pitch, and they’ve been let down when the relationship actually starts. So I think the harder the media landscape gets, the more important things like guarantees will become.

Steve:
You’ve preempted another question I had, which was how many new business prospects or clients come to you that have been kind of burnt in the past? You know, we find it’s quite often, you know, that people will have a bad experience of an agency or of kind of PR or comms work in general. So that obviously is true for you guys as well.

Paul:
I’d really, and this is something that maybe being called hard numbers we should have to hand, but I’d love to know how many of our clients have been burnt in the past. It definitely feels like somewhere between, you know, between a third and or maybe even just over a half. But what I can say is every time someone comments on their earned media guarantee from a client point of view, It’s really controversial with agencies, but it’s really popular with clients. Every time a client comments on that, the story’s always the same. They’ll say, oh, I hired an agency, they said they’d do this, and they never did. In fact, the first client we ever won, going back to those difficult times, who is still a client to this day, four and a half years later. We, as a first client, really important, obviously, really, really delicate economic times we were in. So this client signing on was a real, you know, really, really huge boost for us. We found out two weeks in that the CEO had appointed us, but the largest shareholder was against appointing us. And they weren’t just against appointing us, they were against appointing any other PR agency, because they had appointed three different PR agencies in the past, and for whatever reason, they felt they’d been let down every time. That investor, six weeks in, Not only did we change his mind, another one of the companies that he was invested in then became our client number two. And those guys are still a client four and a half years later. So yeah, it’s absolutely, well, I don’t know exactly how many of our clients have been let down. My hunch is that it’s quite a lot. I think that’s probably true across the industry.

Louise:
I think we mentioned in our introduction how this might be seen as controversial, but you’re so right. There’s no client I’ve ever seen on any LinkedIn post complaining about a coverage guarantee. It’s more from an industry point of view in terms of, well, I guess it makes people look at themselves and it makes people look at their offering and think, it’s very different, so you can see why it kind of ruffles some feathers on that when it comes to the critics. And obviously these aren’t people necessarily criticising you. I have seen other agencies that talk about coverage guarantees or talk about link guarantees. So I’m not saying this is all just being aimed at yourselves or anything, which it isn’t. But there have been people who say that coverage guarantees can lead to low quality coverage or low quality links. I think I may have seen the term race to the bottom being used when it comes to talking about this. But also on the flip side, and you have touched on this, so I think we can preempt your answer on this as well, but in general, talking about coverage guarantees, there is a concern that that working for free will mean that the people working within the teams are going to be kind of overworked. What’s your response to those points?

Steve:
She’s gone full Jeremy Paxman for the first time.

Louise:
I quite like that. Season three.

Paul:
I love it. Go on. Seriously, I do actually think you’re quite kind. I do think we do get, or it feels like it, maybe it’s just from the inside that we do get a lot of criticism. That’s fine. But we’re absolutely fine with us. Also, weirdly, it seems to be we get criticised by lots of American PR practitioners. And generally, most Brits who comment are like, but yeah, just kind of, you know, we don’t do that. But it kind of makes sense. But there certainly does seem to be a lot of controversy over it and the idea is that you could use this media guarantee, the idea of putting media output into a contract as a sort of a way to, you know, gain a system, just deliver low quality coverage and, you know, jobs done, on you go. The thing I never get around that argument is the most common metric to put into contracts between PR agencies and clients is hours. And you can make the exact same argument probably more so with hours than you could with media coverage. You could put, I don’t know, 30 hours or 50 hours into a contract, and the argument normally goes when it’s media coverage. Well, yeah, you can get five bits of coverage a month, but who knows where those, you know, bits of coverage are going to land. And yeah, you can put 30 hours or 50 hours into a contract and you can sit around all month with your thumb in your ass looking at, you know, media coverage reports and doing stuff that doesn’t drive any value at all. But you don’t do that when hours are in the contract, because you know if you don’t drive value for your client, eventually they’re going to cotton on and they’re going to fire you. The idea of media outputs, or earned media coverage in a contract, is it’s more about having a promise for our client saying, look, in the pitch, we’re going to tell you that we’re going to do these things. We’re going to achieve it. And if we don’t, you’re going to have some sort of mechanic that you can pull in order to make sure you get your value out of us.

Steve:
That makes perfect sense to me. And I think he’s very well argued with the hours versus coverage output. Why would a client, and again, just sort of carrying on the Jeremy Paxman theme, why would a client ever leave you, ever leave hard numbers? Because if you’re guaranteeing coverage, you know, it’s like if I was a client, you’d just be like, well, okay. Like this is, this is fantastic. Like what, what are the reasons if it happens at all?

Paul:
We can leave the podcast right there, because I think that’s a fantastic line… No, look, what would happen, look, a lot of the time, so there’s a couple different things. When you’re working in tech, very often you are growing a business that doesn’t have a blueprint for growth, right? So if you work with big blue chip companies, you kind of, you know, yes, okay, the media landscape’s changed. But this product has existed for 20, 30, maybe 100 years. The audience has existed. The triggers of the audience, you know, again, are ever adapting. But broadly, you know, you can learn from past experience if you get this much coverage, if you get this many adverts, if you change behaviours this way, you eventually drive through the profit. A lot of the companies we’re working in, we’re working for, sorry, they’re kind of breaking fresh ground. So they’re kind of thinking what normally happens with a client, this is a real generalisation, but you start with a new product. You figure out whether you can get market fit. If you get market fit, initially you start spending money on paid. You grow it to an area where maybe you’ve got a hundred clients. But those hundred clients, a hundred customers, are maybe the very core of that audience or bullseye, if you think about it. Your challenge really starts when you’ve exhausted that core audience and you need to grow that audience so you can continue to convert people at a cost-effective manner using paid media and things like that. So then you’ve got to start building awareness, retention, change people’s perceptions of you, and there’s loads of different ways in which you can do that. The media landscape is really fragmented now. So very simply, the most common reason that a client will leave us is they will try media relations and they’ll say, actually, while we’ve got the media outputs that we thought we needed, the actual media outputs we need, now that we’ve tested all this, has changed slightly. So then you go through, you iterate, and you go, maybe media relations at X grand a month isn’t what we need to do. Maybe we need to invest in influencer, some other forward pay, et cetera, et cetera.

Steve:
Thank you for answering, because we went off to let the listeners behind the camera. We went off script there. We do send questions in advance, but I appreciate you answering. Thank you.

Louise:
Obviously we’re talking a lot about coverage guarantees, but I would imagine that is not the only way you are reporting to your clients. You are kind of measuring your work in a few different ways. Would you mind going into a few details about how you show your value to your clients?

Paul:
Yeah, of course. So we will always build a full measurement framework before we start working with clients. So that’ll be the pitch stage, we’ll look at a number of different metrics. If a client, for instance, is they’ve exhausted their paid media budget and they’re looking to grow the number of people who are aware of them so they can convert them at cost effective model then they might be looking at something like brand awareness. Brand awareness gets a lot of criticism in measurement circles but actually when used in the right way it can be really really powerful. So if we need something around awareness we might look at say a proxy for that which is something like branded search because the more people are searching for your startup the more people have heard of you. There’s all sorts of other things that will track and might track lease generated, we might track actual consumer sentiment with surveys, but it’s always tailored to what the initial objective is. Yes, we’ll have that common point of earned media coverage as a KPI on the contract, but there’ll be a whole other framework built around it to make sure that we know the needles that we’re moving are the right ones. How that’s all reported on is we will have, every client has dashboards built in Looker Studio, formerly Google Dashboard, I think it is. And that basically pulls in data from various media sources, whether that’s analytics, our own internal tracking systems, media monitoring, all that sort of stuff. And that will then come into a centralised dashboard, which when we talk our clients through, rather than having to put together a full report each week or each month, there’s a sort of a live dashboard where we can see what’s happening, what impact it’s making and contextualise that all as we talk our clients through it.

Louise:
And where’s AVE in all this?

Paul:
What a question. I sometimes want to go to AMEC with a t-shirt on saying AVE is actually quite cool and just see what would happen. No, AVE is dead. It’s been dead in the UK for years. I’ve not seen a single PR agency use AVE for years. I have seen it used in other markets, but I think the I think the battle has been won in the UK on AVE and that’s well and truly in the ground. I actually think the bigger problem in terms of measurement is not the tools or the metrics that we’re using. I actually just think it’s the bigger problem is that the PR industry is afraid of measurement. I think a lot of people deep down have a bit of an insecurity complex about what the channel delivers, which is ridiculous, because in the year, for instance, we’re working with a client, we spent a couple of grand on a campaign that got into every national media going, show me another media platform, another tactical channel that can deliver that sort of an impact on that type of budget. But I actually think that sort of the biggest challenge in measurement in PR generally, it’s a fear of finding out what those test results actually show you.

Steve:
Yeah, I would 100% agree with that. I started out, you know, sort of non-digital PR. The reason I moved into digital because I’d worked with a client who they came to us and said, we basically want to do PR because we want more investment in our business we want to use you know the headlines the coverage that we get as a proof point that we’re kind of legit people talk about think about is all of that kind of stuff and we did that and i got the investment and then they moved on and grew their successful company i found that so exciting. It was like, oh, we played a big part in that, got some national, at the time it was like print coverage, you know, so I was doing the little clippings with my guillotine and stick it in a book and all that. And I was fascinated, moved into digital, but there are a lot of agencies, as you eloquently put, Paul, that are fearful that what they’ll uncover if they dig deep is like, oh, maybe it’s not having the impact we’re claiming it’s going to have. Whereas you are basing your whole business on saying, these are the hard numbers. This is what, you know, we’re sticking our neck on the line saying we’re going to do or deliver what we say we’re going to do. I wanted to come back to tools because you mentioned it a couple of times there. You mentioned Looker Studio. What other tools or platforms do you find most effective for PR measurement? And this would be great for listeners who are, you know, there’s a lot of things out there that they could use, but from an expert such as yourself, what’s your go-to?

Paul:
Probably the one that we use that, you know, we use lots of different tools. Probably the one that we use most in terms of a game changer for our measurement is we’re platformed on HubSpot as a business. So what that means is basically every media pitch that we send out, every email, every piece of communications with a journalist is ticketed in the same way a pitch to like a prospect would be if you’re working in a sales team. And what that basically means is we can look back historically at four and a half years of pitching and say, if we pitch X number of journalists in this way, or if we put, you know, Y number of news releases out a year, we’ll get roughly Z number of pieces of coverage in ABC publications. And that’s really, really useful. You know, we’re the only, I believe we’re the only agency, tech agency at least, that could say, for instance, to our clients, our clients always want to be in TechCrunch. going back to that you know helping people fundraise everyone wants to be in TechCrunch they can put that bit of TechCrunch coverage in the pitch deck document and I would absolutely love to do a you know a bit of research where you look at amount of money raised by companies with TechCrunch in their investor deck and without it because I genuinely think it would have a tangible difference to how much investors are willing to open up their wallets Back to the actual HubSpot, looking at TechCrunch, for instance, when we always get asked by our clients, can you get us into TechCrunch? And sometimes, by the way, we say to them, no. Sometimes we say, actually, that the stories that we’ve got right now aren’t quite ready for you, but we can back that up. We can look at our data and we can say, in the past X number of months, we pitched TechCrunch 28 times and that’s resulted in 12 pieces of coverage and these are the types of releases that work, these are types of releases that aren’t working. It’s really really valuable and I genuinely believe that with your guts you know you should know as a publicist roughly how many pieces of coverage, you know, each item should secure. Again, there’s going to be swings, sometimes it’s going to be over, sometimes it’s going to be under. I think with your gut, you should be able to, you know, roughly say across all of these things, this is how much coverage you’ll get. But the way we’re able to do it really, really accurately is we have the historical data that we can look at. We can say if we do an X number of news releases and if they pass the master on how newsworthy they are, et cetera, et cetera, that result in this number of pieces of coverage. So that HubSpot is probably sort of the biggest game changer in terms of how we measure specifically our activity as it links to results.

Steve:
That’s a great piece of advice. That’s not what I was expecting you to say because often we ask about tools and then people kind of give long lists of the kind of tools they use. And I know there are plenty that you would use at Hard Numbers, but actually just that way of thinking about it, really, really interesting and definitely something I’m going to take away. So thanks for sharing, Paul.

Louise:
Is there anything emerging, AI is obviously like a big thing, but any other kind of technologies that are on the horizon that you think are going to change this space in terms of, I guess, either looking at your own work or looking at your clients results. Is there anything that you think is going to be changing stuff in the next couple of years?

Paul:
I know it’s wildly unoriginal to answer with AI. I know you said that I’m not allowed to say it, but I am going to say it. And the reason being is I genuinely think we’re at an inflection point on the quality of written content. So, you know, ChatGPT came out, it was two years ago, this November, right? And everyone said, oh, God, you know, if you draft content for a living, you know, you’re screwed because it’s going to be able to do all these sorts of things for you. Actually, in terms of content production, most of the quality of work that it puts out is generally fairly naff. But we’ve kept testing with things. And actually some of the later models that have come out in the last six months or so, we did an exercise internally at the agency, and we do these sorts of things regularly, where we called it the byline Turing test. So can you use certain platforms to write content that would fool one of your colleagues as to whether it was written? by a human or by a machine. And actually a lot of the content, in most circumstances, and this was, you know, like a little, you know, interactive exercise we did with half an hour of doing it, most of the stuff it spat out to begin with was naff. It still didn’t pass the, you know, quote-unquote, and I’m sort of misusing this test, but the Turing test. But what we actually found is when we started really analysing why it wasn’t tricking us, there were actually really specific things it was doing that could be fixed with, you know, certain ways of prompting. So, for instance, it’s really weird. It does things like it had, you know, each byline might be six paragraphs, but each paragraph would have the exact same structure. Each paragraph, weirdly, was like disconnected. So it was almost like instead of, you know, a proper argument being laid out start to finish, it was almost like eight, six or eight tweets that were all loosely linked but just bunged together as an article. And all of that sort of stuff you could actually get a long way to fixing by tweaking the prompt, by bringing in examples, all that sort of stuff. So the reason I’m stubbornly insisting on being unoriginal and saying AI when asked this is I do think that the next 12 months is, I think it’s actually ready to write content not better than the best writers, but better than maybe the average writer. And let’s not forget, it only needs to be better than the average writer, not better than the best writers for it to actually uplift the quality of our output. So yeah, I think that’s going to be… I think these things always take longer. I read a quote that was new technologies we overestimate their impact on the short term underestimated in in the long term and while I don’t think we’re quite into the long term I can kind of see why that’s true because the first couple of years it hasn’t quite worked but I do think the next 12 months or so we’re going to see I think we’ll see at least one agency or maybe a couple of agencies actually stealing a march on a proposition where they can deliver a lot more in terms of output based on budget

Louise:
Do you have a current kind of stance on how your teams use AI in terms of maybe writing a press release or writing content, like you said? Do you have rules at the moment?

Paul:
We do have rules and there’s a number of rules and guidelines, but fundamentally it’s about transparency, about transparency with your colleagues, it’s about transparency with your clients, it’s about transparency with journalists. If you do put content in front of them that is AI generated, if they ask, then obviously you have to tell them. That’s the big sticking point. I genuinely think we’re at a point where if you used it correctly, and I think a lot of people might roll their eyes at this because I think a lot of people probably put in like a lazy prompts and, you know, crap in, crap out, right? But I actually think if you use it properly, it’s almost, if not already there at the point, it can produce high quality content. The big thing though is how will journalists react, and that’s something I’d love to understand more. If you said that yes, this release was drafted with the support of AI, I suspect that a lot of journalists would be very adverse to that. but let’s not forget that most publishers are already doing something in the sphere as well. So I do wonder, I don’t know, I think someone’s going to pop their head above the parapet and there’s probably going to be quite a lot of criticism, quite a lot of debate around this, but that’s probably the sticking point is to how will journalists react if you were to start using this on a regular basis?

Steve:
Yeah, it’d be interesting to see what happens for sure because ultimately, even the human created content, journalists will change and edit and work in their own style, their own slight different tone, etc. I agree. I think over time it will become kind of more normal or more normalised. I can’t think of the exact right word there, but one of those will do the job. But yeah, it’ll be very interesting to see how it develops. So Paul, we’re ready for the last question. We’re asking all guests in this third season to answer one of two questions. The first is, could you share a PR campaign or story that you’ve loved recently or which brand you would love to work with and why?

Paul:
I’m going to talk about a campaign that I love, if that’s okay. And it is, to continue the cliche of being a PR person on a podcast, talking about AI, it does have an element of AI to it. And that was the AI Granny that O2 have launched recently in the last couple of days. to help screen fake callers or scan callers. So the idea is how it works is they’ve built this thing it sounds like a sort of a doddling you know old lady and when they detect an incoming scan call they will direct this call to this person or to this you know quote-unquote person and the idea is is the AI will just waste this scammer’s time and I absolutely love this as an idea I love it as a just as you know it’s such a fantastic idea from a creative perspective because it takes existing technology it takes existing but it’s built on like a really key audience pain point and it solves it in in a way that is really really discussable and something that everyone’s going to love but also there’s just so many details to it you know the fact it’s a granny they even gave the granny a name i can’t remember what the name is but they just built this whole brilliant thing and in you know two years we’ve had all these AI driven things launches campaigns and it’s got you know really really tired and boring. They’ve managed to be fresh within that and I think it’s a fantastic way of not just using technology to engage but really to use an audience pain point to build campaigns that people love.

Louise:
It’s nice to have something where you’re using AI for good and for a task that you don’t want to do. I think there was that quote, I’m going to butcher the actual quote, but it was something like, we don’t want AI like painting the next fresco, we want AI to take the bins out. It’s kind of like, that’s where the benefit to life is because you take away the drudgery or you’re taking away the fact that a scammer is going to be busy talking to this robot granny rather than you. So yeah, it’s a lovely take on it. Have you ever seen the, I see it often on TikTok, I think there’s YouTubes as well, where there have been people who do this in the past and film it where they just waste a scammer’s time for hours and hours and they get so angry on the other side. It’s the next level up from that. It’s great.

Paul:
It’s delicious. It’s just built on such a shared human experience. The reason I love this campaign is because if you want to sort of dissect it, there are so many things it does right. And one of them is it picks a common enemy that everyone hates. And it actually uses that. Very often brands are too conservative in how they, you know, they communicate because they’re worried about, you know, upsetting somebody or all that sort of stuff. But they went, no, everyone hates scammers. Let’s have a bit of fun with scammers. Let’s do something interesting. And it blew up. And I can’t imagine the campaign budget was massive for this. It certainly didn’t need to be, because again, it makes use of existing technology. A couple of assets, I think even the images were AI produced, which obviously normally you wouldn’t do because people don’t like that style, but they expect it in a campaign such as this. Yeah, it’s just brilliant bang for buck and brilliant bit of storytelling.

Steve:
An amazing example. Straight after this, I’m going to check it out in more detail because I hadn’t seen all the imagery. So yeah, brilliant example and a brilliant use of AI there. Paul, it’s been such a pleasure to chat to you. We really appreciate you coming on the podcast, giving up your time, talking to us about the journey that you and Hard Numbers have been on and the kind of things you’re doing there. If people want to get in touch with you, either to kind of pick your brains on some of the topics we’ve discussed or ideally to work with Hard Numbers, how should they get in touch?

Paul:
Well, I’ve just this morning set up a Blue Sky account because I’m leaving Twitter, which I think is very stylish. But now, look, drop me a line on any of the normal channels on Twitter, on Apple’s Dollaroom, go to the website, find the email off there. Or slide into my DMs on LinkedIn, any channel you wish, drop me a line. And whether it’s clients, prospects, or even if there’s folks out there, freelancers looking to start an agency, anything like that, we’re always super happy to, or super keen to help anyone we can. So just drop us a note and we’ll get back to you.

Louise:
Great, very kind of you. Thanks again for joining us, and hopefully speak to you soon. And thanks for everyone for listening.

Steve:
Thanks, Paul. Thanks, guys.