The Guardian - Growing the user base The Guardian way with Jesse Wilkins

[00:00:00.910] - Olivier Destrebecq
Welcome to the Subscription League, a podcast by Purchasely. Listen to what's working in subscription apps. In each episode, we invite leaders of the app industry who are mastering the subscription model for mobile apps. To learn more about subscriptions, head to Subscriptionleague.com. Let's get started.

[00:00:22.590] - Olivier Destrebecq
Welcome, everybody. Today I have Nicolas Tissier as my co host, and we'll be interviewing Jesse Wilkins, senior growth manager at The Guardian. He has worked on how to convert Guardian's reader to Guardian subscriber with no paywall and while keeping all the content free.

[00:00:37.870] - Olivier Destrebecq
In 2018, The Guardian with Jessie's team crossed the 1 million financial supports. Jesse, welcome to the podcast. Do you want to tell us a bit more about yourself?

[00:00:46.990] - Jesse Wilkins
Sure. Yeah. Hello, and thank you for having me. As you've mentioned, I've been at The Guardian for five, six years now. I was lucky enough to be working in a team back in 2016 that came onto the idea of voluntary donations, product which we call contributions. That has developed really well and played a significant part in that one million number that you mentioned and also in helping us break even.

[00:01:15.970] - Olivier Destrebecq
Awesome. You're working in an industry that has struggled to find a business model in the digital age. Can you explain what business model The Guardian has landed on, from what you just said, it allowed you to break even with those one million financial supports?

[00:01:30.830] - Jesse Wilkins
Definitely. I'll start with some fundamentals. Our belief, at The Guardian, is that the world is a better place when people understand the world that we are in and have journalism that is freely available to help explain that world. The model that we settled on is, a supporter model that enables our readers to give. But our journalism remains free and open to anyone.

[00:01:53.330] - Jesse Wilkins
Any reader, regardless of whether they give or not, can access our website and they can read an unlimited number of articles every day, anywhere in the world.

[00:02:01.520] - Jesse Wilkins
In terms of how the business model works, on that side, print and advertising revenue streams have always been important, and they remain critical. But the digital reader revenues is where we have seen the growth in recent years.

[00:02:14.780] - Jesse Wilkins
By that I mean our contributions product, digital subscriptions, a former membership scheme. As you've touched upon already, contributions is a donations product, where readers give a one time, a monthly or an annual amount in order to support they get nothing tangible in return.

[00:02:31.360] - Jesse Wilkins
Digital subscriptions does not give access to additional content, so it's not a subscription in that sense, but it enhances the reader's experience with features such as ad free. We've seen fantastic growth on both of those products.

[00:02:46.170] - Olivier Destrebecq
Can you give some background on the ownership model at The Guardian? Because my understanding is that it's not a typical news organization either.

[00:02:54.020] - Jesse Wilkins
The ownership model is quite key to how things came about. The Garden is owned by the Scott Trust,

and the Scott Trust has one purpose and that is to support The Guardian in perpetuity. There are no requirements to pay up to shareholders, there is no owner or model figure demanding returns.

[00:03:10.350] - Jesse Wilkins
That has enabled the organization to take a different view over the long term and build a huge audience and to go back a little bit. Keeping the website free and open enabled The Guardian to go. I think, in the early '90s, it was the 8th biggest UK paper, so it was a relatively small paper in the UK back then.

[00:03:30.000] - Jesse Wilkins
By the 2015, I think, it was in the top three in the English speaking world globally in terms of the website audience. There was already a success story there in terms of audience. But back at that point, as I'm sure you guys will remember too, the rise of Google and Facebook in the advertising world, amongst other things, meant there were very heavy losses at The Guardian and something needed to change in the business model.

[00:03:53.720] - Jesse Wilkins
Kath Viner he became editor in chief in 2015, set things up quite clearly in terms of values and said, we want our journalism to remain global, free, accessible for our readers, and not available only to those who can afford it. We want to give people the opportunity to contribute regardless of how much or how often, where they live, how they get their news.

[00:04:13.100] - Jesse Wilkins
The organization wanted a model that... A way of giving that felt part of that and spoke to those values and was an extension of it rather than something that we stuck onto it. That was that context. Then we tried things. We tried a membership model, we tried digital subscriptions and contributions.

[00:04:30.070] - Jesse Wilkins
Consideration was given to a paywall, but the decision was made to go with a more open model and it paid off. They made a profit by 2019, I would say we found a way that worked for us and it fitted that model and the times we were in. I wouldn't presuppose that would work for everybody. I do think there are lessons that we have learned that could be applied elsewhere.

[00:04:52.050] - Olivier Destrebecq
I'm sure you'll share a lot of those as we move through the interview. Curious, because you're describing a challenging place that you are in, where you have to keep the content free and yet you still have money. In a way, you're trying to get people to give you money for nothing different that they couldn't get otherwise. Can you tell us how you approach that challenge?

[00:05:14.400] - Jesse Wilkins
Definitely. As you've described it, it was a very particular problem at that point in time. Huge readership, amazing breadth of content, high quality content. But also a commitment not to put any of that behind a paywall.

[00:05:28.930] - Jesse Wilkins
Having that guard rail, I described, where we said, there will be no paywall, it actually made things simpler. We didn't have to think too hard about that. We knew there wasn't going to be one within the team I was in.

[00:05:39.340] - Jesse Wilkins
So then you focus on persuading readers to give on a voluntary basis. As you guys will know better than me, that's a hard thing to do in many ways. Most content businesses will ultimately either withhold content or functionality, or they might create friction as a way of persuading users to pay. But not having that option gave us a bit of focus.

[00:05:57.740] - Jesse Wilkins
I'll just describe some of the very early experiments. At that time, we had a small membership scheme

where people could pay a monthly fee and as a reward for becoming a member, they were giving things such as a bag with Guardian branding and a certificate.

[00:06:13.390] - Jesse Wilkins
I remember one key piece of research where some of these members came to us and they said, "Please don't send me these things. I wanted all of my money to go to journalism and you're posting me these physical things and it must be costing you money."

[00:06:27.250] - Jesse Wilkins
That really stuck at the time and it helps us think about the problem. When people are coming to you and saying, "I just want my money to go to the journalism," that's a helpful thing.

[00:06:36.240] - Jesse Wilkins
Another thing that sticks is we had a colored pop up banner, but that worked to some degree. It's one of the first things I worked on and we had the line, "For the price of a coffee, you can support independent journalism." Some of the listeners might remember that if they are long time readers of The Guardian, and that worked well as well.

[00:06:53.100] - Jesse Wilkins
There were some signals that we were getting and we then experimented with a bit more copy to make a pitch just for people to give, getting nothing in return. We initially used our existing ad unit containers.

[00:07:05.880] - Jesse Wilkins
As a online newspaper, The Guardian has a whole host of ad unit containers, basically powered by Google Ads. We put some of our messages within those containers because they were there. It was functionality there to use. It didn't really work. There wasn't much success there.

[00:07:21.920] - Jesse Wilkins
Then we played with some different formats and designs and then we tried a bespoke container at the end of our articles. It wasn't an ad unit, it was a more native feel, and that was a breakthrough. It worked very well in comparison to the ad unit messages.

[00:07:38.940] - Olivier Destrebecq
For websites, for mobile apps, it's always great to be able to experiment and to experiment, obviously we have to use AB testing. Can you tell us in your mind why it's so important to do AB testing and do those experimentations?

[00:07:52.320] - Jesse Wilkins
Sure. From those very early experiments, there were four insights worth pulling out. One of them was that a native design was important. We have ads on our site. They are incredibly non invasive. They've been beautifully incorporated into the user experience, but nevertheless, they are ads.

[00:08:10.340] - Jesse Wilkins
It turned out that having a message that was at the end of the article in a design language which felt like an extension of that article, it put us into the journalistic space and it kept the reader's attention and that was really important.

[00:08:23.780] - Jesse Wilkins
We found that language and tone were another thing and I could talk more about that conversational tone really worked for us. You might remember this, we said, "Since you're here, we've got a favor to ask. Since you're at the end of this article, we've got something to ask." Quite direct, quite conversational.

[00:08:40.150] - Jesse Wilkins
Number three was the message did not have to be short and snappy. Conventional wisdom suggests otherwise. To this day, people come to me and said, you should really consider making that message

a bit shorter. But we found that people-

[00:08:54.810] - Olivier Destrebecq Like we did.

[00:08:56.610] - Jesse Wilkins
We did. We have tested this stuff hundreds of times. We found that a longer message was effective. The thing about a longer message is it gives you the space to say more, to talk about more complex things, more things and articulate ideas, and tell more about story. That was important.

[00:09:13.100] - Jesse Wilkins
Finally, the position in the reader's experience was important. We are talking to the readers when they have just read an article and they have got some value. But all in all, you will still see the message there to this day. It doesn't look super slick, but it ties to our core purpose and it's powerful and it remains so.

[00:09:32.840] - Olivier Destrebecq
You mentioned earlier, you've tested the lengths of the message and that stuff. How much optimization did you put behind the current messaging on each article and subscription page?

[00:09:42.780] - Jesse Wilkins
The short answer is a lot. When we got to this point, I just described. We now had a channel which we could use to try and persuade readers to support us. We found that the words, the ideas, the language, and the tone were all important. That lends itself to AB testing copy.

[00:10:01.810] - Jesse Wilkins
Obviously from a technical perspective, changing some words and running that as an AB test, it's relatively simple. We had to build some infrastructure, but it's not a crazy hard thing to do.

[00:10:10.940] - Jesse Wilkins
That set up meant that we could test that copy quite aggressively when we combine it with research and we used it as a way to develop our core support proposition.

[00:10:21.460] - Nicolas Tissier
Regarding those membership banner, I'm not sure if it's a good way to call them? Do you have anything to share regarding the triggering strategy and the way copier and the messaging is adapted to the content on which it is displayed?

[00:10:34.920] - Jesse Wilkins
Sure. Indirect answer to your question, when do we show the messages and what we say is adapted to the copy, it's all been explored and optimized through copy testing.

[00:10:47.200] - Jesse Wilkins
Basically, no changes are made unless they are validated. If we don't find an improvement, we won't change it. Everything is structured in that way and the learnings have been built up over time in that way.

[00:10:58.790] - Jesse Wilkins
Many of the major news events of the last few years, such as Trump, Brexit, the Pandemic, major elections, have created huge surges in our readers engagement and we have been able to adapt our messaging to speak to those events and the role that The Guardian plays in them.

[00:11:14.340] - Jesse Wilkins
We also adapt our copy to our regions. Some topics resonate strongly with readers, subjects such as climate change, for example, are very important to them. Those data points have led us to create campaigns such as our Climate Pledge, which we have run for the past two years, which you may have seen on the site.

[00:11:33.040] - Jesse Wilkins
I would say, if you were to say, how far can that go? It's been a fantastic tool for us. There is a point at which you run to the edge of the benefit. There is a complexity overhead when segmenting our audience too much. There are also benefits to simplicity where it's possible because it enables you to move much quicker and test much quicker. There is a trade off and it's not a infinite world of always testing segment more, but huge benefits.

[00:12:01.240] - Olivier Destrebecq
One of the things that always fascinate me with AB testing and all that stuff is hearing about the failed experiments and the surprising failures, but also the surprising success. We know what works by just looking at the website and see what is there today. Can you tell us about those failed experiments and surprises?

[00:12:19.600] - Jesse Wilkins
Sure. I'll keep it to ones within the team I was in. But we all have a list of our favorite failures and they all seem like such a good idea at the time, and that's the great thing about them.

[00:12:32.710] - Jesse Wilkins
A few of the ones that jump out at me, we had something that we called, a love button, which was a heart emoji button that people could click on to show how much they loved us. It's fair to say that one didn't go very far.

[00:12:46.630] - Jesse Wilkins
We had a point early on, and I talked about this at the beginning, we really wanted to double down on the point that we did not have a billionaire owner. We thought we would represent this visually with a yacht, because billionaires have yachts so we had a picture of a yacht, a nice little graphic, and our billionaire yacht was sinking to show how little we needed billionaires.

[00:13:07.490] - Jesse Wilkins
Again, it made sense at the time, but it's fair to say that we managed to muddle the message along the way. I remember we got some early feedback from readers who were really quite worried about the sinking ship because they thought that was The Guardian and were like, "Are you okay, guys?" You've got a sinking ship. It doesn't seem to be going well, so that was another one. It didn't get much further. We did...

[00:13:33.870] - Jesse Wilkins
There have been long and passionate debates about some of the color changes, buttons and things. It's fair to say that color changes have not been something that we have seen use changes from.

[00:13:45.170] - Jesse Wilkins
We learned that words were something we were better at changing. Things like photography and images, perhaps on our landing page was a little bit less successful.

[00:13:55.810] - Jesse Wilkins
We tried a Valentine's message, that was another one that didn't get much further. Then the making the messages shorter, it continues to come up and it's another one where, it's not that it can't be done, but it's not an obvious route to success. That's just a few that come to mind.

[00:14:09.860] - Olivier Destrebecq
Interesting. The one that you mentioned about the color is not changing much. There are so many places online where they'll say, "You need to make sure to AB test the color of your buttons and all that. You'll make a huge impact." I guess, it always depends on the website and on the app. It's very interesting to hear that for you guys, that didn't have much of an impact.

[00:14:25.620] - Nicolas Tissier
What about on the mobile side? Because all those experiments are easily connected on the web. How

do you perform on mobile? Is there any difference with what you noticed on the web?

[00:14:35.760] - Jesse Wilkins
When you say mobile, do you mean mobile web?

[00:14:38.480] - Nicolas Tissier Yeah, mobile app, I mean.

[00:14:40.220] - Jesse Wilkins
Mobile app. I wouldn't say we have seen significant differences if we're talking about things like colors and so on. I mean, I would say very broadly, the app audience consumes much more content, and it has often reacted very well to learnings that we have seen on web.

[00:14:58.090] - Jesse Wilkins
Also, vice versa, learnings from our app audience have translated at times quite well onto our web audience. There's some cross pollination there.

[00:15:06.580] - Jesse Wilkins
I guess I would say in terms of the velocity of testing and learning, it has been easier on our web platforms because it's been a bigger audience with more data, more conversions, and we've been able to build tooling to support that more successfully.

[00:15:21.460] - Olivier Destrebecq
Nice. One of the benefits of the approach that you're taking is you have all the content available to everybody. It's not hidden behind a paywall. How does that impact your job and the findings that you can have by having all that content available?

[00:15:36.540] - Jesse Wilkins
Sure. Keeping the journalism on our website free and open is part of our values, and it's actually become one of the reasons people give. People have told us that the idea that they are giving in order for others to access it is actually one of the motivations.

[00:15:52.850] - Jesse Wilkins
I would say giving people unrestricted access has probably helped us understand what is important to a wider range of readers than a paywall site. Obviously, I haven't worked on a paywall site at the same time, but it's almost certainly given us access to a wider range of people and helped build our understanding of them.

[00:16:13.190] - Jesse Wilkins
I would say we probably have more signals as well in that we can really see what people read when there are no restrictions and how they behave. As I said, it's probably made things simpler and it really focuses on persuasion.

[00:16:25.260] - Jesse Wilkins
Another thing that you can see on the website today, it's enabled us to do things like tell the reader how much they have read. You will see a message on the end of article containers and it will say, "You have read 100 articles in the past 12 months." That's really effective in reminding a user how much value they have received.

[00:16:45.020] - Jesse Wilkins
It's quite powerful. It's just telling them this is how much you've enjoyed our website. The more that somebody reads, the higher that number is and the more effective the message becomes. I guess, that's another example that it wouldn't really be the same if you could only read one article per month. But if you have that unrestricted access, obviously it changes that picture.

[00:17:05.860] - Olivier Destrebecq
Were you able to correlate that reading number to donations in any way? I'm curious to understand if

is it your heavier reader they're donating the most, or is there a different correlation there?

[00:17:17.690] - Jesse Wilkins
We don't necessarily see a correlation with how much an individual is likely to give, but we do see the more engaged a reader is, the more heavily they read us, they do become more likely to give. That's not going to be a surprise. Someone reading 20 articles a day, every day for a year is just a heavy user and clearly likes the product.

[00:17:37.700] - Jesse Wilkins
There is definitely that correlation has always been there. We've just through that article count element, we've surfaced it, and we are reflecting value back to people to almost remind them of how much they've enjoyed reading us.

[00:17:50.710] - Olivier Destrebecq
That's a great feature. The solution that you've picked for The Guardian is very different from what most newspaper websites have picked so far. Do you have a guess as to why it's working for you guys while everybody's going the other direction?

[00:18:05.430] - Jesse Wilkins
It's a big question. I feel there could be a whole episode about alternative universes. Where The New York Times might have gone with an open website and where that might have got them.

[00:18:15.550] - Jesse Wilkins
We've talked about some of the factors about ownership structure and releasing open journalism that took us down our path and how we made it work. Honestly, it's hard for me to say that it couldn't work for others.

[00:18:26.390] - Jesse Wilkins
It's been a huge experiment for us. It's not a simple thing, obviously, for our media peers to try and just suddenly completely switch off your paywall. But the landscape does change all the time.

[00:18:37.880] - Jesse Wilkins
One thing I do think it's never black and white. It's never just one thing or the other. There are overlaps between both approaches. One interesting thing that I see is that, quite big chunks of our messaging strategy have been copied by organizations around the world.

[00:18:55.750] - Jesse Wilkins
I have a file with screenshots, and there are gray and yellow boxes appearing at the end of articles in all sorts of unexpected places, from Argentina to the Netherlands to the US. Some of them have lifted the language word for word and translated it.

[00:19:14.690] - Jesse Wilkins
I've described how our solution, how that came about. But just because we arrived at it through our path doesn't mean that it can't be used for other organizations.

[00:19:24.280] - Jesse Wilkins
I chatted to a newspaper in the Netherlands last year, and it was a fairly open conversation. I shared the things that had worked well for us. I explained how the end of article message was structured. They tried it and they had a go and they reported back that it was quite successful for them, and they were essentially a paywalled site. That's quite interesting to me, the lifting certain parts and moving them back and forth.

[00:19:47.460] - Jesse Wilkins
I do think the scale of our audience, which has come about because of that openness, has accelerated how quickly we've been able to learn. As a person at the center of trying a couple of hundred things or so a year, I know how important that scale and that velocity is.

[00:20:03.870] - Olivier Destrebecq
You guys are lucky to have a large audience to do some AB testing that is meaningful very quickly, I'm assuming.

[00:20:11.760] - Jesse Wilkins
Definitely. If you go back to those early days that I described when we realized we had this message and we could change things for improvement, it was really exciting. We found that A B testing was both a means to improve our conversion rate, but it was also just a tool to learn. Bear in mind at this point, there was not really a template for our model, and there wasn't really a manual on how to do it.

[00:20:35.980] - Jesse Wilkins
The messages were very visible, as they are to this day, so the wider world could see what we were doing. There was actually a cartoon in the Private Eye magazine in 2017. It depicted The Guardian as having a begging bowl and asking for money. It's fair to say there was some doubt, whether this approach could really work and whether you could just ask for money.

[00:20:56.420] - Jesse Wilkins
The AB testing really helped us explore and learn. The advantage we had was that a copy was displayed a few times a month to each reader around the world. As you've said, the size and the scale of that audience meant we could get results back quite quickly.

[00:21:09.260] - Jesse Wilkins
We had lots of data coming in, and that created the potential for a very tight and fast feedback loop. We'd conduct user interviews and other forms of research, form a hypothesis, translate that into a copy variant, publish on the site, get answers back within two, three days quite often.

[00:21:24.850] - Jesse Wilkins
We had a very small cross functional team. Sometimes the researcher would sit next to the copywriter, who would sit next to me. The work would flow straight through very quickly. I'm sure, like many of you guys listening, will understand that being able to learn quickly is a fantastic thing.

[00:21:42.310] - Jesse Wilkins
If you can try hundreds of things in a twelve month period, it allows ideas to flourish and be tried. It gives room to fail. Most of our ideas failed, but we learned from all of them, and we were able to double down on the wins and build a foundation of knowledge very quickly.

[00:21:58.220] - Jesse Wilkins
It became a really valuable tool to help us tell our story. We're an organization with hundreds of journalists, but we weren't so used to telling our own story. The research and the AB testing enabled us to see ourselves as our readers saw us, to hold that mirror up and speak to them in an authentic way.

[00:22:18.130] - Jesse Wilkins
One important point we pulled out there was, we discovered that the reason people read us and the reasons they choose to give are not quite the same. You might think they are, but there was a distinction there. Again, having that research testing loop allowed us to get that deeper understanding and to be able again to reflect to readers why they might want to support us.

[00:22:40.780] - Jesse Wilkins
It's interesting when we look back on this process. In retrospect, it's easier to see the pivotal moments and when things changed, but when you're in it, it's less clear. Also looking back, we can see just how much it didn't work and how messy that path was at times.

[00:22:55.640] - Olivier Destrebecq
Because I'm sure most of the changes, as you were making them, were small changes and not huge step forward. Totally get that. Jesse, those were all the questions we had ready for you today. If people want to learn more about The Guardian and more about you, where can they go?

[00:23:11.290] - Jesse Wilkins
If you want to learn more about The Guardian, you could always look on The Guardian site and you will see what we're trying, what our approaches are. In terms of asking me about more details, you can find me on LinkedIn and I'd love to chat. I really enjoy dissecting problems and approaches. I'd love to hear from people about things that they have tried or questions that they have on things that we have done.

[00:23:33.420] - Olivier Destrebecq
Great. Thank you very much for coming today. It was such a pleasure to have you and hear your answers to all those questions so we could share that knowledge. Really, thank you.

[00:23:41.620] - Jesse Wilkins Thank you for having me.

[00:23:42.850] - Nicolas Tissier Thank you very much, Jesse. Bye.

[00:23:46.050] - Olivier Destrebecq
On behalf of the Purchasely team, thank you for listening to the Subscription League podcast. If you've enjoyed what you heard, give us a five star review on iTunes or other audio platforms.

[00:23:56.290] - Olivier Destrebecq
To find out more about Purchasely and how we can improve the subscription business, visit purchasely.com. Please hit subscribe in your podcast player and don't miss any future episodes. You can also listen to previous episodes at subscriptionleague.com. See you soon.