25 minutes reading time (4915 words)

In conversation with... John V Willshire and Philip John

The rapid increase in hyperlocal websites has delivered some challenges to the communications sector, including competition with established local media; a greater depth of audience engagement; and a local, focused approach to advertising and media buying.

With the last point in mind, we caught up with PHD's John V Willshire, and Philip John of the Lichfield Blog and Journal Local, to discuss how to develop a converged view of the new world of local advertising - between advertisers, media planners and buyers, and hyperlocal owners themselves.
Philip John, John V Willshire
Is hyperlocal on the media agenda?
JVW: I have been thinking about this from the perspective of what media agencies used to think of as being “local”. We have local teams, who have dealt with local press for years. What you see in the local paper is a compendium of all of the things that people want to know and do. They want to know about the performance of Kidderminster Harriers; who has something for sale; a sense of community. Hyperlocal has fostered fragmentation: Kidderminster Harriers has a fan site; local selling is on eBay; the sense of community is in small sites. Where I live, there's a community site for four streets. That notion of the “local paper” has fragmented right down into the paper's sections.
From our [PHD's] perspective, how do you buy into that? It's impractical to buy that, individually. No media organisation is geared to deal with hyperlocality, because we used to be able to bulk buy. From that perspective, we have to change the way in which we do things. The model, as it stands, cannot deal with that fragmentation across lots of communities of interest, in lots of different places.
Has it been something which PHD has looked at, in terms of internally restructuring to meet the configuration of hyperlocal, or has there been a view that it's far too fragmented to build business?
JVW: It comes down to the model of the creative agency making the ads, and the media agency placing the ads. If we're being asked by the creative agency to place some ads, contextually, that fragmentation doesn't work. It's too hard, and not the best way to connect people with companies.
What we are trying to do is reframe the purpose. As a media agency, what we always have done, and always will do, is connect people with companies, for the benefit of both. We used to do that through the mass media model, because that was the best way to do it. Now, we're re-considering the best way to connect people with companies. So, we might offer a suite of things on a quid pro quo basis: we could fund your hyperlocal website by having relevant advertising, but we choose the ads for you. Rather than us having to engage with every hyperlocal website owner, we could act as the marketplace.
PJ: There are a lot of hyperlocal website owners now considering that they can't do everything on their own. A lot of owners started their website as a hobby, and now realise that it is taking up a lot of their time. They're looking to other hyperlocals for support, and that's why you're getting sites like Talk About Local and little networks starting to pop up.
I think that there are a lot of hyperlocal site owners that are interested in joining a network that would enable them to go to advertisers. So, if they all have a sports section, they - or their agency - could go to the advertisers [with a collective proposition]. Sky, for example, could place an ad for their sports package.
There's a movement towards that now. Addiply is creating a network which is focused on local, and is now moving into other areas, such as media blogs, so they are now going into subject areas. If you combine hyperlocal with topics, then you have something [compelling] to take to advertisers. The top-down and bottom-up models are moving towards each other.
JVW: It's really interesting that you pointed out the topic areas. Before the hyperlocal movement, you were stuck in your locality. You talked to the same people in your street, and were with the same people at work. You were constrained within your location and social class. You would talk to each other about what you had in common: the big, generic things. For example, if you were really interested in 1960s architecture, you couldn't find anyone to talk to, and there wasn't a newspaper or radio station covering it. So, you can now go online, and talk to anyone in the world about 1960s architecture in the UK, or across the world.
PJ: The example that I always use is Chasetown FC; we have a lot on the Lichfield Blog about them. The editor has now started to go to matches, and do some little vox pops. There's a lot of people on the web and on Twitter, interested in Chasetown FC. The thought occurred to me that they might not be interested in the rest of the Lichfield Blog, so why are we serving them ads that are site-wide, for things like web designers? Why not ask Chasetown FC's sponsors and ask them to advertise Chasetown FC stuff only? It lets the advertisers get to their target audience more effectively. If they sponsor Chasetown FC, it's because they want Chasetown FC supporters to look at them. So, obviously the supporters will be looking at the blogs, and you build up a market.
It's little things like that - becoming a little bit more clever on how you advertise. It's really simple to do.
From a publisher's perspective, is this a case of “tanks on the lawn”? The pre-Internet world produced fanzines, for example, which were often poor quality. Fanzines would never have a voice with advertisers of Chasetown FC's programme. But, now, you're bringing in a high readership that is possibly equivalent to Chasetown FC's website, so in volume terms, you have an equal voice.
PJ: It makes advertisers more accessible, and it's ultimately cheaper, as wastage for the advertiser is minimal. Previously, whatever you did had so much wastage attached to it. If you've got a bunch of hyperlocal websites, and you can target really well - which is easy, with targeting on tags for example - then you cut out so much wastage, then it becomes so much cheaper. Multiples shrink, and hyperlocal websites become a much more attractive proposition.

Philip John, John V Willshire

Do you think that there's still some way to go, in terms of fragmentation? The size and professionalism of the Lichfield Blog is a shining example of that. Is it only a matter of time that this fragmentation evolves, and we see collective operations, or even Newsquest just coming in and buying sites up?
JVW: You see it elsewhere. Big companies come in and think “I can scale this”. What if it doesn't scale, and is the perfect size of site for the perfect amount of people? Advertising is still about millions, not thousands. Community and hyperlocal websites are about being small and valuable. If someone comes in and tries to scale, covering wider issues or a wider area, then it loses relevance.
PJ: The people that create hyperlocal sites do so because they are passionate. That's why they do it for free. You can see a clear difference between these bottom-up hyperlocals and the top-down networks. I don't think that the top-down models get the same audience, or have the same feel. You just don't see it. The bottom-up models are generated by people in that community, who go out and talk with people. There's a real connection, and that's why it works so well, even thought it takes a lot of their time.
The only example that I think does work, is the Guardian Local project. That's fantastic, and down to probably down to the fact that Sarah Hartley is in charge of it, and knows the bottom-up model of hyperlocal. There are four of them - one person in each city. Unless that approach is taken, I don't think that it works. You're still going to have to get the bottom-up hyperlocals into some sort of cohesive network, in order to make it work on a bigger scale. So, instead of taking a site and scaling it up, you have to take a site and do it again.
JVW: It's not about taking one site and scaling it, it's about taking the principles and scaling them. “This works really well here, how can we take the lessons that we have learned and use them over here?” It's like what you saw in Internet advertising, with networks. You would have network buys, and websites had to be on the network; the network would then say “We have the websites, now send us some ads”. This is different.
PJ: It's almost like web rings - lots of websites that all decide to get together. There's not a lot of that going on at the moment, and that's what I'm doing with Journal Local. But, how do you make it worthwhile? Is it money, or a need to get advertising? It's a Catch-22, as to get advertising, you need to be in the network, to make it a good proposition to advertisers.

To get [hyperlocal] advertising, you need to be in the network.
Philip John

JVW: The advertising model works on economies of scale. So much effort would go into making an ad, that sending it to 4000 people wouldn't be worthwhile; it needs to go to 4 million. In the hyperlocal model, you can't serve the same ad to all of those communities. The beauty of it, is that hyperlocal owners have sites which are personal, within their community, and so on. So, I waste my time just by sending the same ad - and it's too expensive for me to produce 50 ads. I wonder if the solution is for the advertisers to let go, and say “What I've got is a toolkit of stuff that's going on in my shops, and I have shops in all of your locations. Take this toolkit, and put something from it onto your site, if it's of relevance and your audience is interested, may click on, or may buy” - and this could be a partnership, rather than just on an advertising rate basis. But, the advertisers aren't very good at letting go of control, and creative agencies aren't very good at letting go of control. It's against the grain, and it's hard work.
It's a different model.
JVW: It is taking the network principle, but rather than bombard the whole network with one ad that kind-of-works, it's about working in partnership, rather than as a supplier.
PJ: You can still go halfway, though. What I'm really excited about, is the potential to use data to help to target. When you have a network of hyperlocal sites, and you know what postcode areas they cover, you could tie it into something like healthcare. Addiply is on JournalLive, a hyperlocal network of 22 websites. One of the sites had a massive leaderboard for BUPA. We saw it and wondered who is going to click on it; that ad was a waste of time and BUPA's money. That got me thinking: what if you could get health data, and the NHS wants to target heart disease in certain areas? You could hack away with some data, and produce a list of sites that have a high rate of heart disease in their area. What you have is a target market. You then make it easier for the advertiser, because you're targeting for them; cutting their wastage, rather than a blanket campaign everywhere, just to get the people you want.
You are removing control of where the ad is displayed - because you are telling advertisers where to display them, but then you give control to the publisher, saying that we should advertise here, where there is a high rate of heart disease. It makes perfect sense to both.
So, there's something here about publicly-accessible, open data, and adding that to the mix, to validate the context.
JVW: Then the network becomes smarter and self-aware. As an advertiser, I can then put things into this, as the network is helping me to steer content. I might have twenty things or one thing, but the network decides your relevance.
There's so much data around, and what you see now is Government departments saying “This data that we collect and hold: we shouldn't be scared of letting it go. As long as people aren't being nefarious with it, let's see what good things can be achieved.” TFL, for example, has realised that they can't do that [internally]. The way in which their organisation works and how their projects work and are billed for, means that they cannot cost efficiently do that with their data. It's free and open. You made the point [Philip] about people running hyperlocal websites for love; people also spend weeks creating stuff with TFL data.
From your perspective within PHD, do you see value in this freely-available open data?
JVW: Yes, but everyone's trying to change the model. The media agency, the ad agency, and the client, want to change. You will find a few people in each agency, trying to pull everyone out of this - because it's a machine. It's a machine that has been created over the past 20 years, ever since media agencies split from ad agencies to create two separate entities. Up to the end of the 1990s, the machine was really humming. The famed years of the “media lunch” used to involve “doing some media planning” - to decide what the split was between TV, posters, and whatever - then going for lunch. Guttingly, it's the era that I missed [!]
We've always used Mosaic, ACORN data and so on, so I think that it will be something which is good. If I compare this data against someone taking the time and effort to look at all of the publicly-available data... can I beat that? If you streamed all of that data into a system, it would be better, more flexible, and more “alive” than commercial data.
It will be continually updated; if you built it right, it's feeding a live, changing picture of the UK.

Philip John, John V Willshire

PJ: Openly Local does a really good thing; there are 300 hyperlocals on there now. You can go on and see the latest news from all the hyperlocals. I helped to create a Google custom search for hyperlocals. You could repurpose all of this - as Openly Local does with government data. It scrapes local government websites, and repurposes it as government data. If you did that with hyperlocals as well, you would have the Guardian Open Platform for hyperlocals. You can then grab that, mess with that, and create some great stuff.
JVW: Going back to your health example: if you had that system in place, scraping all of that data - government and hyperlocal - it would spot things quite quickly. If you set the network up to alert you on a local health scare related to cleanliness, you could match it to a product that helps to clean surfaces. All that I would need to do, is set up this query, and leave it there, so it's almost a minesweeper: I just wait for it to erupt. It's a passive action for me.
PJ: It could also become a newswire. If you're sitting in the Birmingham Post and Mail, you could mine hyperlocals in the West Midlands.
JVW: It's in one place. It would tell you a very interesting story about Britain - area versus area, and so forth, which becomes a very interesting resource for everything: media planning, comms planning and newspapers, who love “stories with a difference” - people in the north loving something more than people in the south, and so on. It becomes a really interesting, live picture, rather than the top-down perspective of a national survey of 500 people. This is based on live movements: things happening from the bottom up.
Has access to open data been beneficial?
PJ: It has, but the problem is that most hyperlocals are started by active citizens, who have enough nous to set up a site. But, beyond that, they don't always know how to set up how to use this data, how to use Yahoo Pipes, and so on. We've got a great system on the Lichfield Blog, where I'm the techie, the editor is a trained local journalist. He produces the articles, and I look after the back end. It's a great relationship, but most hyperlocals don't have that. I can play around with data, but they can't. It's great, but it needs someone that has the time to do something interesting with this data.
I'm sponsoring Hacks and Hackers in Manchester. Its aim is to get such a relationship going between developers and journalists, to help journalists to tell stories from the data. That's really important, because I blogged on this yesterday. One of the issues regarding open data is whether someone will use it for other purposes. There was a recent situation regarding land ownership data in Bangladesh; property companies were going in there and marginalising poorer parts of the community. They were able to use this data, and the community wasn't. Hacks and Hackers is there to make sure that data is used effectively.
Even though the data is coming out, and it's great, it's of little use unless people can make it usable for hyperlocals. Openly Local is trying to repurpose council data, in order to add context to stories. If they are talking about a councillor, for example, and they are tagging through Wordpress, it adds on a bunch of information about the councilor, beside the article. If someone in that ward is reading the article and wants to talk to the councillor about the issue, they get a fact file right there, next to the article. That's really helpful, but it takes time for people to do, and there's not enough of that out there.
So, if hyperlocal owners don't know much about technology, does it follow that they won't know much about advertising?
PJ: Yes. We put Addiply on the Lichfield Blog, and our ads were snapped up straight away; all of the inventory was taken up. There are quite a few sites running Addiply, and they have no advertisers. Those site owners may have the view that they don't want to go out to advertisers; they're not salespeople, and would feel a little weird, going out to their community, asking them to advertise. There is that problem.

John V Willshire

It goes back to those sites needing to be part of a network. Because there is then more collective resourcing, there will be more resources to pull advertisers in, and the network is more attractive to them.

Is there a perceptual thing here? If the hyperlocal owners see themselves as “deliverers of information” and not involved with ad sales, how can they compete with newspapers if they see themselves more as being a public service?
PJ: There are very few hyperlocals that see themselves as a replacement to local media. Hyperlocal is in between the community - people that journalists might want to talk to - and the media. So, all that hyperlocal is, in many cases, is the community simply making its voice heard online. Without this, they would have to phone the local paper. Hyperlocal informs the media what the community thinks.
JVW: It's something that's so easy to organise. Before, you would have had a meeting at the local parish hall, that wasn't necessarily open, and you had to invite a journalist to attend, and someone had to go to that journalist and brief them. The media was in control: what they published, how they published it, the context in which they placed it, whether they were interested. Now, you wouldn't necessarily care if the media was interested: hyperlocal is the place to put a view across.
PJ: It also gives a voice to people that previously couldn't attend those meetings.
JVW: It's a great leveller. People are happier giving their opinion on a hyperlocal site than public speaking.

We're much happier buying things with tall potential than long potential, because long potential is less proven.
John V Willshire



Do you think that there will be a perceptual thing here, for media buyers? It's not a newspaper, it isn't This Is London, it's a thing in the middle for 4000 people.
JVW: This is it. It's that 4000. There's something really interesting in behavioural economics, about the impact of numbers. If I can talk to 4000 people about an issue, then that's brilliant. But then, by putting it into a newspaper that reaches four million, then 4000 doesn't look as good - but not everyone sees the newspaper ad.
It's just reach.
JVW: It's just reach; it's about tall potential, and long potential.
The tall potential is the reach of four million people, but only 4000 people will do something. Community models have long potential: if you start something with people, then you could reach 4000. Media agencies are bad at selling that to clients, because everyone's interested in tall potential: the big numbers.
It's a problem. We're much happier buying things with tall potential than long potential, because long potential is less proven, and gives you smaller numbers.
There's a potentially massive benefit here to hyperlocal owners: they should go to advertisers and convince them that they should be changing the game from tall to long, in this context.
JVW: An example might be Sainsbury's, who is a client of ours. Through the tall model, we would place an ad on every hyperlocal site.

Philip John

PJ: The reach would not matter to us [in Lichfield], as the nearest Sainsbury's is 20 miles away!.. so there's wastage.
JVW: Exactly! You just reach lots of people. The long model with Sainsbury's is finding the hyperlocal sites relevant to the stores - using their data. We could then run a project where store managers are put in touch with the hyperlocal site, and work together to deliver benefits to the community.
Is there client interest in making locally relevant advertising, in either buying or content?
JVW: There is interest around it. CEOs consider it to be very important, but the way in which organisations have been constructed, means that responsibility sits in different departments. The marketing silo is all about those big numbers, where local issues can be with HR, field sales, store managers, and anyone else.
Hyperlocal offers a solution to something which is a bit like marketing, a bit like local management, a bit like HR, and a bit like CSR. It doesn't work in one silo, so it's not always easy to get buy-in. With Cadbury Spots v Stripes, it was clear that it was not just marketing: it was business. So, we needed to talk to HR, about bringing it into the company first - which makes it a lot easier to help that story to travel. And, extending that analogy, all of the other department doors then open up: marketing and HR talk together, then IT become involved, and so on.
There is extreme interest from companies which are not in silos. But, companies were organised before communication technology, and most are still in silos. Departments do one thing, and do it well. We, as an agency, work with not just the marketing teams, but find other people within the organisation, and bring them in... and everyone's really receptive. That's the key. Everyone wants to make their company better, and everyone wants to break down the silos. It's about providing the 'grease' to make that happen.
Hyperlocals probably couldn't get to chat to Justin Rose or Terry Leahy, but they could chat to local store managers. Does that mean that aggregation, whether as a group or network of hyperlocals, is necessary to have such a voice in the market?
PJ: Definitely. At the moment, you can't really advertise on hyperlocals if you're a big company - and a lot of hyperlocals might turn round and say that it doesn't make sense for them.
With the Lichfield Blog, unless an ad is really well targeted, we might say no. I want to attract national advertisers to hyperlocals and to the Lichfield Blog, but it has to be really targeted.
JVW: It has to be right. From our perspective, it's more interesting doing projects [with hyperlocals] than just advertising. Projects are in partnership: this is something that we are doing, rather than something that someone designed in the London office and published to the site. It's getting our industry out of the habit of just serving an ad.
PJ: We do that a lot. With the Fuse Festival, our photographer and I spent all weekend there. We captured as much as possible, including all of the meetings that the Fuse board had over the year - so we were actively involved in organising it. Tesco had a presence there, with children's activities including papier mache and cardboard models. I thought how great it would have been to have the corner of the car park, with this big thing that was produced in association with Tesco, with kids joining in... but we didn't know about it. If we did, we could have phoned Tesco.
JVW: But, that probably slips in the gaps between the PR silo, the marketing silo...
PJ: If Tesco know that this stuff is going on, then they simply get a store map, and cross-reference it with a map of hyperlocal sites, which they send to store managers. If they know that it's going on, it makes the whole process easier.
That sounds like a really quick and easy thing to do.
PJ: Hyperlocals would put themselves on the map, draw a circle around their area, and with the stores, you would then create overlaps. There are your target areas.
It goes back to what we were saying: moving away from display ads, to co-branded projects. Tesco colleagues painting a primary school has a greater contextual relevance than a display ad that may be quite targeted, but not as rich and inviting.
JVW: It's about moving advertising away from the age of information: “I'm going to tell you something about our products”. I have information available: I can search on Google. There's not really much point in advertisers simply telling you something, as people can find out for themselves. Advertising has to be more about stories of things that we've done.
Does that have to be part of an organisational step change within the client: opening the doors between silos, throughout the organisation?
JVW: You have to hand over this control. It's no longer about matching luggage, with everything looking and feeling the same, and the brand onion dictating that we are chatty. It's ceding that control throughout the organisation.
In The Wizard of Oz, the bogus Wizard is the giant glowing head at the end of the room who tells them to do stuff, and they begrudgingly go off and do it. They don't get anything out of the relationship. Brand can be these giant glowing heads, shouting at people.
When they find out that it's a guy behind a curtain, they then talk on an individual basis. On that conversational, one-to-one basis, it's easier to sort out problems. You could do the faceless brand thing in certain local environments, and it would work to a certain extent. But, if you ceded control to the guy behind the curtain - the store manager - and it became part of community engagement, then there's a very specific voice, which would be very different in every town across the country.
PJ: It's like hyperlocal - you know the people behind them, because you've seen them. We get out so much, that people are now starting to recognise us. It's those connections.
JVW: Brand is increasingly becoming shorthand for the people within the company. What you have when you talk about a brand, is a community. "Those people over there did that thing; I'd quite like to talk to those people over there". It's picking apart the notion of a brand tone of voice. People are increasingly used to being able to do that in other aspects, in their world: to talk with individual personalities. You would still have brand cues around that, because the community needs shared values. Talking in the same way in the same tone of voice is still likely, but it's not as prescribed, or as top-down.
John V Willshire is Chief Innovation Officer at PHD Media. His blog is Feeding the Puppy.

Philip John is part of the team behind The Lichfield Blog and founder of Journal Local, sponsor of the first Hacks and Hackers Day in Manchester on 15 October.
TV: refusing to give up the fight
In conversation with... Toby Barnes and Marcus Bro...