Why critical mass for Foursquare isn’t universal adoption

I’ve been chatting with a couple of people recently and they’ve questioned whether Foursquare will ever get sufficient penetration to break through to the mainstream and influence people in the same way as Facebook or Twitter. Sure Foursquare has 10m users, but that’s hardly the 200m that Twitter has or 750m that Facebook claims. Growth at Foursquare may be strong they argue, but checking in to venues and the like will only ever appeal to a minority of people, and the service is unlikely ever to have significant social or commercial impact.

Now the people I happened to discuss this with were both ex tech VCs, each currently working on their own online start ups, and therefore reasonably sensible commentators on the space, but I think they are missing the point. Here is why I think Foursquare is already at critical mass in many places, and whilst utility will obviously grow with adoption, Foursquare only ever needs a minority of people regularly checking in to generate significant value for users and businesses.

Essentially I think a lot Foursquare’s critics fail to recognise the end point for the service. It’s not about being able to broadcast your location to your friends in the hope that they will drop by, or bragging about cool places you’ve been to (though there is some marginal utility here). It’s about creating a Last.fm for physical locations, a service that helps you discover places you like, and rewards you for staying loyal to them.

Critical mass for this is not universal adoption, which the critics are right in thinking will never happen for Foursquare. Instead critical mass for Foursquare is sufficient penetration for it to have enough engaged users to gather rich data on venues. With this “location graph” Foursquare can generate significant value for the majority of passive observers, as well as the small minority of engaged users who are checking in.

Twitter is great because it allows you to discover content and keep up to date with your interests even if you never write a single tweet, and even if your friends aren’t tweeting themselves. I find few of my friends outside of the media and tech world are on Twitter, but this doesn’t matter – I use Facebook to keep up with them. I’m on Twitter because I’m interested in tech start ups and there are only a few people I need to follow to get a constant feed of (often, if not always!) high quality links to content I would not have discovered otherwise. (Incidentally I tweet for quite separate reasons – to raise my profile on the web, connect with people that I don’t know that well, and engage experts that I have never met.)

Foursquare is analogous to this. It doesn’t really matter if your friends are using the service or not. All that you need is enough people who like similar places to you to be using it. Already if you check into one location then Foursquare will recommend you other venues based on where people who checked into the same location also go to. This is an order of magnitude better than the results that you get by Googling or Yelping locations. Sure there are plenty of comments and tips on Yelp, but it’s time consuming and difficult to determine whether the people leaving the tips and you have anything in common. What if their idea of hip is your idea of grungy? A bar that a banker likes is not going to be perfect for a student.

And if you’re not the checking in type? No problem. As long as some of your friends are using the service, then you can use their check ins as a proxy for yours. After all, you are more likely to share your friends’ tastes that those of the average man in the street. The more you check in the better the results should be, but as long as Foursquare can connect you to the rest of the community in some way, it doesn’t matter whether either your friends or check ins are particularly dense or not.

For me, the key component of the value Foursquare delivers is this context that it gives you about locations. You can fit its recommendations into your mental image of the world because of the link it creates between them and the venues you already know, through telling you about users that have been to both. I realised how important such context is when I set up a business called The Crackberry Times back in 2007 – back when Blackberries were the height of smart phone technology. The Crackberry Times was a weekly email recommending (not reviewing) bars, restaurants and events in London, in much the same way that Daily Candy covers women’s products.

Other business interests that were priorities for us were a constant distraction, and the idea never really took off, but we received a fair amount of praise for our content and editorial. One of the most consistent, and surprising, pieces of feedback we received was that people loved the fact that they had been to a fair proportion of the places we recommended. Because of this, and unlike edgier competitors that talked up venues that no one had ever been to, our readers felt they could trust our recommendations much more. They knew exactly how to slot our recommendations into their world view using what we said about places they had visited as a benchmark to judge what we said about venues they hadn’t been to.

I was pondering the whole issue of Foursquare critical mass a couple of days ago when I had dinner at Freemans in New York’s SoHo. A buzzy American restaurant, I guessed it must go through a couple of hundred covers in a night. How many of these people would need to be Foursquare users to it to have a rich presence on the service, with tips, photos and sufficient check ins to get a sense of the sort of people that liked it? Maybe one or two a night for a month or so? That would only equate to 1% of its patrons being Foursquare users. Not a tall order in Foursquare’s back garden, and in actual fact the place has 5,743 check ins from 4,329 people.

Assume the math holds true for other locations though, and I think it’s likely that with 10m users Foursquare already has sufficient penetration in most major cities in the US and Europe to make meaningful recommendations. Additional users will improve the quality of recommendation that you get, and the depth of content that it can offer about each location, but my guess is Foursquare has already hit critical mass over a huge area. Foursquare will succeed because, as with many other social networks, it can create value for the majority of passive observers with only a small minority of truly engaged users.


Monetizing Foursquare


Foursquare has 10m registered users, an estimated 2-3m monthly uniques and partnerships withAmerican Express and a host of daily deals sites including Living Social, Gilt City, zozi, BuyWithMe and AT&T. What it doesn’t have is significant revenue. Whilst it is apparently taking a cut on the daily deals partnerships, that’s not the case for the Amex deal.

Having just raised another $50m in June, the pressure will be on to start pulling in some serious revenue soon, and despite the recent deals, how and when Foursquare will do this remains far from clear for some critics.

There are a few different ways to look at how Foursquare might monetize its service, but I thought that I would start by examining Foursquare as a loyalty company, as they’ve been talking about “socialising loyalty” for quite a while.

Loyalty basics

Loyalty companies are sophisticated marketing devices for savvy brands, and generate value in five ways:

  • Customer acquisition
  • Purchase frequency
  • Transaction size
  • Customer retention
  • Customer behaviour

They do this by offering customers a variety of rewards in return for changing their purchasing behaviour. So, if Foursquare can start demonstrating its ability to influence its users in any of the ways listed above, then it can start charging brand partners for the value it is creating.

Customer acquisition

Can Foursquare increase the number of customers that a brand has?

I would have thought that this was one of the main areas that Foursquare could create value. Already their merchant platform allows retailers to offer special deals to entice new (and existing) customers in. These seem to be passively served in a one-size-fits-all manner however, relying on users searching for them. Allowing retailers to construct more targeted offers and then actively pushing these out would I believe significantly increase the value to retailers.

For example, imagine if Starbucks could send a deal to every user that checked in to a café within 100m of a Starbucks outlet. Foursquare already has a huge amount of data on their users, so this kind of sophisticated targeting is at least theoretically possible. Starbucks might even decide to exclude users that had recently checked into one of their own stores (perhaps offering them a different sort of deal) and thereby solely target new customers.

Furthermore, as Foursquare is already a social service, it could facilitate the viral discovery of brand partners amongst users and their friends. What if checking in to a location allowed you to offer your friends a deal there? Users would be able to forward on these offers to the friends that they felt would value it most, thereby introducing potential customers with a higher than average lifetime value. (NB offering rewards this way round, rather than to the user that sends out the promotion keeps the recommendation honest, and valued by users. No one likes getting spammed by their friends).

Purchase frequency

Can Foursquare increase the frequency with which users visit and spend with brands?

Here Foursquare has already got some great features in place that could target this lever, and the basic points and badge system encourages users to check in to different places as frequently as possible. For potential Foursquare partners however, there is currently limited value to these features – at the moment Foursquare encourages you to check into as many different places as possible, not to be a loyal customer. Instead of this, and in addition to current merchant incentives (e.g. free coffee with your third check in), Foursquare could be taking money off merchants, and potentially offering them a much more economic marketing tool, if they let merchants buy points and distribute these as bonuses to users.

For example, businesses might buy Foursquare points at $0.01 a go, and give +5 to users every timethey checked in – much in the same way that users currently get awarded +3 when they visit new venues. For users trying to climb the leaderboards against their friends, this could provide a significant incentive for them to visit partner brands that little bit more often than they already did. Even if the effect was relatively small, with large numbers of users the absolute effect could be large, and due to its targeting would likely be very cost effective. Of course even at $0.01 a point Foursquare still makes money, because the virtual points don’t cost them anything, and are expunged on a rolling weekly basis.

Transaction size

Can Foursquare increase the average value of an individual transaction a user makes?

This is more difficult. Traditional loyalty schemes typically achieve this is by giving customers coupons along the lines of: “Spend $20 or more and get $5 off”. The coupons are always structured so that the threshold amount is slightly higher than the historical transaction size for that particular customer. Foursquare tracks physical location across a multitude of different categories and is not restricted to a single individual partner per category; it does not however link actual transaction figures to check ins. The Amex deal should go some way to solving this, but this will obviously be restricted to the users that both have Amex cards and decide to synch them. Even this isn’t the greatest solution, as it requires users to pay with their Amex card as well as check in, and this feels like too many steps for all but die hard points chasers.

Perhaps Foursquare could do more research into average transaction size of its customers en masse? This might work for larger retailers that are willing to put in the time and investment to the analytics required. However, working off averages leaves the reward system (i.e. money off coupons) open to abuse and inaccuracy, and such inefficiencies are precisely what loyalty schemes are designed to remove from traditional marketing.

This might be an area for Foursquare to leave for at least the moment. Further developments with Amex might make this more exciting, as could NFC payments eventually becoming a reality. It certainly doesn’t feel like an area where Foursquare has a natural advantage, and as it’s got so many other things going in its favour, I would suggest it sticks to levers where it has a unique proposition.

Customer retention

Can Foursquare reduce the churn rate of customers, and encourage customers that have moved from a client brand to a competitor to come back?

Similar to the targeting of new customers, I think this is an area where Foursquare could do some really exciting stuff. Foursquare already knows how often users visit different locations, and the typical times that they visit those locations, so all it needs to do is allow brands to analyse this data, and send out targeted offers to users that look like they have churned, or are about to churn.

Whilst card based loyalty schemes can tell if you haven’t been back to a particular store as much as you usually do on a month-by-month basis, Foursquare could theoretically target users who visited a different lunch spot on a day-by-day basis, or even hadn’t had their morning coffee on time. With sufficient analysis it should also be possible to determine when to send loyal customers an offer that makes them feel special and valued, and push their loyalty up a notch, before they even consider heading off to a competitor.

Customer behaviour

Can Foursquare help businesses increase their margins by encouraging customers to change their behaviour to ways that are logistically easier and cheaper to serve?

This final source of loyalty value is not always relevant to brands, but an example might be if the cost to a brand of a customer paying by credit card or cash was different. By offering an incentive for consumers to pay by the preferred method the brand could increase its margin overall. Often such behavioural value is related to billing methods, and for the same reasons as transaction size, this is not an area that Foursquare is naturally strong in.

However, there are almost certainly cases where Foursquare can unlock value here. Perhaps helping food outlets spread out their lunchtime rush by offering users who vary their mealtimes significantly, or who had a late breakfast, to buy lunch outside of the usual 12:30-14:00 peak. Again, the granularity of data that Foursquare has over traditional loyalty offerings is a real advantage, because they can track customer behaviour across the entire universe of retail locations, and push offers to users on a real time basis.

Wrap up

Successfully unlocking value from a loyalty scheme requires significant analytics to generate meaningful insight. This suggests that major brands with the resources to invest, and the volume of customers to analyse, are the natural partners for the sorts of services mentioned here. Whilst Foursquare does not capture transaction value accurately, and this precludes it from unlocking all the sources of value to brands that traditional loyalty schemes cover, the additional data that it can gather on location and competitors, and its ability to push out real time deals to users means it has a significant edge over incumbent loyalty schemes. Developing products that unlock value to brands by facilitating customer acquisition, increasing purchase frequency and improving customer retention would give Foursquare a highly valuable service that it could sell, and let them start monetizing their service.

Image is screenshot from my cell 7.21.11 at 08:04

The Bikini School of Organisational Change

Imagine you want everyone to turn up to work next Friday in beach wear. How could you encourage them to do this, or make any other significant behavioural shift, whether it concerns expense policies, sales techniques or bikinis?

Turns out there are four types of action that would be effective, and using these in combination would lead to the best results.

Role modelling

First, it is important that senior figures at work embrace the change in behaviour, and conspicuously so, talking up their personal commitment to it. If the boss takes every opportunity at lunch to describe his Hawaiian board shorts and matching flip flops, then other people start to believe that this whole beach wear thing is actually going to happen, and start to shape their own behaviour accordingly.

‘People take their cues from those who they consider as ‘significant others’ and model their behaviour accordingly’ – Lorenz

 ‘Proximal others, those in close psychological proximity to a focal individual, including his or her co-workers and leader, are likely to have a strong influence on the individual’s perceptions’ – Lewin


Second, it helps if you give people clear incentives for switching over to a new set of behaviours. If everyone coming into work in their beach wear gets a free lunch, an extra day’s holiday or a boost in their annual review, then they have a very tangible reason to do just that. As always, remember that incentives do not have to be financial, or even material. A massive bunch of flowers and a bottle of champagne is appreciated more than a £50 voucher, whatever people say beforehand.

‘The prize for behaving differently must be greater than the perceived pain involved in entering into the new behaviour’ – Lewin

‘The surrounding environment sends us signals which we each interpret in different ways but which cause greater likelihood to behave in a certain way’ – Skinner

Necessary tools

Third, you could provide people with everything they needed to make the behavioural shift. In this hypothetical case, people are likely to understand what you mean by beachwear, and know how to put a bikini on, so training might be less important. But how about telling people that come Friday you will be providing an assortment of trunks, bathing suits and sarongs to rival the wardrobe of Baywatch? By removing the barrier of finding and remembering to bring in their beach wear, you make it even easier for them to take part.

‘ People don’t leap from unskilled to highly skilled in one bound.  The job of a parent (organisation) is to underpin (or scaffold) the learning at any stage of development so that it becomes embedded in patterns of behaviour before the next developmental leap occurs’ – Vroom

‘Adults learn through a trial and error approach’ – Kolb


Fourth, it would help no end to explain why you want everyone to come into work in their beach wear. Maybe it’s a charity event, maybe the office party theme. If people understand why they are doing something they are not only more likely to get behind the decision, but they can also extrapolate from your stated goals and help out in ways above and beyond the call of duty. Suddenly you’ll find that someone is organising a tombola, or has gone to the trouble of buying miniature umbrellas for all the cocktails, to help reinforce the mood.

‘It’s difficult to behave in a different way if the behaviour is inconsistent with your view of the world’ – Festinger

Wrap up

Whether you are trying to get everyone to turn up to work in bikinis, or have a more important goal, getting people to change their behaviour is about creating the right context. The most effective contexts for behavioural change are ones where role models and incentives are aligned, people have the necessary skills and tools to make a change, and they understand why they are doing it.

80% of your Facebook friends have already read this post

There’s a sign that inhabits the bathroom of almost every hotel. It informs you that by indicating you don’t wish to have your towels replaced, you can help save the world. By reusing your towels, you’ll save them from being washed. Washing would require detergent, which once used will end up in our drains, our rivers, and eventually the ocean. Baby seals live in the ocean. Did you know that? THINK OF THE BABY SEALS.

This is of course, all true. I haven’t seen the sums on how much detergent the average towel wash takes, or how much detergent makes it from the washing machine to the ocean, or the levels of detergent that are required to slaughter a baby seal, but on some level, there’s no denying it. Of course, as a cynic, I would point out that the main benefit is the money that it saves the hotel on detergent, but either way we all have a vested interest in reusing towels and in doing so reducing detergent consumption.

However, despite the imagery, these signs are not a very good way of stopping people from demanding fresh linen. Baby seals may be cute, but they have surprisingly little emotional hold on the average hotel guest. People don’t feel that bought into the idea, even if they like baby seals, and don’t feel that they will have much effect on the World in general. Some people recycle their towels, but many don’t.

With the fate of baby seals and hotel margins at stake though, not to mention a situation that the average consumer can easily relate to, a couple of pop scientists saw the perfect opportunity to run a few experiments and write them up as a chapter in an intriguing book [LINK] on behavioural change. I found it super interesting, and hence a quick summary…

The base case in this hotel example is a sign informing guests that they have a choice over whether or not their towels are washed every day. The details of the sign are then changed to try and increase the proportion of guests that actually do reuse their towels. How did the effectiveness of different messages go? Well, from worst to best:

  1. “You can reuse your towels”
  2. “Reuse your towels to help us save the environment”
  3. “Most guests staying in this hotel reuse their towels to help save the environment”
  4. “Most guests staying in this room reuse their towels to help save the environment”

[Note: room vs. hotel in last two]

The interesting thing is that effectiveness of the message increases significantly with seemingly trivial statements about social norms on towel reuse, and that referring to the room that people are staying in is much better than just referring to the hotel (you can buy the book if you want the actual percentages).

It all sounds a little odd until you realise that by referring to the room, the sign is creating a social group that the guest is a part of, which is smaller than by just referring to the hotel. Fewer people have stayed in whichever room the guest is in than in the hotel as a whole. Because the social group is smaller the bonds between members are stronger – there are fewer members to diffuse the responsibility amongst. Should one person fail in their “duty” – in this case reusing their towels – then the performance of the group as a whole will be dragged down that much more. Of course, the sign in each room could be the same – guests will only read one – the point is not that the guests ARE part of a small group, but they feel like one.

Translate to advertising

This form of influence is only just beginning to permeate advertising. Traditionally advertisers have spent huge amounts of money on celebrity endorsements, and understandably given the effect that such endorsements can have. But could brands get a similar or even better effect by placing consumers in tight social groups, and then defining the normal consumption patterns of the group? This is effectively what the Facebook “like” button does. Brands may only be tapping into this psychology crudely at the moment (many brands do not currently translate well online), but it certainly seems like a tactic that will much more popular as social networks become even more ingrained to our way of life.

Wrap up

It’s obvious stuff that our behaviour is influenced by those around us. What is slightly less obvious is that placing individuals in smaller groups has a bigger effect than putting them in larger ones. Rather than convincing people that “we’re all in it together”, influencers should try to persuade people that “us few are in it together”.