I have watched the evolution of Foursquare and now Swarm for years. Usually as a user, but often also as a curious onlooker. As somebody who can not understand, how exactly could they manage to mismanage product design so much and what exactly has pushed them to make one questionable decision after another.
I don’t have an answer for that, but I do have a complete plan of how they would need to redevelop the check-in product for it to become the darling of travelers, collectors, patrons, bureaucrats, entrepreneurs, and investors alike. Read along to see.
When Foursquare started, it offered a rather innovative concept of checking in into places you visit, competing for the mayorship of the location, and collecting various badges in the process.
That was fresh and cool. I was immediately hooked and started using the app wherever I went. The concept of badges especially appealed to me. Like to so many other humans, it feels good to collect things, achieve some recognition for collecting even more, and so on.
Check-in itself was not so interesting, beyond the fact that it was part of the badge collecting process. There was always this notion, that I must be able to find meet friends that I would not otherwise bump into during night-out, or maybe people I know from another city would show up in my Foursquare when visiting, and we could have some surprise lunch together.
That never happened. Often we were in the same party already or communicated our plans in advance. And any surprising visits from other cities were too rare to bother. And I could notice them by pictures posted on Facebook anyway.
Becoming a mayor was never a big thing for me — for popular places it was never really possible to hold the title very long, as there were most certainly someone else who was a more permanent fixture there, while for unpopular places it was pointless — no fun being the mayor without a city to rule.
Then Foursquare decided to split itself in two. Keep reviews and location discovery in the Foursquare app and move check-in functionality into another app called Swarm. Named for its most coveted badge — awarded for every participant if a large number of people checked in in the same location.
Named for the badge, but badgeless. And mayorless. Just pure check-ins. OK to keep getting your check-in fix, if you were really addicted, but not really delivering any extra shots of dopamine.
Later Swarm brought back mayorships. Approximately the same as before. And later on, also points for check-ins and social leaderboards to see how you stand among friends. One of the simplest game mechanics to use. And one with the least possibilities to expand. And ‘Fun Facts’.
Desperation is evident.
At first, let’s declare strategic goals that our application should have. (I will write ‘our’ and ‘we’, as it is hard to place myself in the role of product director of foursquare and then not use those words). Three goals should be good, round number.
- People should use the app daily
- It should bring not only intrinsic joy but also added value
- Our product should become embedded in other products
The first goal is what will eventually bring us money. The second goal supports the first goal by delivering more value for the time spent using the app. And the third goal defends us against the competition, which is rather strong. (e.g. Facebook).
Now, I will list things that need fixing in the core application structure and elements, later we will move to supporting activities.
The first to fix is check-ins themselves. It is a core feature, yet it is almost fundamentally broken. By action of checking in, you create a small piece of value, that later is used either to cook up badges, or advance you in the leaderboard, or give you coveted mayorship. Yet this value does not feel very large, as you can check in multiple times in a row in multiple different locations. Locations that vary by scale (Country, county, city, street, pub), by actuality (events, districts, houses, bridges, bus lines, private apartments with odd names), by distance (pub you are sitting in, pub next doors, pub three miles away), by honesty (restaurant you are sitting in, restaurant you think about going to, restaurant you will never go to).
All those wildly different variability factors diminish intrinsic (and extrinsic) value appointed to one check-in. How can I value my honest daily check-in into the office highly, if I can see that hyperactive friend is checking into every park, restaurant, tram, office, house, and party in the daily commute. Tens of times per day. Putting some easy-to-understand and highly visible constraints on check-ins would make them scarcer and boost their value.
- First should remove redundant scale check-ins. It is silly to pollute the list of locations by items that are basically parent locations. If you check in F-Hoone restaurant, you should be automatically checked-in in Kalamaja district, Tallinn city, Harjumaa county, and the state of Estonia. And if your previous check-in was in a different district, different county, or a different state, only then you should get a higher scale notification and extra value for your check-in.
- Should clearly separate events and places. They should not get mixed up. Events can be disabled completely at first, later they could come together with some good ticket provider integrations and other features to bring real added value. Also, all non-place non-event locations should be removed. No ‘Tram Nr4’ or ‘Bus Tallinn-Tartu’.
- Separate the self-generated and the approved locations. It is fine to be able to create locations yourself, but they should be visible only to you and your friends, and any value from check-ins should be calculated accordingly. Different categories of approved locations could then have several levels of approval — for some, it is enough to pull in map data, for others it is better to engage a local partner to verify details, for yet others, there could be completely third-party sources. For example, the Estonian Tourism Board could provide a verified list of all the castles, museums, sights, hiking routes, and beaches. Basically anything with a brown road sign. It would be win-win cooperation, and could probably be easily automated.
- Add restrictions on the number of check-ins you can do without actually moving around. You can and should check-in when you come to the new cafe or museum, but it should be one check-in per place. there could be also some time limitations — for example, the minimum time between check-ins could be fifteen minutes.
After we have increased the perceived value of check-in, we can move on to various interesting ways how those check-ins could be combined, collected, and counted. It is time to fix Rewards and Mechanics.
Currently, there are points, a leaderboard, and mayorships. Previously there were also badges, most of them in sets with simple progression — One airport, five airports, ten airports. Some of the badges were secret.
If we look at a list of commonly recognized game mechanics, then we have previously used only a small subset of available tools, and we have used them badly. Let’s create for us a new and revamped set of rewards using more precise mechanics. It has to feel the same at the core, but it has to work a lot better. Target is to increase engagement, gratification, and intrinsic value.
The simplest to fix is mayorships. Currently, as I described above, they are broken, because they are too scarce. It is often impossible to win the mayor’s title, as there is too much competition, and also for location owners, it is a bit silly to create interesting extrinsic rewards for the mayor, as there is only one of those ever around. To fix it, we shall introduce several new levels of ownership (with names drawn from different bodies of representatives):
- Emissaries — up to twenty people. The amount is set depending on the size of the venue and amount of people checking in. You achieve status when you have checked in at least three times and there is an empty spot. If all the spots are taken, then every week one check-in is removed from each emissary, eventually dropping the status for some, and opening the spots.
- Legates — up to ten. Legates are unlocked for location when there at least ten emissary places are taken. To become legate, you have to check in five times within two consecutive months as an emissary. When all legate spots are taken, the same attrition starts as for emissaries.
- Senator — one. This is the new mayor. Even higher status. You can really be proud to achieve it. Legate who checks in five times within one calendar month can become Senator. Attrition is always on for Senator. Senator is unlocked only when there are five active legates.
As higher levels are unlocked only if there are enough people participating, we can ensure that senators will appear only in the places where many people are checking in, thus focusing the energy of application users.
For busy location owners having up to 31 potential privileged persons provides a real incentive to create discounts or personal offers. For users, it provides a chance to achieve something, even if you are not exactly going to the place every single day.
Harder to fix are Points and Leaderboards. Point system, in particular, is a hard element to work with, as once you introduce points, then everything suddenly gets valued in points, thus enforcing comparison of various things by how many points do they bring. The comparison that we do not necessarily want users to make.
Upgrading the sticker-badges by spending points, to later earn more points using the upgraded stickers — that is a rather complicated mechanism to drive the usage and improve the experience of Leaderboards. It is way too much effort, for little gain. But it is good to have some feedback connection from badges to check-ins.
Leaderboards are working better, as they are working only within the social circle, but they are fully dependent on accumulated points. And without points, they would not work very well. So, considering all this, for the first iteration of our renewed application I would like to omit points and leaderboards completely. Focus on the important parts — check-ins and collecting badges.
Next in line for fixing are Badges. They used to work quite all right for the first couple of badges you would get, but then it would become increasingly complicated to understand what needs to be done to get the next badge. Getting a badge often felt like a surprise after doing some random check-in.
Now, after conversion to swarm stickers, badges retain a similar setup, only are less visible and not so central for the application experience. They have somehow become less fun while retaining many of their previous attributes.
While there is certainly room for surprise events in Foursquare mechanics, it should not be part of the core experience. Instead, it has to be a more transparent process, with the possibility to highlight some subset of badge lineups, to not be overwhelmed by all the thousands of possibilities.
And of course, lines of progression for badges should not all be exactly the same in terms of application functionality. There should be several completely different categories of achievements, that would each appear differently in the app. And if you would want, you could keep visible only the categories you enjoy the most. So, let’s dig into categories we will get:
Geographical achievements. Everybody has seen those maps shared on Facebook where people check which countries they have visited. Foursquare needs this global map, as well map of each part of the world, and of course county/regional level maps for all the countries. And city district level maps for the biggest/most interesting cities and towns. And each level would be its own lineup of badges. For example: by checking into 5 neighborhoods of Helsinki you get the ‘curious visitor of Helsinki’ badge, by checking into 15, you get the ‘Helsinki explorer’ badge and by checking into all 59, you get badge ‘You must be a taxi driver. Or courier. Or simply most Helsinkian of Helsikinkians.’
And each level of the map is nicely graphically represented. You can see which parts of a particular map you have uncovered, how many check-ins you have done in each part, and what is left to do to get all.
There can also be Geographical Combo Badges. For example, you can do a road trip through California, and try to catch most counties. And by visiting all the countries on ocean drive, unexpectedly get a special hidden badge for this scenic road trip. Or you can visit one pub in every borough of London, to get a special badge to brag to your untappd buddies. Those combo badges also can be their own lineup of achievements — so you can get something already after visiting pubs in five or seven boroughs.
Classical Category Badges — all sorts of badges similar to those in original Foursquare. Five pubs, ten airports, twenty parks, etc. For this to work, there has to either be fewer categories, or a more rigorous categorization process with extra verifications. Otherwise, it has happened way too often, that two very similar places are categorized completely differently. It has to be predictable.
Also, if needed, we can introduce two-layered categories. Like ‘pubs & bars’ as a top category and then ‘Irish pub’, ‘English pub’, ‘Ice bar’, ‘Vodka bar’, ‘Cocktail bar’, ‘Beach bar’, etc as subcategories. That could be the solution for this unpredictability, and then people who do not use the app so much, or just do not visit that many places, could focus their collection efforts on the top category, while others could dive deeper and try to become experts in something really specific. There could also be some Combo Category Badges for visiting establishments in several related categories. Like a library, museum and school could yield one badge of ‘Education & Learning’ lineup, and adding to the list observatory, university and research laboratory would give the next.
And of course, nobody forbids combining it into Geographical Category Combo Badges. Visiting five types of exotic kitchen restaurants in London would yield ‘Hungry Londoner’ and 30 distinct types would yield ‘Greater Foodie explorer of London’ badge. Opportunities are endless.
Then there come Event Badges. They are kept quite separate from the rest of the application and are arranged in similar lineups as other types of check-ins, only they have extra time dimension to play with. So, ‘5 summer festivals of 2016′, ‘Glastonbury 6 years in a row,’ ‘Wacken & Brutal Assault combo’ type of badges is a possibility.
Of course, various Event Combo Badges can be combined to reward visiting various related events in the same country, or in the same season, or by the same venue, or any other combination.
Just need to keep in mind, that each new badge lineup that is unlocked and presented to the user, needs to be actually completely collectible. We do not want the user to see a list of things that he is very very unlikely to ever achieve. Goals need to be within reach, and only a limited amount of goals should be visible at the same time.
Also, as a side feature, when organizing social aspects between players we also have to keep in mind their interests. It could be similar to how social news network tree.social has done it — by allowing you to follow your friends only in categories that interest you. Those could be auto-selected based on your currently highlighted and also previously completed badge lineups.
That’s was it about core elements and rewards, but now we shall talk about supporting activities of embedding our product into the world. We can think of Foursquare or similar check-in solutions as a virtual overlay to the physical world. As we do not have too many augmented reality gadgets yet, the only connection between this overlay and the world is the app in the phone from one side, and various venues recognizing the app and its contents as a real factor influencing their offerings. We need to increase those connections, make them stronger and more permanent.
Previously there have been two usual connections, offering discounts in return for check-ins and some special deals for the Mayor. Both have always worked kind of awkwardly. Not much value was delivered. That is clearly not enough for the future.
We should offer deals to business owners that would make it worthwhile for them to bother setting up a tighter connection with Foursquare, and more added value to end-users, to complement their intrinsic fun experience.
First, we have to make physical integration simpler. Venue owners have to be able to easily recognize that people have checked in, without looking at their phones (who likes strangers looking into their phone, or, god forbid, touching it?), and that they are actually present in location when doing so.
To do that we introduce NFC/QR code sticker labels. They can be attached to every table in the pub, to the entrance of the club, or at the wardrobe line in the theater. It should be simple strips of paper, that are easily attachable, they contain read-only NFC tags and have small QR codes printed on them for people who can not use NFC. After beeping the tag, the venue owner can receive a signal to their POS software, and apply a discount or special offer automatically to the table that has sent the signal. After scanning the tag, it is also immediately visible on the screen, so it can be shown to entrance attendants at the club to get fast track entry for example. Or to get a free starter cocktail.
You might think that distributing and coordinating such stickers would be an expensive and complicated logistics process. That is not necessarily the case. NFC tags are cheap. Printing the stickers is not expensive either. Delivering the tags to the business can be simplified by doing provisioning post factum. That is — the business owners can pick up a bunch of tags at a local distributor (say chain bookstore), then install them in their shop and only after that, by doing a scan, provision them to the backend system as ‘theirs’, thus setting up the connection. In the same way, any damaged tags can be replaced quickly and easily. Foursquare backend system then can store extra parameters for each tag, like table number, to make later integration with POS system more useful.
Why would venue owners want to have POS integration? Because their current loyalty systems and offers are either really bad or do not exist at all. It is very hard to compete for wallet space with another plastic card when all you can offer is a 10% discount for an infrequent purchase. And that is for those who bother to create such systems at all — most of the small restaurants, cafes, and bars do not have any sort of loyalty program. And that means that they do not have any way how to learn more about their customers. How to engage them. For those small businesses, a very attractive product will be Foursquare Loyalty System. It will include more chances to interact with frequent visitors, offer deals to emissaries, personalized offers to legates, and a named dishes for Senator. It will allow small business owners to participate in localized badge combos, analyze review data more precisely and correlate visitor behavior with any other big data input that they might have. It would bring them into the 21st century of customer relationships.
Also, government agencies, such as various tourism boards, can be engaged with similar integrations. Everybody loves their dose of big data and can be in turn persuaded to give some benefits to people participating. Discounted entrance fees, special tours, guides, and other perks.
For larger events and festivals, this could be even more tightly integrated with wristbands, that are going to go the NFC route soon anyway. There could be perks that give extra access to bands, the possibility to use priority entrances, or other event-specific things. all without having to roll out complicated custom IT systems and with possibilities to get access to a wider big data pool.
Of course, customers can easily get worried about Foursquare sharing their data with third parties. This issue we need to solve preemptively — there needs to be a good log of all the data sharing within the web version of Foursquare, so any user can see what has been done with their data. You should be able to opt-out with certain granularity, we do not want customers to just say ‘do not share at all’ — you will have to pick and choose. And with each opt-out, you will be notified of what benefits or functionality you will lose.
OK, let’s look back and see if we have addressed our three strategic goals with the plan outline above:
- People should use the app daily — if you have a really nice and playful check-in and rewards system, and a lot more visible chance to also earn real-world benefits, then yes, we can see people using app very often.
- It should bring not only intrinsic joy but also added value — we have discussed possible gained value via integrations with location owners. There might be even more possibilities in the future.
- Our product should become embedded in other products — if we become an integral part loyalty offer by the location owner, then we have succeeded in becoming part of their product.
This post is part of the series that could be titled ‘What would I do, if I would be in charge of product strategy of company X’:
- How to fix Foursquare. My take on Product.
- How to disrupt European TV market.
