How would you go about designing a new location sharing app for Google?

Clarification and Understanding

Location sharing means sharing the location of one user, as determined by the location of their phone, with another user or several. What is our company like, what are we trying to accomplish with this app? Monetization or Engagement? Location sharing features exist today on smart phones but aren't very compelling or often used, for that reason let's focus on engagement.

Use Cases

1. Trying to locate another person - Think finding your friend at a concert or large crowd

2. Keeping track of someone - Parents keeping an eye on kids, or an unhappy wife/husband tracking  an unfaithful spouse

3. Scouting out a location - Is the supermarket busy right now? If I go out tonight is downtown going to be dead?

Out of the above use cases, there already exists solutions today for use case number 1 and I personally don't find it very useful. I think the use case for number 2 is compelling but the market size for users who would use it is realistically too small. 

I would like to focus on number 3. This seems like a use case that hasn't already leveraged location sharing so there should be a lot of room for improvement which wil help use increase engagement amongst our users.

Needs / Pain Points

1. I want to know if a place is going to be over crowded so I don't have to take longer than usual

2. Need to know if there are going to be other people around for me to interact with

3. I want to know more about the type of people at a location

4. Want to give other people information about the location they are at without being concerned about their privacy or safety

Solutions

1. Scout - This would be a feature inside of Google maps, that tells users approximately just how many people are at a specific location. For example, if I wanted to go to the park and play pick up basketball I could search for the basketball court on Google maps and see a status report on how busy it is based off of how many other smart phones are currently detected within x distance of the basketball court's coordinates. 

2. Cupid - This would be an app focused on helping single individuals find where to go to meet other single individuals in their area. Users would fill out a profile indicating basic demographic information and their preferences, then Cupid would display a heatmap of the different local areas where other singles are currently hanging out. This could be critical in helping people pick out where to go on a Friday night.

3. Friend Alert - When you are out and about in public, and someone you know is nearby you will both receive an alert that the other person is nearby. For example, imagine going to a concert but not knowing that your friend from highschool was also going to be there. With this feature you will be given the chance to connect in person. 

We don't have the resources to build out all of the solutions listed above in parallel, so let's go ahead and pick one to focus on while keeping in mind that our goal is to increase app usage with this feature. To help ups prioritize let's go ahead and compare some feature tradeoffs:

Customer Satisfaction, Monetizability, Ease of Implementation

1. B, C, A

2. A, A, C

3. B, C, B

Scout seems like it would be a value add but it is only an incremental improvement as Google already currently offers estimates of busy times for locations inside of Google Maps. While I think Friend Alert is a solution that a lot of people would enjoy and appreciate, I don't think those sorts of missed connections would happen frequently enough to cause a significant increase in engagement. 

I am going to suggest that we focus our development efforts on solution 2, Cupid. While it is going to be the most resource intensive to build out, I think it will give us the biggest lift in terms of engagement. I can really see singles and groups of people incorporating Cupid into their weekly going out routine. Checking out the app while pregaming for a night out on the town, or opening it up when planning their next move.

One of the downsides I see with this solution is it really would be more of a killer feature for an existing social or dating app. I know with the sunsetting of Google+ Google has started to focus less on social products; however, with Google Maps Google already has a lot of the technical infrastructure and data collection in place to make this feature a reality. Perhaps Google could build out this feature but then license it to all of the different dating apps which is a competitive market to be in. The market is so competitive that this new feature could just become tablestakes for competing in that space.

Summary

To summarize, with the goal of increasing engagement amongst users we are going to satisfy the scouting out a location use case by offering a new app called "Cupid". Cupid will provide users with a heatmap of where other eligible singles in their area are located. This will help increase engagement, as measured by weekly active users, by giving users a reason to open up the app any time they are planning for a night out on the town.

I have been providing a lot of feedback on the platform and not answering questions lately.  So I decided to find a recent product design question that no one has answered and give it go - this one was fun.  

1.       Clarifying questions

a.       Location sharing – do you mean an app that shares the location of one user with another user or groups of users?  Yes

b.       I know that android phones have a built-in function that allows a user to opt in and share their location with other users, it is built into maps, we want to do something different then this?  Yes – the feature has low usage (I made that up) and we want to start from scratch and come up with ways that location sharing might be adopted and more useful.

c.       Are there any specific goals besides adoption and usefulness?  I’ll leave that to you.

d.       Do we have a reason for launching this feature?  As you mentioned location sharing is already built in as a feature of android phones but we haven’t explored how this function could be turned into a product.

e.       If this a monetized feature?  Assume that this is going to be a free app and we will find monetization latter

f.        Is there a specific geo or time frame for development?  No

                                                               i.      I’m not sure if it will be important but I’d like to limit to the US – the functionality may very well be universal but for the question I’d like to limit to a geo that I know, is that OK?  Yes

g.       There has been a lot of assumption that app and location sharing = mobile app?  Is it Ok that I assume this is a mobile and limit to android devices?  Yes

2.       Goal – Create a mobile app that we believe would be adopted and used related to sharing location. 

3.       Personas

a.       Individuals that want to share their location with friends and family

b.       Parents that want to share their location and know the location of their kids

c.       Business that want to track their employees

4.       Pick a persona based on criteria – I’d like to focus on sharing location with friends.  Family sharing and parent child sharing already exist and likely are mapped based solutions (just show where they are).  Business employee tracking is the same – already exists and likely focuses on utilitarian location sharing.  I am going to try to come up with non map based solutions (not just showing where people are) that are more exciting.

5.       Pain Points

a.       I want to know where my friends are at

b.       I want my friends to know where I am at

c.       I’m board and I want to meet up with friends

d.       I’m meeting a group of friends and we need to find a place to meet in the middle

e.       We are at a large location and need to find each other. 

6.       Pick a pain point based on criteria

a.       This is the point where I tell people not to boil the ocean and pick a pain point based on criteria.  I’m going to break my own rule – I think the solutions I am going to propose are fun and may solve multiple pain points. 

7.       Solutions

a.       Marco Polo – Marco pollo is an app that lets users announce their location to set of friends.  It’s a Saturday night the user selects a list of friends and clicks the Marco button.  The app broadcasts to the selected friends that user says Marco – if they respond polo it shares their location and draws an arrow on the users screens that shows the direction of the friend that responds.  Why show arrows and not just directions.  I am thinking about all those times you went to the mall, a concert, a club with friends and get separated.  The map would show location but at a club we might just need to know the general direction to walk.  Another use case – your friends are already seated at a large restaurant you walk in and look around – you don’t see them.  You press the Marco button, your friend presses pollo – you are directionally guided to them - almost like a game of Marco Polo . 

b.       Meet in the middle – you want to meet up with a group of friends for drinks, dinner, etc.  You add everyone to meet in the middle they consent, and the app determines the bet place to meet in the middle.  You could imagine functionality that takes into consideration everyone’s preferences (AI).  We could add in functionality where people vote on options being presented or input what they want and the AI sorts out recomednations.  Basically its like an AI night planner that takes into account everyone’s location and preferences and recommends where you end up.  Imagine a tie in with reservations and reviews. 

c.       Proximity meter - want to see who is close by – open the app it shows you how far away your opted in friends are.  From there you could text or message friends that are close.  The app doesn’t show exact locations till you allowed and we can limit the precession so it just shows you like who is within a mile, 5 miles, 10 miles etc.  That way you really can’t stalk anyone – you just know they are a mile away – note you still have to opt in even for that level or location – think uses cases like a college campus and surrounding bars.  Open the proximity meter and see which friends are close by.  "Marco"  - maybe this one and Marco Polo are a pair.  

8.       Pick a solution based on criteria

a.       Marco polo – medium reach – medium impact – medium effort

                                                               i.      The idea of compass directional arrows really works when people are walking distance from one another.  I think this limits reach and impact.  I think it is great for college students on campus and surrounding local bars. 

b.      Meet in the middle – high reach – high impact – high effort

c.       Proximity meter - medium reach – medium impact – medium effort

Of the solutions I think meet in the middle has the biggest impact of the solutions.  It also allows for monetization opportunities as we promote locations and make reservations.  I also  think it is really helpful for users - who hasn't tried to cordiante that night out with multiple friends.  As a go to market strategy, I would try to tie this into existing google maps and Google assistant functionality rather than attempt to make a free-standing app.  The plus side of integration into the existing Google infrastructure is a built-in user base, the downside is that updates and v2 may be slower as we need to include them into the main code of maps and assistant rather than as a stand-alone app where we could iterate faster. 

9.       Measuring success – unlike maps and assistant where the user might use the functionality every day / multiple times a day – meet in the middle only gets used when groups need to meet up.  This means that measurements like  DAU and MAU and the ratio DAU / MAU will be much lower than maps and assistant – and the tool could still be a success.  I think we can design some meaningful metrics

a.       Acquisition - 1st time users per period.  We should see period over period increases if the tool is gaining traction

b.       Engagement and Utility - % of uses where the group reaches consensus and meets in the middle using the app.  The higher the consensus percentage the better the tool is operating. 

c.       Retention - % of multiple uses or uses per user.  If the tool is making group meets ups easier than we would expect to see repeat uses.

d.       Hidden (maybe) Counter Metric – the tool may be highly successful and also not get a lot of use.  It is possible that some groups will use the tool once and keep going back to that place but not use the app.  We could look at if user locations group together without using the app. 

                                                               i.      I’d also track features like reservations and POI / restaurant search when not using the app – it’s possible the poor experiences on the meet in the middle could spill over to deceased confidence in the recommendations of the maps ans assistance.  

10.   Limitations –

a.       preferences of the individuals in the group may not be known because they don’t use google maps and assistant

b.       HUGE – if the app is built into maps and assistant then it likely does not work on IOS by default which would kill the solution – Let’s meet in the middle (oh half of you have Apple phones).  I really didn’t realize the fatal flow until I got here.  We would have to have an IOS campaign app or we would have to launch this as an app instead of native functionality (maybe this is why the  app doesn’t already exist)

c.  Lowest common denominator  - I really love indian food but when organizing a group of multiple firends we may never eat indian based on A.   The strong downward prefences of one could reduce the options and the AI might end up alwasy guiding the group to the lowest common denominator  - this happens in real life as well but you balme the person for the fact that you ended up at tex mex again - here you start balmming the app.  

11.   Summary – I was asked to create an app that featured location sharing and would actually get used.  I looked at a few persona and settled in on friends because I felt I could create some really fun apps that weren’t just show the other person on a map.  I looked at the pain points and broke one of my rules and decided to solve the problem generally.  I came up with three well named solutions and decided to go with the solution that had the biggest user impact “Meet in the Middle”.  I identified some key metrics including % of group consensus using the tool and % of returning users (do they come back).  Finally I looked at limitations and may have killed the solution before we wrote a line of code (which is way cheaper than killing it 3 years latter!).  The solution needs to work on both IOS and Android which likely means it is a free standing app rather than built within Google maps and assistant.  This will change the adoption model of the app and needs more thought.