App Invites, Friend Finding and Drenched Spaghetti

There was a time when successfully inviting people to use your app or play your game was as simple as implementing Facebook invites, but once FarmVille peed in everyone’s spaghetti things got a little tougher.

That trend has continued.

So, for the last couple weeks I’ve been doing a deep dive into how some of the top trending social and photo apps in the AppStore handle user invites nowadays. Along the way I’ve learned quite a few things that I wish I could have learned by reading a too-long Medium post rather than creating 30-something new accounts and poking through apps I’m not interested in using regularly. Allow me to save you the same pain.

Here is a sampling of the apps I took a look at over the last couple weeks. For simplicities sake we’ll toss these apps into a couple of easy to digest buckets:

The Big Four:

• Snapchat
• Facebook
• Twitter
• Instagram (Instagram Engineering)

Those that Prioritize Finding Friends:

• Live.ly
• Flipagram (Flipagram)
• Musical.ly
• Periscope (Periscope)
• Wishbone
• Vine
• Disney MIX
• Kik
• Live.me (Live.me)

Those that Prioritize Inviting Friends:

• WhatsApp
• Marco Polo
• Houseparty (jsneedles)

On to the nitty-gritty. We’re going to take a look at two things, 1) How and when an app will choose to prompt invites and 2) what mechanisms are being used to send the invites.

The Big Four

The Big Four are almost not worth talking about since, by and large, they either don’t bother with invites or they bury them so deep in the bowels of their settings, it would make more sense to just send an Uber driver over to your friend’s place with a handwritten note that says, “Yo! Download Twitter already.”

I include them because there’s still something worth learning there. The best way to get people to invite others is always going to be word of mouth and that’s not something you get by using the right SDK or onboarding flow, but it is something you’re more likely to get by being awesome at what you do. (Or just be big. That works too.)

Those that Prioritize Finding Friends

The apps in the “Prioritizing Finding Friends” bucket are characterized by never actually prompting anyone to invite friends.

Instead, the onboarding is often focused on adding your friends who already use the app.

This bucket stuck out to me for a couple of reasons. Mostly, it’s interesting that in a startup world that worships the K-factor the overwhelming majority of apps that I dug through don’t seem to care about it at all. Or, at the very least, they don’t seem to prioritize its optimization.

Of course, the same goes for the “Big Four” which leads me to believe that one reason for such an overwhelming percentage of social apps not having prominent invite features might be folks simply copying what seems to work.

That could be a false promise though. The Big Four have a might and polish behind them that many other apps don’t, so they may be conflating best practices for a large, well establish social network with best practices for a smaller early stage social network.

(Then again, all of the apps I looked at for this are charting fairly high so what do I know?)

So what are these “Prioritize Finding Friends” folks for sure doing right?

To pull myself into the story for a second, in our own onboarding at the work in progress that is Mission: Pic, we stripped out everything we thought was non-essential to your first experience, and that includes the finding friends portion of onboarding that nearly all the apps in this bucket utilized. We wanted people to get to the ah-ha moment as quickly as possible and this seemed like an extra bump in the road.

After digging through a metric 💩-ton of other apps for this exercise…I think we were wrong.

Often times having “friends” in a social app is either necessary or important to the user experience. What is Instagram if you have no friends’ photos in your feed? That’s less true for Mission: Pic so it was easy to write friend finding off as unnecessary. But now I believe that the most important element this priority brings to the table has nothing to do with usability, ease of onboarding or even actual friendliness and everything to do with social proof.

When you dig through 30 apps, many of which serve a demo you aren’t in, it makes a powerful statement when you come across one that can tell you that you already have a handful of friends using their service.

I did run across a couple of missed opportunities in this bucket though. For every app I loaded that started with “Oh! Spence has this app already! Cool!” there was another that started with something more like this:

Don’t get me wrong, that’s some solid error handling, but wouldn’t BETTER error handling be something like:

Holy cow! None of your contacts are using Live.ly! We should fix that. Select everyone you’d like to invite to this magical world of friendly friendness from this list of your contacts you just gave us access to.

Your user is already in the door, presumably they have reason to believe they’re going to enjoy your experience and it’s reasonable to assume the experience would be better with friends. So why show them an empty house full of cobwebs when you could show them a full beer cooler and a phone to call their friends?

Musical.ly was guilty of this too. But they didn’t even have the decency to handle the error like gentlemen. Just a blank screen with a CTA instructing me to follow all the no ones.

Vine’s solution was to look harder for your friends, allowing you to find friends through multiple networks — Facebook, Twitter or your contact list. AND, they front load some suggested users for you. I accidentally followed hundreds of people right off the bat, so they must be doing something right. I guess.

At the very least it probably gave hundreds of people a warm fuzzy when they scored a new mystery follower. ¯\_(ツ)_/¯

Just a couple last things before we move on to the next bucket. None of the apps in this category prioritized inviting friends but after some digging through settings, several of them did have some pretty interesting implementations of invites.

Most took advantage of the 98% open rate of SMS messages. And some streamlined that SMS invite experience still further by using Digits for Twitter as part of their signup process. (Both the .ly’s mentioned above used Digits.) And other’s, like Kik, use their own homespun version of Digits.

Below you can see examples of the prewritten text messages from Flipagram, Musical.ly and Wishbone as well as documented proof that I’m a failure at providing sustenance for my family.

It would be interesting to see how adding images plays into conversions here. Two of the three decided to add images of some kind even though there is reason to believe that there is some correlation between load times and link clicking and that might be starting you off on the wrong foot in places with low reception.

Even stranger, the image that Wishbone shares is just your avatar. I’m not sure why I would want to send a picture of myself to convince a friend they should download an app but maybe it sends a message of trust? It reads to me as automated spam in a way that the more ad-like image choice that Musical.ly made doesn’t for some reason though. (Let’s talk about it in the comments.)

The downside of inviting via SMS (from the developer perspective) will always be the 1:1 nature of it. There’s really no such thing as bulk inviting for SMS unless you do it as a group message and that marks you as a worse friend than the guy who always shows up after you carried the couch up three flights of stairs but before you’ve poured the drinks. Not the place you want to position your app before people have even downloaded.

The most interesting work around for this issue I saw came from Live.me. They still had the usual SMS, FB, etc. invites but they also used Facebook Messenger! I couldn’t find any open rate or conversion stats but this brings a whole slew of bonuses:

• Complete integration of your friend graph
• Allows users to send multiple 1:1 messages with one extra click each
• A brilliant solution to having to look for all your closest friends to send the message to by giving you a chronological list of the people you most recently Messaged
• Penetration equal to Facebook and
• What I would expect is probably a pretty high open rate. (Let me know if you have any data on this!)

As you can see, I only speak to people without names.

Good stuff. Stealing that. On to the last bucket.

Those that Prioritize Inviting Friends

There’s less science behind this statement than essential oils, but those that prioritize inviting friends on this list outrank all but the Big Four in their respective lists. (Granted, WhatsApp is a juggernaut in its own right.) Although, this does have some logic behind it. If Virality(k) = Invites(i) * the Percent of Conversion( c ), then you’re going to need some invites in there.

In fact, all three of the apps in this bucket do some clever things to try to have their cake and eat it too.

Marco Polo gives up a lot of real estate to convince you to invite your friends. The first page after signup is a full screen to convince you to connect your contacts and get busy inviting. All inviting is done via SMS, presumably to streamline things and avoid death by choice. They also have the top of the screen filled with “Suggested” friends to invite…and those friends appear to be COMPLETELY random. I’m not sure there’s any wisdom in that. If you’re anything like me then half your contacts list is people you don’t even remember meeting.

But then here’s the clever bit, since inviting via SMS is still a 1:1 proposition, they don’t rely on you to do all your inviting right away. They still take 18% of the bottom of your screen once you’re using the app to remind you over and over about all the friends you should be using Marco Polo with.

And, AND, eff that prioritizing finding friends BS everyone else was focusing on…Marco Polo just defaults you to following everyone in your contacts list who already uses Marco Polo, so the second you open the actual app part of the app you already have a lively page full of faces you trust. Brilliant.

WhatsApp takes a more all encompassing approach, allowing you to signin via Facebook (who own WhatsApp) and then treating you with a front and center plea to “Tell a Friend about WhatsApp” (why ‘Friend’ is capitalized and ‘about’ isn’t we can only guess. *shrug emoji*).

You’re then given a slew of choices for how you’d like to “tell a friend” — Mail, Message, Twitter, Facebook. Standard stuff.

Okay, I saved this one for last because, although I put it in the Prioritizing Inviting Friends bucket, Houseparty does a great job of marrying the concepts from both Prioritizing Finding Friends and Prioritizing Inviting Friends.

Perhaps because the folks behind Houseparty have been burned once before by building on top of someone else’s graph with Meerkat, this time around they’ve decided to create their own graph from scratch by only allowing signups via email and using their own version of Digits (although, I’m not sure what function the not-Digits part serves since they also force you to enter a password and email…unless they’re just forcing two-factor? Feel free to explain this to me.)

Immediately after signup you get the warm fuzzy of social proof by finding out you have friends using the app already via your contacts. Friends found. Check!

The danger at this point would be, if you as the developer now choose to ALSO prioritize invites, you’re committing to potentially doubling the length of your onboarding flow and losing new users before you even get to the good part. But Houseparty does something clever here.

They let you FEEL like you’re using the app by showing you the UI in the background, but the foreground presents you with a brightly colored notification compelling you to invite friends now. The onboarding is effectively done but the user gets to choose to take it a step further and in doing so, potentially improve their user experience and Houseparty’s download numbers. Win-win.

Welp, that was a lot of typing for me and reading for you so I’m going to bow out but I’d love to continue to discuss all of this in the comments.

Hope this helped! Let’s converse.