The Smartest Messenger App

iMessage, Google Allo, Facebook Messenger, Hangouts, Viber, Telegram, WhatsApp, Slack, Discord, Skype and last but not least WeChat. Obviously, I am talking about the most popular messenger apps currently on the market. Of course, the key question is: which one of those is the most popular? The answer to this question completely depends on the geographical location as can be seen in the figure below.

Most popular messenger app in every country (Android). The study has been conducted by Similar Web (April 2016)

Note that above data originates from April 2016. One month later, in May, Google announced its own brand new messaging app called “Allo”. This app has to compete with the “big boys” in this market: Facebook Messenger and Whatsapp (both owned by Facebook). According to the study by Similar Web, Whatsapp and Messenger are the #1 most popular mobile messenger apps in 109 and 49 countries respectively of a total of 187 examined countries. In other words, Facebook dominates this market with a share of 84.5% (on Android). On iOS, the default messaging app iMessage got a lot of attention in the last couple of months. The fact that Apple spent almost 20 minutes talking about all the new features of iMessage during the WWDC 2016 speaks for itself: messaging apps are a big thing these days. They get more and more features and become more playful. As an avid iMessage user myself, I cannot deny that iMessage that comes with iOS 10 is packed with all new crazy features. Balloons 🎈, invisible ink 🔏, confetti 🎉, drawing 🖌, you name it..

All in all, there has been going on a lot in the messaging industry recently which makes it even more attractive for reflection and comparison. Given the enormous barrier of entrance for new messaging apps, I think it is especially interesting to see how Google’s Allo will perform. Therefore, I will dedicate this article to the app with the tagline “a smart messaging app”. If you want to know whether this smart messaging app will also be smart enough to be a serious competitor to Facebook and Apple eventually, please keep on reading!


1. Excellence — Great Is Not Just Good Enough 🌟

1.1 Research 🔎

1.1.1 Feature Comparison 👉 👈
Although all messaging apps essentially have the same basic set of features there are some differences that make apps stand out from the rest. This difference in functionality can mainly be found in secondary features, such as extensions and stickers. As most messaging apps are becoming more and more advanced by incorporating additional features the differences also become more noticeable. Therefore, I have compared 5 popular messaging apps (iMessage, Whatsapp, Facebook Messenger, Telegram and Google Allo) on 55 attributes to map all the features (data here) and get a good overview. In particular, Allo’s functionalities are compared with its competitors: what features are unique and which are entirely missing?

Unique features of Allo
- Google Assistant: You can ask this virtual assistant a lot of things, for example what the weather is going to be for tomorrow, to set a time for 1 minute and you can even play some silly word games against it. Furthermore, you can ask for its assistant within messages with friends by simply typing “@Google”. Yes, with Allo there is always a listening ear in case you need it.

Allo understands this pain point and closes the loop by providing a Google Assistant which can pull in a list of things from Google, right inside the chat so that you have to change apps. This is pretty neat and hats off to Allo for solving a real use case. No other app is doing such stuff and Allo is a clear leader here.
- Udit Sajjanhar (Quora)
Things you can ask Google Assistant (@google)

- Advanced machine learning: by default “Predictive text” is available for all apps on iOS, though Allo has a feature called “Smart Reply” that is even more advanced. It not only predicts possible words and phrases based on real-time user input, but also takes into account the full context of the conversation. So it predicts what you want to say, already before your fingers have reached the keyboard. Even more so, it can use image recognition software to suggest responses when pictures are being sent to you (e.g. suggest “Ahh 😘” when you receive another cute cat picture).

- Resize text size: This feature may seem like a minor detail though users love it. Also, during Google’s I/O conference it was received with much enthusiasm: swipe up to SCREAM or swipe down to whisper.

Well.. Google’s Assitant didn’t really get it, haha :-)

Missing features of Allo
- Desktop client / multi-device support: Allo is only available for iOS and Android. Neither a desktop client (e.g. web.whatsapp.com) nor multi-device support is available (the latter allows you to continue the conversation you originally started on your phone on your tablet) .

Desktop support is totally missing..

- Share documents: Similar as all other messaging apps you can share photos and videos, however other file formats such as PDF are currently not supported. Please don’t force us to use e-mail for that purpose.

- Last seen status: iMessage and Allo are the only apps that don’t indicate the last time the receiver was online. However, both apps let you know when the users are typing (and thus online).

From top to bottom (WhatsApp, Telegram, Facebook Messenger)

- Calling feature: Hangouts and Duo already allow users to make voice and video calls. However, these features have not been incorporated in Allo in any way. For example, Apple’s iMessage automatically switches to Facetime when you make a video call within iMessage. Also, note the video and voice calling icons in above figure (WhatsApp and Messenger).

Translation: “Send X a WhatsApp message. What do you want to say?”

-Siri support: Since iOS 10 third-party apps can finally make use of Siri’s API. Contrary to iMessage (duh), Telegram and WhatsApp Allo doesn’t support it yet.

- Native SMS support: Contrary to iMessage, Allo sends messages from another randomly generated phone number (e.g. “3145”) in case the receiver doesn’t have Allo installed on his phone. Compare that with iOS: some users don’t even know the difference between an iMessage and a regular SMS thanks to the seamless integration within one app.

iMessage gained traction by replacing SMS — you just did what you used to do before iMessage existed and the messages went over iMessage instead of SMS if both people were signed into iCloud.
The way Apple usurped SMS for their own users and let SMS remain as a fallback for texting with everyone else was simply genius. — John Gruber

1.1.2 Customer Development 👨‍👩‍👧‍👦
Before I dive into the nitty gritty details, I’d like to show how Allo factually performs in comparison with its competitors on both iOS and Android. As you can see in the two figures below, Allo does a much better job on Android (Ranking: #15 vs #73, Star review: 4.2⭑ vs 3.5⭑). However, if you investigate a bit further you’ll discover that for this full selection of apps on average the Android alternative greatly outperforms iOS: the average AppStore rating is #22.6, whereas on Android it is #14.1 (the lower the better). The same holds for the average star review: 4.2⭑ (Android) and 3.9⭑ (iOS).

Average AppStore Ranking (NL): 22.6 | Average Play Store Ranking (NL): 14.1
Average #Stars AppStore (NL): 3.9 | Average #Stars Play Store (NL): 4.2

As expected, if you look specifically at the sentiment breakdown for the Google Play Store the positive sentiment is higher and the negative sentiment is lower (figure below). In other words, Android users seem more satisfied with Allo than iOS users. Moreover, this is clear from the relative percentage of 1-star reviews: 23.6% on iOS while only 7.8% on Android.

Red = negative sentiment, orange = neutral sentiment, green = positive sentiment (tool: Appbot.co)

For Google it is worrisome that there is downwards trend for the average AppStore star review (worldwide) whereas the star review in the Google Play Store stays about constant (see the trendlines in the figures below).

The variability in the left figure can be partly explained by the lower number of iOS AppStore reviews.

As both apps are literally copies of each other I wondered whether there might be a difference between iOS and Android users? You could argue that iOS-users have higher expectectations, because iMessage raises the bar higher and higher while Android users don’t know any high-quality and comparable pre-installed messaging app. In the second place, from the data it seems like iOS reviewers express themselves more actively compared to Android reviewers (i.e. they are less passive / more assertive):

So then the question becomes: what are users complaining about? A few topics come up over and over again illustrated with representative citations of customer reviews (sources: App reviews, Quora, Reddit, Tech blogs):

1. No native SMS support 
Everyone I messaged got a weird text from a weird number and already hates Allo and hates me for even trying to use Allo because of no SMS support.”

2. Another standalone app
Google: why not just add all these features into Hangouts and make that app truly awesome? We need a single app that runs on everything that has all the features in one package: SMS support, audio and video calling, web integration, etc.”

3. No desktop integration (web client)
“I remember being excited when I first heard about Allo, but then I found out there’s no desktop integration and all my hopes died. I don’t want or need another messaging app that’s tied to my phone and nothing else.”

4. No multi-device synchronisation
Definitely can’t become your new go-to messenger without multi-device support, including web access for your desktop/laptop.”

5. No chat back-up 
Google can’t afford missing out on basic features like chat backup when other apps are doing so much more.”

6. My friends and family don’t use it
Why would people change their messages app to this one? Nobody uses it. Same with Google+. Too late.”

7. No default End-To-End Encryption (privacy)
I am an avid Google user but care about my privacy. If you care about your privacy do not use this app until Google implements end to end encryption as default.”

As for this last complaint many reviews refer to the tweet below from Edward Snowden:

Although Google Allo has a built-in incognito mode (=end-to-end encryption) similar to Facebook Messenger and Telegram it is not set as default (see Amnesty International’s figure below). That is because many of Allo’s unique features don’t work in incognito-mode.

Finaly, if you look at the sentiment distribution of topics related to these complaints you indeed find the picture you would expect (predominantly negative):

Sentiment score for topics related to user complaints (based on Google Play Store)

All in all, we can conclude that there are 3 main causes the undermine Allo’s success: 
1. Firstly, there are some crucial missing features that competing messaging apps currently do offer (#1, #3, #4, #5, #7). 
2. Secondly, users want an all-in-one experience rather than switching between Allo, Hangouts and Duo (#2).
3. Thirdly, friends and family aren’t using Allo which is indirectly related to the former two causes (#6).

For now assume that the first two causes can be fixed on the short term. Then you may wonder, do Allo’s delightfulness factors — such as machine learning — convince users to choose for or even switch to Allo as their default messaging app? It is very hard to accurately predict that, but at least in the figure below you can see that Allo’s unique features receive overwhelming positive reactions, especially those reviews that contain the phrases AI and machine learning.

Allo’s unique features generally result in significantly more positive sentiments (Green = positive, Red = neutral/positive).

1.2 Measure ♥️📊

In the previous section several hypotheses were made about features we assumed most customers will enjoy. Nevertheless, this doesn’t mean that we should take that for the only truth. In fact, measuring is knowing, right? Hence, in this section we define success and introduce metrics to quantify to what extent objectives are reached.

1.2.1 HEART Framework 💙
Google’s HEART framework helps you determine the right UX metrics for your product by measuring the quality of user experience. HEART is an acronym for Hapiness, Engagement, Adoption, Retention, Task Success*. For Google Allo I have focused on the Adoption, Engagement and Retention categories as they demonstrate whether people download the app, regularly use the app and keep using it respectively.

*Note, the HEART sequence is not ordered chronologically. In fact, the processes take place in this order: Adoption > Task Success > Engagement > Retention > Happiness (ATERH).

Google’s Heart Framework (categories: Engagement, Adoption and Retention)

Explanation:

  • Engagement: Most people use messaging apps daily. The number of daily sent messages indicates how much consumers make use of it and thus how engaged they are.
  • Adoption: The number of app downloads gives only a limited image of user adoption (1/4 installed apps is never used (Tiongonson, J., 2015)). On the other hand complete account registration — and in particular the fraction of configured accounts — tells a lot more (the sign-up process requires several actions before you can start using Allo).
  • Retention: Especially for new product releases it is important that people not only try the product out but keep using it (i.e. are retained). That is why the formula in the metrics column in above figure calculates the percentage of users that have used Allo for at least two consecutive weeks.

Hero metric 👸🎯
From these three metrics the so-called hero metric can be derived. This is the one metric you should make better at the end of each month. As the retention metric (% retained users) inherently entails engagement and assumes adoption it is the one metric the whole team should care about, aka the hero metric.

Moreover, a large AND active user base (the latter is related to retention) is required for a messaging app to be successful. For example, think of Telegram which became extremely popular in a very short amount of time after rumours that Facebook would acquire WhatsApp. People massively switched to Telegram because they thought their privacy was at risk (figure below). However, after a couple of weeks less and less people became active on the messaging platform and people returned to WhatsApp. That is also why you find in the figure by Similar Web (at the beginning of this article) that 2.5 years after the acquisition WhatsApp has a much larger share of the pie than Telegram.

Google Trends report for the search term “Telegram”

2.2 OKRs 👌🎛
Objective and Key Results (OKRs) can be used to quantify the progress to reach an objective. To achieve any objective several substeps have to be carried out. Let’s say this quarter we want to improve the hero metric we’ve just determined, then our objective can be “Improve weekly retention rate”. The key results contribute to reaching this objective. Then, all these criteria (key results) are independently rated and its average is taken to derive the overall objective score.

The sample figures on the right are only to illustrate the principle of OKRs (Objective result = AVG(Key_Results))

2. Distribution — Is Everything 🚗

Messaging apps fully rely on network effects. After all, from the user data it turns out that if your friends and family don’t use the app there is simply no reason to stick to it. In other words, next to excellence distribution is everything.

“It must be excellent otherwise you cannot distribute it. If things are just excellent no one will hear about it unless you distribute it.” — Robert Gaal

In the previous section we particularly looked at the first criteria: how can Allo be excellent. Now let’s take a look at the second criteria: how can we make sure that others will hear about how great Allo is and actually start using it?

2.1 Status quo 🗓
Seriously, the latter still requires plenty of work. Although my contact list mostly consists of teens and twenties, only a fraction of them have actually installed Allo. Let alone use it as their default messaging app. In fact, of my entire address book — which counts 314 contacts of which I have any phone number — only 18 have installed Allo (5.7%) as of today.

“Invite” indicates that the phone number associated with a contact person has not been used to set-up Allo so far.

Of course, the Allo app has only been publicly available for about two months but if you consider the search behaviour on Google Trends (duh..) you’ll find that there isn’t much going on except for two peaks. The first peak round May can be explained by Google I/O and the second one by the release date of the apps. Some folks argue that Google already failed by postponing the release until a couple of months after the official announcement during Google I/O.

They blew it by not just releasing it the day it was announced. It’s going to have a soft launch, garner a minimal user base and then just taper off with people going back to SMS, Whatsapp and FB Messenger. (Reddit Allo thread)
Google Trends report for the search term “Google Allo”

2.2 Search ads 🔍
Okay, they may good have done a better job at the time but if I were Google what actions would I undertake to improve the distribution channels right now? Again, let’s consider the facts from nota bene Google’s own marketing department:

  • 40% of smartphone users browse for apps in app stores.
  • Search is a major source for app discovery, according to their research: One in four app users discovers an app through search.

In the end it is also concluded that “search ads are among the most effective ad formats for driving app downloads”. So it seems like an open door to invest in search ads.

2.3 Default messaging app 📢
As some users already suggested Google can greatly accelerate adoption by making Google Allo the default messaging app on Android with a new software update. In essence, Google can impose people to use Allo. But that is only part of the equation. After all, it is only possible on Android, not on iOS. In one of his blogposts John Gruber describes in great detail how he thinks this will work out on the long term:

It’s no coincidence that two of Google’s major Android initiatives this year are Allo and Duo, their answers to iMessage and FaceTime. I don’t think it’s going to work. iPhone users on the Google ecosystem might install Duo and Allo, and those who switch to Pixel phones will have them installed by default. But I don’t see why iPhone users on the Apple ecosystem will install either Duo or Allo in large enough numbers to make a difference. Anyone who switches to a Pixel phone from an iPhone is still going to miss iMessage and FaceTime.

2.4 Growth hacks 📈
So on top of the previous point we can definitely benefit from some growth hacks to make it work.

A famous and common growth hack is the timing of requests for an app review to get a higher star review. For example, after a happy moment (e.g. when you have won a game). Not for nothing you want a high star review as it is a critical factor users consider when deciding to download an app. Since the Allo app on iOS has only 3.5 stars, there’s work to be done.

I bet you have seen those yellow Snapchat codes by now. It is an easy and fast way to add Snapchat friends. But it’s also some kind of growth hack: in Instagram and Snapchat, the most popular social media apps right now, it is rather complicated to share hyperlinks (URLs). However, sharing images is a no-brainer.

“My Code” — Shareable Facebook Messenger Code

You may not know it, but Facebook (Messenger) has something similar to Snapchat codes called Messenger codes (see figure on the right). I think Allo shouldn’t miss this trend and build a similar sharing feature. Especially given the fact that you can currently only add new contacts by knowing or memorising their phone numbers (Gmail accounts won’t do).

Image by Pixlicity

Lastly, it should be noted that Google has already a very smart growth hack in place using its inherent channels: once Android users who haven’t Allo installed receive a message which is sent from Allo a pop-up card will appear that allows them to install Allo with only one click (figure on the left). Moreover, even without installing the app users are able to respond. In other words, bring the experience to non-users.

2.5 Habit forming 🚬
Cigarettes? No, I am just kidding. And in case you wondered, I don’t smoke. However, habit forming is very important for apps to ensure they are used not just once but again and again. After all, 65% of people stop using apps after only 3 months (Adrien Montcoudiol). According to Google the solution to this problem is apparently straightforward: consistently prove the value and utility of your app. This can be achieved through creating habits. Interestingly, Google’s Assistant has a tool called “Subscriptions” that will send you daily push notifications about the news, poems, funny videos and a lot more cool stuff you want. In other words, the perfect habit forming tool. Therefore, I believe the subscriptions part should be incorporated in Allo’s onboarding process. Don’t just ask for only the requisites, ask for a small time investment from your user that will result in a personalised user experience (e.g. similar to Pinterest and Blendle).


Exponential Growth 📈

One thing is certain. Google Allo needs exponential growth to catch up with its competitors. Robert came up with the following expression for that:

E^G = E x D

It implies that exponential growth (denoted by E^G) can only be achieved through excellence (E — section 1) and combining that on a equal level with distribution (D — section 2). Note the multiplication symbol (x) which emphasises the importance of both E and D. Both factors are crucial. One factor can be huge, but if the other one is nearly zero exponential growth will never happen.

Roy Klaasse Bos

For more personal details:
🌐 Personal website
💼 LinkedIn

I think you may also enjoy this piece, in which I extensively research and evaluate the iOS app of the Dutch Netflix for journalism: Blendle.

N.B.1. Apple’s Messaging app is actually called “Messages”, however to avoid confusion with its competitor Facebook “Messenger” I have decided to call it by its previous name: iMessage.

N.B.2. Google’s #1 value is “Focus on the user and all else will follow.” I fully believe in this philosophy an therefore I have tried to back up my findings with user data whenever possible. To illustrate this, I have regularly included relevant tweets or citations.

N.B.3. Thank you Robert, for your great talk during the YC-Summit. Not only did it inspire me to learn more about Product Management, I also used it as the “backbone” of this article.

N.B.4. Another word of thanks for Nick who has suggested me to write this article about Google’s newest messaging app Allo. It took a lot of time but it was pretty fun :-)

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.