BUILD

or

BUY
Evaluating in-house and outsourced
solutions for Mobile Growth Stack

The mobile growth stack Ebook from GetSocial

 “Do what you do best. Outsource the rest”
https://www.marcusbuckingham.com/

Most companies already have a mobile product in existence before they

start thinking seriously about growth. That’s where the mobile growth
comes in.

The mobile
https://blog.getsocigrowth
al.im/implementing-the-mobistack
le-growth-stack-2017-wiis aal/framework for app developers, founders
th-getsoci
and marketers to develop and evolve a growth strategy for mobile. It
includes every little aspect of what an app needs to focus on, for
growing sustainably in a competitive market: User acquisition, user
engagement, retention, monetization and of course, insights and
analytics that are important for growth.

“Colored blocks represent the components of growth stack covered by GetSocial technology”
 BUILD OR BUY - 2

Now when it comes to implementing the mobile growth stack, the

key drivers are user acquisition, engagement and retention. Once
you have nailed the three, you have achieved half your growth
goal already and your insights are shining with impressive
numbers.

The mobile growth stack describes an entire range of possible

tools, techniques, and channels that need to be used by an app
developer or marketer. You don’t have to implement everything. It
is much better to start with basics, understand performance and
iterate on the results.

We believe the most important aspect of an app's growth is

https://blog.getsocial.im/9-ways-to-build-a-viral-loop-in-your-game/

Virality (Word of mouth marketing) i.e. making it frictionless for

your users to evangelize your game with their family and friends. It
is important that Virality becomes a part of the core loop in your
app, rather than an afterthought. The most important aspects
here are:

• Reliable deep linking that creates seamless transition from the

invitation to engaging content

• Attribution analytics to find influencers and understand the

highest performing channels

In this ebook, we’re going to discuss what you need to consider

when implementing these critical aspects of the mobile growth
stack for your app. We’re addressing the buy vs build when it
comes to finding the right solutions to implement this growth
framework.

But as with most build vs buy decisions, the decision to develop

something in-house boils down to a strict evaluation of the
current and future business needs. The decision is also influenced
by the available in-house expertise, core competencies, time to
market and the overall cost of R&D to build and maintain these
solutions.
IN-HOUSE VS
OUTSOURCING
SOLUTIONS
 BUILD OR BUY - 4

The build vs buy debate is a healthy one to have at any stage of a

product development - be it a company, an indie developer or app
sponsored by publishers.

The minimal viable feature set necessary for any app changes over
time. Since the level of functionality in the initial stages is
minimal, the success of an app heading towards millions of users,
requires robust and more engaging features.

TI ME E X PE RT I S E COSTS A NA LYT ICS T ECH NOLOGY

This ebook gives you insights into what you need

to consider when contemplating the build vs the
buy decision for your app(s).
 BUILD OR BUY - 5

Time

The study, entitled “Doomed From the Start? Why a Majority of

https://www.geneca.com/blog/software-project-failure-business-development

Business and IT Teams Anticipate Their Software Development

Projects Will Fail” states that 80% of product teams claim they
had to spend majority of their time on reworking a product. This
includes technology and platform updates, change in project
scope, unclear requirements etc.

For instance, a typical referral marketing MVP can take anywhere

from 4 months to over a year to build.

If you’re a company or a developer focusing on faster growth, you

don’t have so much time to wait! Re-evaluate the following before
deciding to build something in-house:

? Does your tech team have the time to build products alongside
your core products?

? How long will it take for them to build the required product
solutions?

? How much time and effort is required to maintain these products

in lieu of ongoing platform and technology changes?

? What will be the impact of building something in-house on the

roadmap of your core products?

You don’t really want to spend your resources building a solution

re-inventing the wheel that already works.
that is just https://blog.codinghorror.com/dont-reinvent-the-wheel-unless-you-plan-on-learning-more-about-wheels/
 BUILD OR BUY - 6

Expertise

While an engineering team might be able to build 80% of the

solutions available on the market, it is a fact that they are not
experts at everything. It might take them some time to
understand what is required for user engagement and marketing,
how they can implement it and most importantly, to optimize it
for a seamless experience.

So before you decide to build something in-house, ask yourself:

? Does your engineering team have experience in building the

entire solution stack.

? How long will they take to understand the technological

requirements, before building something?
 BUILD OR BUY - 7

Cost

One of the most crucial parameters of deciding to build in-house

vs buying a solution, is the cost of engineering. Sometimes in their
zeal to optimize their spend, most companies end up developing
solutions in-house, increase their overall technology cost.

Here are a few cost calculations you need to make in order to

decide:

Cost of building the Time required to N° of resources Cost per

tech in-house
= build a technology x required x resource

Initial costs to get Cost of Potential costs on

Cost of infrastructure = a highly available, + maintenance & + app growth
redundant and improvements
secure product
system up &running

Opportunity costs for scope or schedule creep i.e. the cost to the organisation for forgoing the
opportunity to allocate its resources to other projects.

Simply put, ask yourself one thing after you have the costs in
hand:

? Can you afford to hire experts to build and maintain another

product in-house?
 TECHNICAL
CHALLENGES
“Every app has a different need.
The challenge is to find the right solution.”
 BUILD OR BUY - 9

We’re going to break down the technical challenges you need to

consider into smaller segments to walk you through better
decision making at each level of product development.
In order to build some of the key components of the mobile
growth stack the bare minimum you need to consider is:

1 - Support various deep linking technologies in popular versions

of iOS and Android (iOS Universal Links, Android App Links, URI
schemes, etc).

2 - Create links that work across every browser.

3 - Create links that work across different versions of iOS and

Android.

4 - Create links with complete tracking and attribution, including

mobile web to app and desktop web to app attribution.

5 - Cater for moving target i.e. constant updates to the platforms,

browsers, messaging and social apps etc.

In additional to the above, you also should:

6 - Build marketing dashboard that provides useful and

actionable insights by collecting and analyzing data to make
data-driven decisions

7 - Prevent installation fraud

8 - Native integrations with social media and messaging apps

instead of using native share sheets.

9 - Setup manual and automated QA

10 - Localize content to optimize conversion

Let’s dive into each of those technical challenges deeply.

 BUILD OR BUY - 10

1 - Various deep linking technologies

When it comes to marketing your app, be it content marketing,

influencer marketing, performance marketing or content sharing
to drive word of mouth it is imperative that you build links that
works across every browser, platform, app stores and complies
with various deep linking standard.

Nowhttps://blog.deep
getsocial.im/deep-linking-what-ilinking ist/ an important aspect of mobile apps to
s-it-why-does-your-app-need-it-and-how-to-use-i

onboard their users and keep them engaged effectively.

While looking for a solution, you need to consider URI Schemes,
iOS Universal Links, Android App Links, Facebook App Links and
Android Chrome Intents. You have to support all of them to make
onboarding process seemless on any platform.

Deep links are the most effective way to offer users a smooth
in-app experience. The deep linking standards mentioned in the
paragraph above are used to bring a user directly to a piece of
content inside the app, that is contextual to his journey.

But implementing deep links on both Android and iOS on your

own, can be time consuming and challenging.
 BUILD OR BUY - 11

Apart from the time consuming development, none of these

standards are supported on every platform or OS version:

iOS < 9 iOS > 9 Android < 6 Android > 6

URI Schemes

Universal Links

Android App Links

Chrome Intents

R E CO MM E N DAT I O N :
If you’re building the deeplinking tech in-house, ensure that you support ALL the
above standards and additionally also add support for Facebook App Links.
When deciding to use third party service for the same, ensure the service supports all
the aforementioned standards.
 BUILD OR BUY - 12

2 - Links that work across every browser

You want to create a link that is smart enough to take user to the
app if it is installed, otherwise open Google Play on Android, or
App Store on iOS. You also want to ensure that you don't lose the
attribution data on desktop and are able to pass content through
the app stores post install.

You want to make the process seamless with as few clicks as

possible to install the app and deeplink to the appropriate
content that drove the install in the first place.

Most links clicked by the users go through a web browser, with the
exception of iOS Universal Links and Android App Links when an
app is already installed. In process you'll discover a "forest" of
mobile browsers. On Android only there are over 5 popular
browsers. And it gets worse as every manufacturer tends to ship a
phone with its own default browser.

Each of these may behave differently in the way they redirect

users to the Google Play store. And a bunch of in-app browsers
from Twitter, Facebook and, other messaging and social apps. To
get a better understanding on the breadth of Android browsers to
consider, check this presentation.
http://slides.com/html5test/the-android-browser#/5

It is therefore imperative to address the different quirks of each

browser before deciding on a solution.

RECOM ME N DAT I O N :
When deciding to build in-house ensure you prioritize a list of browsers (including
various browser versions) that you need to support at minimum to make sure
deeplinking to app and fallback to appstore works as intended.
Also ensure on desktop you implement a feature for users to sms the link to
themselves and/or QR code so that you don't lose the attribution and tracking
information. When evaluating third party services, make sure your QA team
validates the above.
 BUILD OR BUY - 13

3 - Links that work across different versions of iOS

and Android

In addition to different browser versions, operating systems have

different versions as well.

Let’s take iOS as an example first. The usual way to implement

deep linking before iOS 9 was via URI schemes. It was easy to
implement, easy to support on both JS and native app sides, and
it just worked. iOS 9 introduced the concept of Universal Links:
they are more secure, look familiar on both websites and native
apps.

But there is an extra requirement: you have to host a special

verification file on your HTTPS domain. iOS10.3.3 introduced a
security popup in Safari browser, which broke the fluent deep
linking flow. Nobody knows what else will Apple introduce in the
future.

The story with Android is fortunately a bit simpler despite the OS

fragmentation. Even though Intents work since the inception of
Android, Android App Links were introduced in Android 6
Marshmallow and must be incorporated for better user
experience.

R E CO MM E N DAT I O N :
This is fairly straightforward if you already support the deep linking standards and
the various browsers mentioned above. You however need to make sure you are on
top of all the changes introduced by Apple and Google that affect your existing
solution.
 BUILD OR BUY - 14

4 - Links with complete tracking and attribution

Let’s imagine you want to send name and some data attached to
the link from the sender invite to the recipeint device after install
(e.g. username and game session identifier).

The problem with the install flow is that you can’t attribute invites
easily. Stores are doing a poor job of sending the data through
them.

• Google Play does better job with its referrer URL parameter and
the corresponding INSTALL_REFERRER Android broadcast. But
from what we see, there are situations, when broadcast is not
delivered.

• Apple App Store has no proper method to send specific

parameters after install to the app on the first launch. The only
solution here is to “guess” somehow if the user who clicked the
link is the one opening the app.

So is there a good way to mitigate these problems and attribute

every install?

Yes: Device fingerprinting.

Device fingerprinting, is the most common and widely used

identification method to match link clicks to installs for the
purpose of attribution. It uses publicly available parameters (i.e.
device model, device screen dimensions, OS version, platform, IP
address, carrier, and many more), to form a digital fingerprint of
 BUILD OR BUY - 15

the device.
Fingerprinting is a probabilistic model (i.e. there is probability of
two devices with same fingerprint) and therefore it’s not 100%
accurate. However when clicks come from mobile web (i.e. when
invite links are shared outside of the app via email, sms,
WhatsApp etc) deterministic identifiers like Google Play Referrer
and device identifiers such as IDFA or Google Advertising ID are
not available for attribution.

MOB I L E W E B TO A P P AT T RIBUT ION

The biggest challenge is to uniquely identify mobile users. Since

they access content through both native apps and the mobile web
(disparate domains), you need separate identifiers for each.

• There is no single standard and technology for tracking and

attributing installs

• In addition, there is also no single standard and technology that

can detect if the user has your app installed. If not, it redirects
them to the appropriate app store.

• Most links don't function without appropriate Open Graph tags

https://dev.twitter.com/cards/overview
(and twitter cards)

• App Store drops any data you try to pass through install, so you
have no information about where new users are coming from.
Utilize device fingerprinting technology to facilitate this.

• Google Play on the other hand, sometimes passes referral post

install through their INSTALL_REFERRER broadcasts, however
these are highly unreliable. These referrals also don't work when
user is coming from Chrome browser.
 BUILD OR BUY - 16

• Attribution window, defined as the maximum duration of time

between the link click and install needs to be considered. The
longer the time gap between a click and install, the greater the
chance that the user would change a setting in his/her device,
thereby creating a new fingerprint (this is especially true with IPs
as mobile users constantly change their location). That’s why the
attribution window is short, usually 2 hours.

• Facebook’s App invites technology has it own tracking and

attribution system and requires app developers to integrate their
SDK.

DE SKTO P W E B TO A PP

When users click on your links and install your app from the web
version of the stores you lose all the attribution and tracking since
the web version of stores does not pass the referral data when the
app is installed on the mobile device.

RECO MM E N DAT I O N :
For Mobile web to app, you need to build a proprietary device fingerprinting
technology and ensure you address all of the challenges above.

For Desktop web to app, one of the ways to mitigate the issue is to allow your users to
SMS themselves the link to continue the install process from their mobile device.
Setting up such SMS technology involves third party integration and costs for sending
SMS. Alternatively, generate QR codes.

When considering a third party service for the same, ensure your QA team validates
all of the above.
 BUILD OR BUY - 17

5 - Moving target

On implementation of a solution, there is always a moving target.

You always need to be on top of the constant updates to the
platforms, browsers, messaging and social apps etc and also cater
for the undocumented changes in the process. Look for solutions
that offer you stability and quick response times rather than
de-focus from your core developments.

Let’s look at some of the examples to validate the point:

1 - Introduction of universal links with iOS 9:

In June 2015, Apple introduced Universal Links to replace the
previous deep linking mechanism through URI schemes. With
URI schemes, the routing logic was in Safari, and there was no
way to check if the app was installed or not. Developers had
to call the URI scheme 100% of the time to check whether
there was an app that can handle the scheme and fallback to
the app store otherwise. Universal Links was introduced to
address this.

2 - End of URI Schemes with iOS 9.2:

Within just 6 months of introducing Universal Links, Apple
made changes to Safari breaking URI schemes where user
was now redirected to the App Store regardless of whether or
not the app is already installed.

3 - Google release App Links with Android Marshmallow (6.0) :

Similar to Universal Links on iOS, Android App Link is simply a
way to turn your existing website links into app links where if
a normal HTTP/HTTPS link is clicked and the corresponding
app is installed, it will open immediately.
 BUILD OR BUY - 18

4 - iOS 11 changes:
With iOS 11 Apple removed Facebook and Twitter
integrations. iOS apps so far had the ability to sign into
Facebook and Twitter directly, making it easy for the users to
share updates on their social profiles. Simply put, the two
platforms will now be treated like any other apps on the
store, having to follow the same iOS sharing extensions as
used by others. There were some other notable changes,
which you can read here.
htps:/blog. etsocial.im/se -how-ios-1 -changes-impact-your-ap -sgrowth/
https://blog.getsocial.im/see-how-ios-11-changes-impact-your-app-s-growth/

5 - Device identifiers get locked down in Android 8.0:

Google has locked down a number of ways for apps and other
third-parties to identify the device. The Wi-Fi MAC address can
now be randomized and "Android ID" value is now scoped per
app and per user.

RECOM ME N DAT I O N :
Make sure you are on top of all the technological changes (browser, OS, deep linking
standards, messaging and social apps etc) that affect your existing solution and
make necessary changes and/or improvements. When considering third party
service ensure the company has a history of product updates that comply with these
technological changes.
 BUILD OR BUY - 19

6 - Marketing dashboard that provides useful and

actionable insights

Right from the point of soft launch, you need to dig deeper into
analytics, to get actionable insights. This data helps you
understand the market response, identify the growth
opportunities and focus on the product value adds and features.

Building a marketing dashboard to analyse cycle time, k-factor,

invite funnel, campaign performance, retention per marketing
channel, cohort and segmentation analysis etc can be
demanding. Once you have insights into what you need, ensure
that there is some form of marketing automation in place to
encourage user behaviour for the growth of your app.

RECOM ME N DAT I O N :
Chances are you already use some third party analytics services (GameAnalytics,
Google Analytics, Mixpanel, Amplitude etc) or have an internal BI system in place.
Ensure you collect and process the data necessary to make informed decisions, and
create/update your in-app marketing tool to run targeted campaigns.
Alternatively, find a third party solution that provides you with such key insights and
lets you act upon those insights.
 BUILD OR BUY - 20

7 - Fraud detection and prevention

When you are running in-app referral programs, it is important to

focus on fraud detection and prevention simultaneously.
Even the top most apps like Uber have lost millions on their
http://www.businessinsider.in/This-Guy-Got-50000-In-Credit-From-Uber-And-Ended-Up-With-Free-Uber-Rides-For-Life/articleshow/33533647.cms

referral programs. But you don’t just lose out money, you then
lose the resources that are needed to rebuild the product and
re-strategize the campaigns.

Following are some of the most common frauds that you need to
cater for:

1 - Invite sent from and installed on same device

2 - High volume of installs from same device

3 - High volumes of installs with short click to install times

RECOM ME N DAT I O N :
Build a solution that can identify and prevent such fraud, specially when they
impact your in-app economy or ensure the third party solution caters for the same.
 BUILD OR BUY - 21

8 - Native integrations with social media and

messaging apps

to a report bycom/the-messaging-app-report-2015-11?IR=T
BI Intelligence, messenger platforms
According http://www.businessinsider.com/the-messaging-app-report-2015-11?IR=T
http://www.businessinsider.
such as Facebook Messenger, WhatsApp, Line, KakaoTalk, WeChat,
etc have surpassed social networks in terms of Monthly active
users. Given the rise of messenger platform, the mobile growth https://mobilegrowthstack.com/the-mobile-growth-stack-2017-edition-7fbac938420d

stack 2017 has been updated to include them as an additional

https://mobilegrowthstack.com/the-mobile-growth-stack-2017-edition-7fbac938420d

channel for growth.

According to the mobile growth stack, messaging apps provide

opportunities for apps to reach new and existing users through
content sharing, direct messaging and invites/referrals.
 BUILD OR BUY - 22

Both iOS and Android platform provide native share sheets to

enable your users to share your app with their family and friends
through messaging apps and social networks.

However these share sheets are cluttered with apps that are not
relevant for sharing e.g. notes, dropbox, google drive, evernote,
reminders app etc. and has a tremendous impact on user
experience.

RECOM ME N DAT I O N :
Provide your users with an optimal user experience by natively integrating with
popular social media and messaging apps. Having such native integrations will
further allow you to gain insights on the performance of each invite channel and
optimize your invite funnel.
https://blog.getsocial.im/user-acquisition-funnel/#acquisition
 BUILD OR BUY - 23

9 - Quality Assurance

2 platforms, at least 6 versions of Android, 4 iOS versions, some

messaging and social apps’ integrations, and a bunch of browsers
is very conservative list of what you’ll have to continually test
against to make sure your deep linking, attribution and referral
tools are working as intended.

RECOM ME N DAT I O N :
In addition to setting up test automation, set up a QA team for manual testing as not
all flows can be automated. Test automation is a big investment, and absolutely
necessary to ensure the technological changes (as mentioned in sections above)
does not break your solution.
 BUILD OR BUY - 24

10 - Localisation for a global market

For an app to be successful, it needs to offer high level

personalization to its users. The goal is create a seamless in-app
experience for the user, and localisation is one of them.
One of the things to keep in mind is that there are more than 40
https://support.google.com/googleplay/android-developer/table/4419860?hl=en

languages that the app stores cater to and for better user
https://support.google.com/googleplay/android-developer/table/4419860?hl=en

experience you need to ensure your app’s landing page (that links
to the app store) is also localised.

Localisation for:
In-app content - app copy, in-app referrals, in-app promotions etc.

RECOM ME N DAT I O N :
Optimize your screenshots, in-app content, invite and sharing messages etc,
prioritise the target geographics you want to launch your app in and then work
towards localising the content. As a ground rule, don’t trust the Google Translator.
 ?
FINDING YOUR
ANSWER
 BUILD OR BUY - 26

We wanted to make our ebook an easy and

answerable piece of content that could instantly
lead you to an answer, and end the ongoing buy
vs build debate.

Whatever your answers are, the result needs to

be a consistent app growth that is in sync with
your project timelines.

As a general advice, we believe you should continue focusing on

building the core product, before venturing into building another
technological solution. This helps you build faster and gets
industry expertise onboard, without added hiring costs.
 BUILD OR BUY - 27

For instance GetSocial enables you to implement over half of the

https://www.getsocial.im/
mobile growth stack within a week. It offers you the best of tools
that you need for app growth, without the extra headache of
developing and maintaining them yourself. As the saying goes
“Don't Reinvent The Wheel, Unless You Plan on Learning More
About Wheels”.

https://www.getsocial.im/customers/

Acquire Connect Engage

Expand your viral channels Build your own social graph Build in-app communities

Our customers benefit from:

7 x 25 % 15 %

HI GHER D3 0 R E T ENTION CONV ERS I O N O N REFERRA LS I N C REA S E I N LT V

 BUILD OR BUY - 28

https://www.getsocial.im/customers/
GetSocial is
used and trusted by some of the top companies in the industry.

“GetSocial is the foundation layer for us to put the club

into Gear.Club.
For a small yet ambitious independent studio like Eden
Games, the GetSocial tech and their marketing-friendly
dashboard save up a lot of development time in building
our grand vision on a global scale.”

Pascal “Scal” Clarysse, CMO at large at Eden Games

Still not sure if you should buy or build for the mobile growth stack?
Let’s get you talking to an expert.

mailto: sales@getsocial.im
G E T FREE CONS ULTAT ION