WWDC2015 Session 303

Transcript

X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
>> TRYSTAN KOSMYNKA: Thank you.
Thank you.
My name is Trystan Kosmynka.
I'm in iTunes Engineering.
Myself and my colleague David
Hopkins are excited today
to help you get the most
out of App Analytics.
App Analytics today is publicly
available for all developers
with a valid developer account.
Sales admin or finance
role inside iTunes Connect.
There are no restrictions to
your use of App Analytics.
It's available to
all developers.
Because there are
no restrictions
and it's publicly available, we
are excited to tell you that as
of today it is no
longer in beta.
[Applause]
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
[Applause]
Thanks. So why does this matter?
Why do we care about
App Analytics?
Why is Apple excited to
bring this product to you?
We care because App
Analytics provides answers.
There are a lot of big
questions that you have
as app developers or marketers.
Are customers finding your app?
Are they installing the app?
Are they purchasing goods
inside of your application,
and are they returning?
Our goal with App Analytics
is that we can provide some
of these insights so you
can make great decisions
going forward.
Through these answers, we also
reveal missed opportunities.
A good example of this is if
your application is available
across all storefronts but
you have not yet localized
in a particular country,
with App Analytics what
you can see is you can see
that you have really
high App Store page views
in a particular country but the
conversion rate isn't so good.
If you maybe localize your
store presence to that language,
you make it more
accessible to more customers.
This is the type of data that
we deliver with App Analytics.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
This is the type of data that
we deliver with App Analytics.
With all of these,
our ultimate goal is
to help you build a better app.
You are here to learn tips
and tricks from experts
so you can build a better app.
That's on the development cycle.
We want you to know that your
responsibilities do not stop
after launch.
There is a whole new set
of things to focus on,
marketing your app, advertising
your app, adding more content,
new versions, maybe
there's price changes.
Your Store presence
is important as well.
All of that goes back into your
product development life cycle,
and it's really important
to creating a successful app
in a successful app business.
We do all of these things
with Analytics data.
Analytics has three
key data sources.
We have App Store data,
sales data, and usage data.
These three data sources are
combined to tell a single story
across acquisition
to engagement.
We answer those key
questions with this.
Are customers finding your
app, are they purchasing it,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
Are customers finding your
app, are they purchasing it,
and are they using
it and returning?
To understand these
measures better,
let's take Trip Guider 2.0.
We updated this this
year for WWDC.
Don't download it.
It is not real.
Here we are seeing
our App Store views.
What App Store views are is a
potential customer visits your
product page in the App Store.
That triggers an App Store view.
This is an example
of our Store measure.
Next ,we have app
units and app sales.
This is an example of
our sales measures.
In this example, I have
$2.99 for Trip Guider.
That's going to generate
$2.99 of sales.
This is what the customer
spent on your app.
App units registers
the transaction,
a single app unit would
be present in this case.
Let's move along to
the usage measures.
We have installations.
Installations is
different than app units.
This is a common
question we get:
What is the difference
between these two?
If I was to purchase Trip Guider
for $2.99, that's one app unit,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
If I was to purchase Trip Guider
for $2.99, that's one app unit,
but it may generate
multiple installs.
It's one to many.
I could have purchased
it and installed it
on my iPhone and my iPad.
That generates two
installations.
Additional usage measures that
we have available: sessions,
active devices, and active
devices in the last 30 days.
What is important with these is
that it is again one to many.
I may have opened
this app, Trip Guider,
and visited Pigeon
Point in the morning.
That's one session.
I came home and shared
the experience
with family and friends.
That's two sessions.
But all of this was
from my iPhone,
and that's one active device.
So sessions is an overall
idea of engagement,
and active devices is, how
large is your device base?
Active devices in the
last 30 days is the same
as active devices; however,
the time period is longer.
I'm not using Trip
Guider every single day.
But if I used it in the
last 30 days, I register a 1
in the active in last 30 days.
The last measure
we're going to share
with you is in-app purchases.
This is again a Store
or sales measure.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
This is again a Store
or sales measure.
In-app purchases is like units
where it's the transaction
itself, not the price.
In this case I bought Northern
California maps for 99 cents.
What this will do is
it will take 99 cents
and adds it to the app sales.
So it's app sales and
in-app purchases combined,
but it will generate a1 for
the in-app purchase measure.
Combining all these measures,
we get to tell that story
of acquisition to engagement.
App Store views all the way
down to in-app purchases
gives you the idea
of the entire customer
life cycle.
That's great.
But how do you actually
get started?
We have good news.
This is probably the
easiest session in terms
of work generated for
you, the developer.
There is nothing you need
to do to get started.
No code you need to write.
There is no SDK that
you need to integrate,
no iOS framework
you need to hook in.
This is something completely
built into the operating system
and that you get for free.
You have to have a
valid developer license.
Sales admin or finance role.
Not a lot involved to
get going with this.
It is available for the
App Store, iOS apps only.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
It is available for the
App Store, iOS apps only.
And it's available
for iOS 8 and above.
That means that data
is available
when it is generated
from apps on iOS 8.
Another important point is
that privacy is built in.
With App Analytics, we
wanted to provide developers
with a powerful functionality,
powerful feature and tool,
and we wanted to do so in a way
that we would not
compromise customer privacy.
This is built all the way from
the operating system to the UI.
iOS has privacy constraints
built in.
This means that our usage
measures, customers are prompted
with this display
where they can choose
to share sessions
and active devices.
We believe that customers
should have a choice,
and this is built
into App Analytics.
Data is available
in aggregate form
and not personally identifiable,
it's completely anonymous.
What this means for
you as the developers,
in the App Analytics
UI you see counts.
You do not see the raw data.
It is all available
in iTunes Connect.
It's available today,
it's no longer in beta.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
It's available today,
it's no longer in beta.
Once you log into
iTunes Connect,
you'll see the App
Analytics module show up here.
That's it for the
static part of this.
I'll turn it over
to David Hopkins,
and he'll show you an
overview of some key features
within App Analytics
in an actual demo.
[Applause]
>> DAVID HOPKINS: Hi, I'm
David Hopkins, and I'm going
to give you a quick demo of
the App Analytics product.
When you log into
iTunes Connect,
you'll see App Analytics
alongside your other modules,
such as Sales and Trends
and Users and Roles.
When you click on App Analytics,
you will be taken
to the app list.
That's the entry point to
the App Analytics product.
From here you can see a
snapshot across all of your apps
across four key measures,
again measuring all the way
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
across four key measures,
again measuring all the way
from acquisition
with App Store views
down to engagement
with sessions.
You'll see the value for
the current period as well
as a percentage change
from the previous period
to help provide some
context as to whether
that value is good or bad.
If you hover over the percentage
change, you'll see the value
for this period and the value
for the previous period as well
as the date ranges
associated with each.
The default time period
here is the last 30 days,
but you can change which
time period you're looking
at from the date
picker in the top right.
You can select from
any predefined ranges
or select a custom range as well
as selecting any
day, week, or month.
Now I'm showing you the app list
where you can view
all of your apps.
Let's dive into one of the apps.
For those of you who build lots
of great apps, the easiest way
to find your app in
here is to search.
We can search for
our app Trip Guider.
When we click on the app
Trip Guider, we will be taken
to the app overview page.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
The app overview
page is designed
to give you a high-level
overview of your app.
The first thing you are going
to see is your all time data
for your app since it has
been in App Analytics.
Again we are seeing across
the same four measures we saw
in the app list with
App Store views,
app units, sales, and sessions.
Now is a good time to call
out that Trystan mentioned
that we take privacy
very seriously.
All of our usage data is
available only for the customers
who have agreed to opt in
and share their data
with developers.
You'll see this text
that says "opt
in only" below these
usage measures.
If you want to know your opt-in
rate specific to your app,
click on the question mark next
to About App Analytics Data.
We click on that, and we can
see for our app Trip Guider
that in the last 30
days, 23 percent of users
that installed Trip Guider
agreed to share their data.
As you move down the page,
the next thing you'll
see is a breakdown
by day across six measures.
Again, we are looking at
the last 30 days here.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
Again, we are looking at
the last 30 days here.
You can see the value for
the current period as well
as the percentage change
from the previous period,
but now we have a little
bit more granularity.
We can see the graphs by day.
So for example, we knew that our
app Trip Guider 2.0 is killing
it in terms of App Store views.
We are up 82 percent.
But what we didn't know
is that the majority
of this increase came over this
five- or six-day period here.
We can also hover
over any day to find
out the value for that day.
We can see that on May 15
we had our peak in terms
of App Store views, when we had
around 160,000 App Store views.
As you move down the page, the
next thing you'll see is a map.
This is a map of your
app units by storefront.
Here we are looking at which
territories have the most
app units.
For Trip Guider, you can see
that most of the app units came
in the United States,
followed by Thailand.
Again, here we are revealing
missed opportunities.
Maybe we didn't know that our
app was so popular in Thailand,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
Maybe we didn't know that our
app was so popular in Thailand,
and there's an opportunity for
us to go in and localize our app
to Thai to really
optimize the experience
for the customers in Thailand.
You can change which
measure you're looking at,
so this default's app units, but
you can select any other measure
that Trystan mentioned.
Choose sales, and
now we have a map
of our top sales by territory.
As you move down the page,
the next thing you'll see is
on the left-hand side.
You'll see a graph of
your users' retention.
Retention is designed to show
you whether your newly acquired
customers are staying
engaged with your app.
On the right-hand side,
you'll see a breakdown
of your app units by platform.
Again, you can change which
measure you are looking at here
to select from any
of these measures.
You can see that for our app
Trip Guider we have 68 percent
of our app units
coming on the iPhone,
followed by 29 percent
on the iPad.
Now I'm showing you the app
list where you can see your list
of apps as well as
the app overview,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
where you can see a
high-level view of your app.
I'm going to hand it back over
to Trystan, who will walk you
through how we can dive
into some of this data.
[Applause]
>> TRYSTAN KOSMYNKA:
Thank you, David.
Gave a quick overview
of App Analytics.
We had the app list.
This is the best way to see all
of your apps at a single view.
And the app overview, the best
way to look at a single app,
you can see all of the
key measures in one place.
We have the beautiful
map, the beautiful chart,
and you can see it
broken down by platform.
That's just touching
the surface.
As a developer or marketer,
you want to see and explore
and discover new trends.
And you can do this with
a feature called Metrics.
What Metrics does, we
take those key indicators,
all of the measures
we talked about,
and we add in another
layer called dimensions.
These dimensions can
be used multiple ways.
First, as filters.
If you wanted to
answer the question:
How many sessions am I getting
on my iPhones or my iPads?
I can do that by filtering
sessions by platform.
What if you wanted to see
users, are they coming back?
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
What if you wanted to see
users, are they coming back?
You can filter by purchase date.
Is my app version being adopted?
You can do this by filtering
by app version as well.
Dimensions are a great way
to see certain measures
filtered by certain things.
You also can group the data.
If you wanted to see the
composition of your units
by a particular platform, you
can view by that data as well.
We will show these things.
So we have the measures.
We have the dimensions.
You can tell interesting stories
by combining these two things.
We are excited that as of today
there are new measures available
as well.
You can tell all-new stories.
The first one we are
introducing is Crashes.
This is a crash rate.
This will show up just
like sessions and installs.
It's a usage measure.
What it will allow you to do
is you can improve the overall
stability of your app.
If you've really focused on
quality on the next version
of your app, once the
app is in the Store,
you can filter crashes by
2.0 of your app and see
if you made the impact
you intended to make.
Next, we have a new
sales measure.
This is Paying Users.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
This is Paying Users.
If we look at sales,
sales is a number
that hopefully is
increasing over time.
But if you have users that
are purchasing multiple times
within the app, Paying Users
can give you that visibility.
It is the unique count of
how many users have purchased
within your app on a
given day, week, or month.
It's really powerful.
Now, on the topic of power,
we have all of our measures.
We have the two new measures
we are excited about.
There's also power
in comparing things.
I'll give you some practical
examples before David demos
of things you should
be commonly comparing
as you are marketing your app,
as you are making
product decisions as well.
The first is product
page conversion rate.
This is the overall
effectiveness
of your product page.
It is important to focus
on quality screenshots,
using app video previews, making
sure your text sounds great.
You do this by comparing app
units in App Store views.
It will give you an
overall score, so to speak,
of how that page is doing.
And you can compare this
across all of your apps.
Next is average revenue
per paying user.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
Next is average revenue
per paying user.
We take that new measure, Paying
Users, and compare it to sales.
This gives me an
apples-to-apples comparison
across all my apps.
Are my users in app A spending
more than my users in app B?
This is a really important
metric that you can use
to grow your overall sales
and grow the overall
health of your app.
Next, crashes per session.
Again, Crashes is a
measure you want to use
to improve the stability
of your app.
You can use crashes per
session as a comparison
to see how you are doing
against your other apps.
The last example we'll share
is sessions per active device.
This is taking the
Sessions metric
and Active Devices metric,
and by using these
two I get a comparison
across all of my apps.
I will turn it over to
David, and he's going
to demo the power
of the metrics view.
And again, these new measures,
Paying Users and Crashes,
this is data that you do
not integrate to SDK for,
and it's data that
only Apple can provide.
Paying Users comes straight
from the sales data.
It is something that is
unique to App Analytics.
>> DAVID HOPKINS:
Thanks, Trystan.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
>> DAVID HOPKINS:
Thanks, Trystan.
So when we left off, we were
looking at the app overview.
As Trystan mentioned, if you
want to dive into your data,
the best place to do it
is in the metrics view.
The metrics view is a tab
within the context of your app.
We can click that, and it will
take us to the metrics view.
You can click on any of
the data in the overview,
and it will take you to the
metrics view that data selected.
For example, we are looking
at our app units
broken down by platform.
Say we wanted to
take a deeper look
at our iPhone users
and our app units.
We can click on that, and now
we are taken to the metrics view
with our app units
selected, and we're looking
at just our users on the iPhone.
What you'll see in
the metrics view is
in the top left you'll see the
value for the period as well
as a chart where we break
down your app units by day.
If you hover over a day, you'll
see the value for that day.
If you scroll below the
chart, you'll see a table
that contains the
value for each day.
And if you scroll to the top,
the next thing we will talk
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
And if you scroll to the top,
the next thing we will talk
about are your measures.
On the left-hand
side you can see
that we have app units selected.
Your measures are broken
down by sales and usage.
If you have any questions
about any of these measures
at any time, you can
hover over the measure
and you'll see a question mark.
If you click on that
question mark,
you'll get the definition
of the measure.
So we talked about this
new measure, Paying Users.
You can see it says the number
of unique users that paid
for the app or an
in-app purchase.
Great. I talked about
how we are looking
at just our app units
on the iPhone.
To do this, we have
a concept of filters.
To add a filter, click
on the top right,
and you can add up
to two filters.
Right now you can see
that we are looking
at just our users on the iPhone.
We can click to add
a second filter.
So let's look at our users on
the iPhone in the United States.
And now we are looking
at just those users
on the iPhone in
the United States.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
on the iPhone in
the United States.
You can also change which
version you're looking at.
If you want to switch
between platforms,
select from iPhone
and switch to iPad.
Great. Now I've shown you how
to look at a specific value
within any given dimension.
Oftentimes, you want to
instead compare the values
within a given dimension.
To do this, we have a
feature called View By.
If we remove our filters, and
let's say we want to look at all
of our top territories
for our app units and see
which ones rank the highest.
We can select View By,
and select territory,
and now you'll see all of
your top territories plotted
over time for app units.
Again, if you hover over
the graph on any given day,
you can see the value for
each of your top territories.
If you scroll down below the
chart, you'll see a table
that contains the total
value for each territory.
And you can see here that we
have five territories selected
by default.
These are the top territories.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
But you can also
deselect any of these.
We can deselect Brazil
and then plot Japan.
Now the graph will update.
The default view here
is as a line chart,
but you can also change this to
an area chart or a bar chart.
We can switch to area.
That gives us a nice view
of the overall curve as well
as a breakdown for
each dimension.
Okay. So now I've shown you
how to compare the values
within a given dimension.
But sometimes you're
more interested
in the relationship
between two measures.
One example of this that Trystan
talked about is the relationship
between app units
and App Store views.
You'll call this your
product page conversion rate.
What percentage of users are
landing on your App Store page
and going on to purchase
your app?
To do this, we have a
feature called Compare To.
From here, we're
looking at app units,
and we can select
another measure.
So we'll select App Store views.
Now you'll see both
your app units
and App Store views
plotted on the same chart.
The default view here, we're
looking at a dual-axes graph.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
The default view here, we're
looking at a dual-axes graph.
You can hover over any
day and see the value
for each of those measures.
In the top-left corner you
can see the total value
for each of those measures.
This is a really nice way
to view the relationship
between these measures, see
how they trend over time.
We are excited to announce
some new enhancements
that we're bringing to
Compare To as of today.
Instead of just being able to
view this on a dual-axes chart,
we added the ability to
view this rate over time
with the new feature
we call Ratios.
So from here, you can
select new chart types,
from Single Access and Ratio.
We'll select Ratio.
And now you can see the
ratio between your app units
and App Store views
plotted over time.
That gives you a nice picture
of that product page conversion
rate that Trystan talked about.
You can see here that
we're at around 18 percent
for the last 30 days,
but a few weeks ago
for Trip Guider 2.0 we made
some updates to our description
of the app on the
App Store page.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
of the app on the
App Store page.
And thanks to App Analytics,
we are able to see the product
page conversion rate drop off
right here.
We got down to about 6 percent.
So what we did is we
went and updated the copy
to the description of
our App Store page,
and we saw that product page
conversion rate rise over time.
As you go on you can see
that now we are back up to
around 30 percent
conversion rate.
Great. Now I've shown you how to
use some of the power features
in the metrics view, and I'll
hand it back over to Trystan.
>> TRYSTAN KOSMYNKA: So
David went through metrics.
This is the best way to
explore and discover things
and reveal those
missed opportunities
within App Analytics.
If you want to dive deeper
than the overview,
this is that feature.
One question we have
not answered yet is,
how do people find your app?
This is a question that is
hard without App Analytics.
To do this, we have a
feature called Sources.
App Analytics has
two types of sources.
The first is completely
organic and free.
You do not need to do any work
to take advantage of this.
This is websites.
If you are actively marketing
your application on websites,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
If you are actively marketing
your application on websites,
forums, maybe it's getting
picked up by the press as well.
Websites, any time a link to
your App Store page is present,
we capture the web-referring
URL and present
that to you in App Analytics.
It is important that these,
both websites and campaigns,
they slice across
your entire data set.
You are able to filter App
Store views, in-app purchases,
the new measures
such as Paying Users.
You can filter these
by the website
and see how effective those apps
are at increasing these metrics.
Next we have campaigns.
So websites is something
that you don't need
to do anything to
take advantage of.
If you log in today, you
should be pleasantly surprised
that there are websites
available for your apps.
Campaigns: you do
need to create them,
and we will talk about this.
The easiest way,
the vanilla way,
to create them is your App Store
URL, you add a provider token.
In our example, we have
one, two, three, four.
That is not your provider token.
If you use 1234, you won't get
campaigns in your dashboards.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
You can then add
a campaign token.
My recommendation to you
is that we take an event
and placement approach
with this.
Think of WWDC as the event
and placement as newsletter
or forums or home page.
You combine these in a
single string, and that way
when you see it in App
Analytics, you know what it was
for and where it was running.
There are all kinds of ways you
can do this, and we want you
to experiment with it.
We think it's really important.
The next way to create
a campaign is StoreKit.
If you're using StoreKit
today, all you need
to do is add two parameters.
Add the Provider Token
parameter, it's a constant
in this dictionary, and you add
your Campaign Token parameter.
If you are not using
StoreKit, it's an easy way
to create an in-app App
Store-type experience.
You can cross-promote your
own titles within StoreKit,
and it's the best way to
get a nice conversion rate
for installs within
your own app.
It is designed by Apple.
You don't need to create
your own mimicked App Store
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
You don't need to create
your own mimicked App Store
within an app.
The last way to create
campaigns is smart app banners.
If you have a home page on which
you're advertising your apps
today, you've probably
seen this already.
If you're creating an app
and this is the first
time you're here,
you can add a meta tag to your
home page and that meta tag,
you can add a provider
token and campaign token
to your affiliate data, and
that campaign will be attributed
to all of the traffic
coming through.
To demo this, David
will come up again.
He's going to walk through
sources, how you see it.
He's also going to talk about
how you create a campaign
and how you can get that
provider token from the UI.
>> DAVID HOPKINS: Great.
So as Trystan mentioned,
there's two types
of sources we track
in App Analytics.
You have your top websites
and your top campaigns.
Websites are tracked any
time a user follows a link
in mobile Safari to
your App Store page.
For example, we have our
website for Trip Guider,
which is Trip Guider.apple.com,
and we've implemented a
smart banner that links
to our App Store page.
Any time a user follows
that link,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
Any time a user follows
that link,
it's going to generate
a App Store view.
Any time the user goes
on to purchase the app,
it will generate an app unit.
And so on for sales,
and sessions,
and all of your measures.
This view should be familiar.
It's the same view we
saw for our app list.
Now we are looking
at our top websites.
You can click on any of the
websites, and you'll be taken
to the source overview
for that website.
Again, this view is going
to be familiar to you.
It is the same view you
saw for your app overview.
Now you are looking at
just your users who came
from Trip Guider.apple.com.
So the great thing
about websites is
that you get them for free.
You don't have to do anything
to implement your websites.
They just show up automatically.
Any link that goes to the App
Store page will just show up.
The other type of source
we have is campaigns.
Campaigns are great because
you can use them anywhere.
They are not just
working on mobile Safari.
Anywhere on iOS.
We can send out an email,
put them on social media,
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
We can send out an email,
put them on social media,
but they require setup.
So for most of you, when you
log in you'll see something
like this like we have
for Trip Guider.com,
which is an empty
list of our campaigns.
I'll walk you through
how to set up a campaign.
For Trip Guider, we want to
promote the use of Trip Guider
in order to find
things to do for WWDC.
I'll show you how to set
up that WWDC campaign.
As Trystan mentioned,
all you have
to do is add a provider token
and campaign token to your URL.
But to make this even
easier, we have a tool
that generates the
campaign link.
When you follow this
link, you'll see here
that you have a tool
and all you have,
is you have your app selector,
so you can switch between apps,
we have Trip Guider selected.
And that will populate
your app's app ID as well
as that provider ID that
Trystan talked about.
And here we can enter
the campaign name,
so we'll say WWDC --
whoops, there we go.
2015. It generates
a campaign link.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
2015. It generates
a campaign link.
Now you can copy
this campaign link
and you can put it anywhere, in
an email, on your social media.
You'll start to see
your campaigns show up.
It is also important to note
that generating the
campaign link does not mean
that you'll see the campaigns
in your top campaigns list.
Campaigns will only
start to show
up once you have five users.
Follow that link and then
go and purchase your app.
So now that I've shown you
how to generate a campaign,
let's take a look at an app
where we've already
started use campaigns.
The easiest way to
switch between apps is
to use the app switcher
in the top left.
So we switch from Trip Guider
to our other app, Tide Minder.
And now we can see
the list of campaigns
that we set up for Tide Minder.
So, for example, we
set up a campaign
to track our home page here.
So we have our display
home page campaign
that is showing up in the list.
Similar to what you
saw in the app overview
and the top websites
overview, you can click on this
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
and the top websites
overview, you can click on this
and you will be taken to
your campaign overview
where you can see a high level
of how this campaign is doing.
So now I've shown you how
to track these two
different types of sources.
We hope that you go
out and start setting
up campaigns right away.
I'll hand it back
over to Trystan.
[Applause]
>> TRYSTAN KOSMYNKA: All right.
Thank you, David.
David covered the campaign
list, the website list.
Very familiar to the
app list as well.
It filters all of your data
by a particular website.
We show the top campaigns
and the top websites here,
and then we have the
overview so you can see all
of it in a single shot.
If you're running a campaign,
perhaps you're paying
for the campaign as well.
This is a great view to
just look at and see,
how is it actually doing?
Is it money well spent?
Last question we'd
like to answer is,
are customers returning
to the app?
David mentioned there's
a retention widget inside
the overview.
We will go into this
in some more detail,
go over the retention feature.
Before we do that,
before we show the UI,
I want to give a lesson on
how this actually works.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
I want to give a lesson on
how this actually works.
We have a purchase date.
And what makes sense is that
100 percent of the customers
who purchase the
app are customers
on their purchase data.
They've remained customers.
They are retained.
What isn't as obvious is
that a couple days later
and all the way up
to 30 days later,
less of them are
actually active customers.
So here we see day two, 45
percent are remaining customers,
and then on day 30,
we see 10 percent.
So what we're looking at here
is a single purchase date,
and then seven days later,
20 percent of the
customers are retained.
We're going to bring this down,
and we see that June 1
is our real example here.
So June 1, two days later,
45 percent of the
customers were retained.
One purchase date is great
to look at, but you want
to compare this because
you are making many changes
over the course of
time and you want
to compare multiple purchase
dates in a single shot.
We added June 2.
We have June 1 and 2 now.
Let's expand that even more.
That is not enough.
Now I see June 1 to June 9.
I'm looking at nine
purchase days.
We've changed the
numbers now as well.
At the top I have 45 percent
for two days after June 1.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
At the top I have 45 percent
for two days after June 1.
At the bottom, June
9, I have 60 percent.
So it's a little hard to read
because it's a single color.
So what we've also done to
this heat map is we've added
colorization, so I can
see that 10 percent
of my customers 30 days
later have returned
on the June 1 purchase date.
That's a lighter
color indicating
that the retention is not
so good at that point.
Go down to the bottom right,
and I see a darker color
indicating there's a hotspot
there, those improvements
we made.
Maybe it was my App
Store presence,
but likely it was the
new version of the app
or new content to the app to get
people to come back and use it,
and I have a 60 percent
retention rate two days
after June 9.
That's how retention works.
The easiest way to
understand it is to do a demo.
So we are going to
do another demo here.
David will walk through
retention.
>> DAVID HOPKINS: Great.
So retention is available
as another tab
within the context of your app.
We are looking at the retention
for our app Trip Guider.
The first thing I'll talk
about is the retention grid,
which Trystan introduced.
On the left-hand side, you'll
see the purchase date for each
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
On the left-hand side, you'll
see the purchase date for each
of the last 30 purchase days.
Next to that, you'll
see a number of devices.
That's the number of
devices that installed
on that purchase date.
On the top of the grid
you'll see a number.
That is the number of days
since the purchase date.
So when we look at this number
one, it means we are looking
at one day after
the purchase date.
The easiest way to
understand the retention grid
if you are getting lost
is to hover over any cell,
and it gives you a
nice explanation.
So we'll hover over this
32 percent, and it says,
"32 percent of the devices
that installed the app
on May 13 were active
two days later."
So oftentimes you'll
hear people talk
about their day seven retention.
And as an app developer,
you know that people will
purchase your app and a lot
of times they come back
on day one or day two,
but then the usage
starts to fall off.
We like to measure a day
that is a little further out.
That's why day seven
retention is important.
We make this easy to see
within App Analytics.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
We make this easy to see
within App Analytics.
All you need to do
is click on the seven
and you can see your day
seven retention over time.
As the app developer,
your goal should be
to keep the users
engaged and try to drive
up the day seven
retention over time.
You can see that we start
out around 18 percent.
As time goes on, we are
increasing a little bit,
but our goal should be really
to keep our users engaged
for day seven and beyond.
You can also look at any
single purchase date and look
at the retention for that date.
So, for example, if we want
to look at our retention
on May 12, we can click on that.
You can see in the top left, now
we have a graph of our retention
for our users who
purchased on May 12.
It starts at around 48
percent on day one and goes
on as time goes on and
goes to around 8 percent.
Again, we want to focus here
on driving up our retention.
The last thing I want to point
out on the retention view is
the ability to add a filter.
Just like you saw on the
metrics view, we can also filter
by any dimension and
look at the retention
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
by any dimension and
look at the retention
for just that dimension.
We can add a filter and
look at just our retention
for our users on the iPad.
Now the retention updates.
Great, that's the
end of my demo.
I hand it back to
Trystan who will wrap
up our session on App Analytics.
[Applause]
>> TRYSTAN KOSMYNKA:
Great, thank you, David.
That was retention.
This is answering the question,
are users returning to the app?
We covered a lot today,
first thing was the app list.
This is bird's eye view
of all of your iOS apps.
We also support bundles.
So we have page views, units,
and sales for app bundles.
These are storefront
and sales measures.
Next, he covered the overview.
This is your app at a glance.
All of the measures
are available.
There's the nice
retention widget.
You have the map as
well as the pie chart
for the platforms as well.
Metrics, you use this to explore
and discover new
trends within your app.
You can see the dimensions
are available.
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
We have the new metrics as well.
We have Paying Users
and Crashes,
those are new as of today.
Sources answers that
question of,
how are customers
finding my app?
Are they actually turning
into real customers?
We have the page views, which
are potential customers.
Are they turning into engaged
and purchasing customers?
Finally, are customers
returning to your app?
We answer this question
with retention.
We are excited about
App Analytics.
We hope you use it
to build better apps.
We hope this conference has
been a great week for you.
For more information, look
at the App Analytics
Developer Guide and look
at the developer forums.
We encourage you to talk to the
Evangelist as well, Mark Malone.
We will be downstairs
after this in a lab
for a short amount of time.
We also will be available
for Thursday's lab as well.
There are a couple
related sessions,
Promoting Your App with an IAd.
This already aired, so
we encourage you to look
on the video afterwards, and
the What's New in iTunes Connect
that occurred this morning.
This is a great session.
We walked through new
metrics that will be available
X-TIMESTAMP-MAP=MPEGTS:181083,LOCAL:00:00:00.000
We walked through new
metrics that will be available
in the fall for TestFlight,
and there's How
to Manage Your Store
Presence as well.
The labs again this
afternoon and tomorrow,
we will be available, the
analytics team will be there
to answer your questions.
Have a great week.
Thank you very much.
[Applause]