Transcript
[ Music ]
>> Good morning.
Welcome to our --
[ Audience Applause ]
Thank you.
It's great to be here.
Thanks everybody for coming.
Welcome to our session on
Introducing Podcast Analytics.
I'm James Boggs.
I manage the Business Team for
Apple Podcasts and Siri Audio
Briefs, and we've got a great
session planned for you today.
First, I'm going to share some
details of our podcast business
over the last year, including
some very interesting and
exciting milestones.
Then I'm going to welcome Anne
Wootton from our Engineering
Team to the stage, and she's
going to take us through some
updates to Podcast app for iOS
12.
After Anne, Alec Reitter will
take us through a demo of
Podcast Analytics, and then we
will finish up with Anne sharing
some requirements and taking a
look at resources available to
podcast makers.
First off, an update on our
podcast business.
Well, we're just shy of the 13th
Anniversary of the launch of
iTunes 4.9.
That's the moment when we took
podcasts mainstream for the
first time, and podcasts have
come a long way since then.
Our content catalogue has never
been more entertaining,
inspiring, and educational, and
we've never reached more
listeners around the world than
we do right now.
We're reaching them with some
great programming like Sandra
from Gimlet.
This is a brand new program from
one of the leading independent
podcast networks out of New
York, and it stars some really
big talent, Kristen Wiig, Alia
Shawkat, and Ethan Hawk.
It's a really fun scripted
comedy drama about the inner
workings and quirks of a virtual
assistant.
Los Angeles-based Wondery is one
of adventure capital backed
network startups working in
podcast today, and last year,
Wondery collaborated with the LA
Times to produce the thrilling
true crime story of a charming
but sociopathic con man, Dirty
John.
The Hollywood-style production
value is powerful investigative
journalism and remarkable
storytelling, which is just
riveting.
It drove this show to the top of
the charts in the US several of
the markets around the world.
If you haven't had a chance to
check it out, I recommend it.
It's a great binge.
And of course, podcast fans
everywhere are thrilled and
enjoying Oprah's Super Soul
Conversations.
This is the audio instalment of
the Queen of Talks iconic
interview franchise, and we were
overjoyed to have Oprah on Apple
Podcast.
These examples are just three of
the more than 550,000 shows
available today on Apple
Podcast.
That's quite a bit more content
to listen to than the original
3,000 programs we launched with
back in 2005.
We're continually refreshing and
managing our directory,
automatically retiring shows
which become technically
unavailable or those that run
afoul of our directory content
guidelines, such as those with
spammy content or shows seeking
to manipulate our top charts.
Don't do that.
Just please don't do that.
At the same time as we're
managing the directory, we're
accepting thousands of new
submissions per week.
That way, our catalogue is
continually refreshed, and our
users around the world can
always find something great to
listen to.
Amongst those more than 550,000
active shows, we have over 18.5
million episodes of content.
That's more than 8 million hours
of programming.
So, you'd better, better start
listening, better start
bingeing.
As a reminder, Apple Podcast is
available in 155 countries.
That's not only for the
distribution of existing shows
but for the submission and
acceptance of new programming
from each one of those markets.
The Podcast Directory serves an
audience in all 155 of those
countries with content in more
than 100 languages.
So, that means if you're a
Korean native speaker, ex-pat,
living in Toronto, you can find
some great audio programming
from home right in Apple
Podcast, and we're particularly
proud of this, particularly
proud to have such a diverse
catalogue and be one of the
easiest ways to reach a global
audience for audio programming
that there is.
The data from our beta version
of Podcasts Analytics shows some
remarkable listener affinity and
engagement.
We looked at several hundred of
our recent top episodes, those
with more than 7 days of
playback data, and we found some
really notable average
consumption.
Of course, this data is
nonamized [phonetic] and
aggregated and a sample, so no
personally identifiable
information is included, and
it's not a comprehensive
statistic but interesting
nonetheless.
It shows that once an episode is
started, on average, about 80%
of that episode content was
consumed.
This is great stuff and a real
testament to the affinity and
engagement the podcast audience
has with this high quality
content.
Last fall, we successfully
integrated more than 50,000
iTunes U collections into Apple
Podcasts.
This is content from some of the
world's best educational
institutions, including
developing iOS 11 apps with SWF
from Stanford's Paul Haggerty.
Naturally, we're eager for the
new instalment of this class
series later this year.
As part of Apple Podcast,
teachers can now leverage the
episode serial ordering support
in Podcast app.
They can reach a new broader
audience of lifelong learners
around the world, and it can
distribute the content on tvOS
for the first time.
We're seeing some great use of
modern podcast tags, which we
introduced last year right here
on stage at WWDC.
Many podcasts include trailers
to thematically introduce the
feel and sound of their show.
This is the best way really to
launch a new podcast, and ESPN,
for example, makes great use of
trailers and, in addition, is
amongst the chart shopping shows
that uses seasons and bonus
content.
This allows Jody Avragon
[phonetic] and the team to
easily present the thematically
different collections of
episodes in each season.
In addition to seasons and bonus
content, 30 for 30 feeds their
super fans with bonus episodes.
These complement the full season
episodes.
A best practice regarding bonus
content is to make sure that
your bonus episodes are tagged
appropriately to the full season
episodes.
Just want to ensure proper
ordering of the content in Apple
Podcasts.
And of course, episode number,
we're using this handy reference
for your audience inside Podcast
app.
We're also making use of episode
numbers that you provide in your
RSS feed in Podcast Analytics.
An innovative crop of
venture-backed hosting startups
has started to grow.
Organizations such as Anchor,
with some great new iOS-based
creation tools, ART19, which is
an easy to use solution for your
enterprise podcast needs, and
Simplecast, with brand new
migration tools and a 2.0
version they just launched.
These next generation mobile
optimized podcast creation tools
can help make your programming
available to the world, and
these folks join our lineup of
long-standing hosting partners
from around the world, including
Libson, Megaphone, Omni Studios,
and Wushka from Australia,
Spreaker, and more.
Check out our resources now for
our full list of Apple podcast
hosting partners.
HomePod, which arrives soon in
Canada, France, and Germany is
an awesome podcast machine.
Siri has been optimized for
podcast playback, so you can
just ask Siri to subscribe to or
play your favorite show.
We also offer audio news briefs.
Just ask Siri to play me the
news.
We're proud to be distributing
headline news bulletins from
some of the best news
organizations in the world,
including NPR, the BBC, ABC
Australia, ESPN, Bloomberg, and
more.
You can ask Siri for audio
briefs and general news, sports,
business, and even get an update
on the goings on in music from
our own Apple Music Team.
We're thrilled to celebrate the
incredible milestone of 50
billion episode downloads and
streams on Apple Podcasts since
the launch of the service in
2005.
With over 10.5 billion episode
downloads and streaming plays in
2016 and more than 13.7 billion
last year, the growth trajectory
of the platform is amazing.
We're also thrilled that the
shows that are driving all this
consumption are of such a high
quality, and it's a pleasure for
us to be able to champion this
content on top of our platform.
We celebrate the podcasts that
contribute to this 50 billion
milestone, and one show in
particular, "Stuff You Should
Know" from Atlantis How Stuff
Works Network, the first
podcast, an Apple podcast to
exceed 500 million downloads and
streams.
Wow! Congratulations to Chuck,
Josh, Jerry, and the rest of the
team.
Well done!
We're looking forward to the
next half billion downloads and
streams.
So, with that business update,
let me welcome Anne to the stage
to take us through our app.
[ Audience Applause ]
>> Thanks, James.
Thanks James, and hi.
I'm Anne, and I run Podcasts
Engineering at Apple.
Before we dive much deeper into
analytics, I want to highlight a
few of the latest updates to the
podcast app.
As you probably heard in
yesterday's keynote, we are
thrilled to be launching
podcasts for Apple Watch this
fall.
It was one of the most requested
features for Apple Watch and for
good reason.
Between HomePod and Apple Watch,
listening to podcasts has never
been easier or more hands free.
We're also really excited to
share that we've enhanced our
support for chapters in the
podcast app.
In the upcoming IRS release,
we're extending full chapter
support to MP3s.
In addition to enabling
podcasters to segment their
episodes into shorter discrete
chapters, this functionality
also includes chapter art that
displays as the play head
advances across chapters.
So, podcasters can supplement
their audio with visual content.
In the upcoming iOS release,
we're also offering an improved
experience for podcast
notifications.
The new podcast notifications
are grouped across shows and
episodes into a single
expandable stack that reduces
clutter on screen, and we also
provide one-tap access to
notification settings for all of
your shows.
That means if you only want to
get notifications for your
favorite show, 99% invisible,
it's as simple as toggling the
switches on this single settings
page.
So, last year, at WWDC, we
announced that Apple would be
launching Podcast Analytics for
the first time.
We were super excited to launch
at the end of 2017, and Podcast
Analytics is available in beta
right now today to anyone with a
podcast in our catalogue.
We know that this is a service
that was long desired and speaks
to a critical need in the
industry as evidence by work
like the remote audio data spec
being spearheaded by NPR.
So, it's important to us that we
provide this anonymous data back
to publishers to help them
better understand how their
audiences are consuming their
audio.
And all of the data and
analytics is based on the use of
the update Apple podcasts app
from iOS 11 and iTunes 12.7.
We're collecting data only from
those operating systems, Forward
and the Analytic Dashboard
includes listening data from
HomePod, AppleTV, and will
include Apple Watch listening
data starting in watchOS 5.
As you know, privacy is really
important to Apple, and we are
only showing aggregate data of
those users who have agreed to
use the podcast app.
It is anonymous, not personally
identifiable, not tied to your
Apple ID, and we only reflect
the data points in aggregate
that we've collected from at
least five unique devices.
And it couldn't be easier to get
started.
You don't need to write any
code, you don't need to use an
SDK, and it's totally free.
If you have a podcast, all you
have to do is submit your feed
to Podcast Connect after you
signed in, get it approved, get
some listeners, then check your
performance through the Podcast
Analytics dashboard.
Analytics reports the data from
unique devices that engage with
the Podcasts app.
As my colleague, Alec, is about
to show us from the Analytics
Dashboard, publishers can access
an overview that includes tables
and charts with data-like unique
device counts for a selected
period of time, total time
listened, time per device,
average consumption, top
countries, and more.
So, some of you may be familiar
with our podcast, The Very
Hungry Tourists.
They're back for season 2, and
they even have some new cover
art.
Alec is going to take you on a
deeper dive of how to get the
most out of analytics.
We're excited to continue
providing key listening metrics
for this industry, and we are
listening to your feedback and
continuing to improve the
service.
So, thank you.
And with that, I'm going to hand
it over to Alec to take us on a
tour of Podcast Analytics.
[ Audience Applause ]
>> Thank you, Anne.
So, podcasters, every day you're
hard at work making great shows
and episodes for people to enjoy
all over the world.
But how much do you know about
your audience?
Do you know how many listeners
you have or how much they
listen?
Do you know where they are
around the world?
What are their favorite
episodes?
With Podcast Analytics, Apple
offers you transparency into how
your show is heard, so you can
better understand your listeners
and create the best show for
your audience.
My name is Alec Reitter, and I'm
going to do a walkthrough and
demo of Podcast Analytics.
So, let's get started by heading
to the login page.
When you see this bright purple
page, you know you've made it to
the right place.
Here, we're going to login with
the Apple ID that we used to
originally submit the podcast
feed to the Apple Podcast
catalogue.
Now, if you have any trouble,
such as you forget your password
for your Apple ID, or you don't
know which Apple ID to use, or
if after you've logged in you
don't see all the shows you
expect to see, go to the top
right corner of this page and
click on that question mark
icon.
That will take you to Resources
and Help, and there you can find
Contact this Link, where people
can give you the assistance you
need.
In this example, we're going to
use John Appleseed's login
credentials.
So, let's login now and see what
he sees when he enters Podcast
Analytics.
So, this is the show's page.
Here, we see five shows for
John's account.
In this table, there are three
metrics, and we're going to take
a look at those in a minute, and
all of the data in this table is
driven by the date picker, which
is in the top right corner.
Right now, it's set to the last
60 days, and that is the
default.
So, let's take a look at the top
row for The Very Hungry
Tourists.
Here, we can see the number of
unique devices for the last 60
days.
This answers the question, how
many listeners do I have?
Here we see 256,000 unique
devices.
Next to that, we can answer the
question, how much do they
listen?
In this case, the total time
listened is 327,000 total hours.
Now, if we do a little math, we
can find the average, which is
the time per device of 1 hour
and 17 minutes.
So, this is a very popular show.
It's doing really well.
So, next, we're going to switch
tabs from the show's tab to the
episodes tab.
Here, we can see a cross-section
of episodes across all of the
shows in John's account.
Here, we see a few of the most
recently released episodes.
Now, this data table has a lot
more information in it, and
let's take a look at the top
row.
This is the most recent episode
from The Very Hungry Tourists,
starting on the right-hand side,
we can see the name of the
episode, then the name of the
show, and now, the third column
is the episode column.
If you're making use of the new
tags, like season, episode, and
episode type, they'll show up
here.
Following that, we have the
release date and then the
duration.
So, in this case, it's a
35-minute long episode.
If, when you login, you don't
see the right number in the
duration column, it's because of
what you have in your feed.
We pulled this number straight
from your feed.
So, make sure it's accurate.
Following the duration column,
we see three columns that we've
seen before, except this time,
for this episode specifically.
So, we have devices, total time
listened, and time per device.
And for the first time, we see
average consumption.
This is the last metric in the
far right column.
Average consumption is really
great because it's really just
another way of looking at time
per device.
So, in this case, it's 28 minute
for this 35-minute long episode.
That is 81%.
What's nice about average
consumption is it allows us to
compare episodes of differing
durations against each other to
really get a sense of how
healthy each episode is doing.
And from here, let's click
through on the Show Title and
look at the show overview for
The Very Hungry Tourists.
So, here's the show overview.
There's a lot going on here.
Let's start at the top.
Across the top, we have three
headline numbers.
These are the same numbers that
we saw before in the shows
table, but now we have a mini
time series underneath these
numbers to give you a sense of
how things are doing.
Maybe you can see, just looking
at that little line, if your
show is getting more or fewer
listeners over time.
Now, let's look beneath this row
and take a look at the chart on
the left here.
This is the donut chart of total
time listened.
So, for all of the time spent
listening to your show, were
those devices subscribed or not
subscribed?
Here, we can see that 83% of
that 328,000 total hours of
listening was done by a
subscribed device versus 17% not
subscribed.
So, these are pretty good
numbers.
But why might these numbers be
different when you take a look
at your data?
Well, if you have a very small
podcast, one with a very
dedicated following, you might
not have too many new listeners,
but your subscribed count might
be above 90, even 95%.
Conversely, if you're doing a
lot of marketing or promotion,
or your show is growing rapidly,
you're not subscribed count
might be fairly high.
Now, when somebody first starts
to listen to your show,
obviously, they're not going to
be subscribed most of the time.
But once they hear what you have
and they hit that "Subscribe"
button, that's when they end up
in the other bucket.
Now, let's take a look at the
chart to the right.
This is the Top Countries by
Device.
Here, we can see the top five
countries for the show.
As we can see, the United States
comprises the majority of the
listening at 78%.
If we look at the fifth spot, we
see Brazil.
Now, what's neat about this is
that Brazil wasn't on this top
five before, but as Anne said
earlier, this is season 2 of The
Very Hungry Tourists, and on the
first episode of this season,
they actually went down to
Brazil.
They did some marketing and
promotion, and it's starting to
pay off.
So, next, let's scroll down to
the bottom of the page, and here
we can see the ten most recently
released episodes for the show.
This is the same sort of data
table that we saw before for
episodes, but this time it's for
a single show.
What's nice about this is we can
compare apples to apples and
really see on an episode by
episode basis how things are
turning out.
Let's go back up to the top, and
we're going to switch from the
Show Overview to add to the
Trends tab.
So, here we are in the Trends
tab.
This is a Time Series Chart
where we can see day by day how
many people are tuning into this
show.
And again, by default, we're
looking at the last 60 days of
data.
In the top left corner, we see
that headline number again of
256,000 unique devices, and down
below the Time Series Chart we
see a table with episode titles.
We can choose up to five
separate episodes to compare
against one another in the chart
above.
So, let's first now talk about
all of the different drop downs
that control the chart and what
we see in it.
First up, we have a drop-down.
This is the metric that we're
going to chart.
By default, we start off with
Unique Devices, or you can also
choose to have it render Total
Time Listened.
To the right of that, we have
another drop-down.
This is what I call the primary
filter.
By default here, we're looking
at Episode as the primary
filter, but you might want to
choose Country or Subscribed
Device.
Now, whatever you choose for
your primary filter, that's
going to dictate the table
beneath the chart.
So, whatever is selected allows
you to choose multiple entries
to compare against one another.
For example, we could choose
Country, and we'll have
countries in that table below.
Whatever filters are not
selected from this menu, they
move over to the right-hand side
of the page in their own
drop-downs.
For example, countries or
subscription state [phonetic].
For each of these, you can
choose up to one filter.
Now, with that, that's enough
talking through slides.
Let's look at a demo, and I will
take you back to that page but
in a hands-on fashion.
So, here we are.
We're in that same view that we
saw before of the Trends page.
But now, we can see what happens
when I interact with it.
When I move my mouse cursor over
the chart, we have this nice
data legend with details about
the specific point that I'm
hovering over.
In this case, I'm looking at May
18th, where 34,000 unique
devices tuned in.
And I can go down the page, and
I'm going to select a few
episodes to chart against each
other.
We get this nice colorful view,
and as I mouse over, I can also
again see the details of what
I'm hovering over.
In this case, I can see episode
titles and their corresponding
counts for those days.
Now, I'm going to go change my
primary filter, and I'm going
switch it from Episode to
Country.
So, this updates the chart once
again.
This is the same actual graph
that we saw the first time
because now, again, we're
looking at all countries.
But I'm going to choose the top
five countries and compare those
against each other.
Once again, we get this colorful
chart.
But in this particular case,
because the United States
represents so much of the
listening for this show, it
makes it difficult to see the
other countries and how they
stack up against each other.
So, I'm actually going to go and
deselect the United States.
The chart adjusts, and we can
really see with more detail how
each country is doing.
And now that I can, I'm going to
show you the date picker.
It's up here on the top right
corner.
By default, it sets at the last
60 days.
You can switch that to weeks or
a specific month, but we'll
leave it at that for now.
Now, there are lots of
combinations that you can make
here, but I'm going to move on
after this, but I really
encourage you to take a look at
your shows and see how they're
doing and really play around
with this tool.
I think it's a lot of fun to
use.
So, let's switch over to the
Show Episodes tab, and in this
view, we can see the top
episodes or all episodes for the
show, not just the top episodes.
In this case, the show has 11,
11 episodes total.
And now, because all episodes
are together in this view, we
have the opportunity now to
answer some other questions
like, what are the most popular
episodes?
One way to answer that is by
sorting by the Unique Devices
column.
When I do that, I can see that,
essentially, season 1, episode
4, Barcelona, Barbacoa, that is
the most popular by devices, by
unique devices.
But maybe you think that it's
actually total time listened
that should be the determinant
of popularity, so I'm going to
sort by that instead, and we see
that it's actually season 2,
episode 1 that takes the top
spot.
Or for myself, I personally
really like average consumption,
so I'm going to sort by that,
and I can see that at 89%, the
trailer is actually the most
popular episode, which makes
sense because it's a short
episode.
So, now, we've actually seen all
of the show level views.
Now, I'd like to take you into
an episode level view.
So, I'm going to select, I'm
going to click through on the
title for the most recently
released episode.
Here's Broken Heirloom.
So, it was just released on June
1st.
Front and center, we see a
timeline chart.
This shows you the audience size
over the duration of the
episode; actually, at every five
seconds, and when I hover over
the chart, I can see a data
legend again with details about
the specific time in the
episode, its total count, and
its percentage of the max.
And in the top left corner, we
see this headline number for the
total number of unique devices.
Something to note about this
chart, unlike all the other
views on this, in Podcast
Analytics, this chart is
actually for all time, not just
a specified date range.
Now, what's really nice to see
here is that, in this case, most
of the people who started this
episode actually hung out for
the entire duration.
This is a really healthy-looking
episode.
And beneath the chart, if I
scroll down a little bit, we can
see for the last 60 days the
metrics for this show, for this
episode, I should say.
This gives you a more recent
sense of how this episode is
doing, not just all time.
And at the very bottom, we see
the top ten countries for this
episode.
But I'm going to scroll back up
and cover two things that I
breezed past before.
There is a drop-down menu above
that all-time devices number,
and when I click on this, I can
see that I can also choose
Devices or Total Time Listened.
Now, by choosing either of
those, it will actually take us
into a view which is a time
series view, just like the Show
Trends page except this time
it's tied specifically to this
episode.
And now, the part of this graph
that I like the best is actually
the ability to play back the
audio while seeing how the
audience is reacting over the
duration of the episode.
And I can do this by going down
to the bottom left corner of the
graph and hitting this Play
button.
>> Hey there, Hunger Heads, and
welcome back to another episode
of The Very Hungry Tourists.
I'm Maria Sanchez, Ph.D.
>> And I'm Johnny Appleseed.
>> In today's episode, we're
headed to Brooklyn based Viny's
Pizzeria.
As we're here, even Johnny's
massive wonder couldn't get in
the way of our culinary
adventure.
>> I'm still blushing from that
one.
>> So, this is actually a very
entertaining episode, so if you
get a chance, you should
definitely check it out.
But like any basic player, you
don't just have to start at the
beginning.
You can also click on the
timeline to jump ahead to any
specific spot in time, and if
you want to fine tune, you can
grab the play head and scrub
around.
So, I'm actually going to scrub
to the beginning of what looks
like an ad.
And so, just to validate my
assumptions, I'm going to move
to just before and take a
listen.
>> Five years, you wouldn't
believe it, but in the back
there is a big window--
>> As the pizza goes in the
oven, let's take a moment to
hear about out great sponsors.
>> Just say your hands and face
are smothered with Tahini sauce.
So, what do you do?
>> Get a paper towel.
>> Ah, too rough.
>> Tissue?
>> Too soft.
>> What about napkins?
>> That's a great idea.
>> Napkins to you is a new drone
delivery service bringing you
napkins where and whenever you
need them.
>> Simply open the app, request
a drop off, and the closest
drone is deployed to your
location.
By typing in promo code
"Hungry."
You can have your very own
napkins delivered for just $4.99
per month.
>> Again, that's Napkins to You.
>> So, that's sounds like a
really entertaining ad, and it's
too bad that some people decided
to skip it, but if you actually
do the math from before the ad
to the bottom of the ad, it's
only about a 14% drop, and it
seems like it's a temporary
drop.
So, it's doing pretty well.
So, with that, I want to switch
back to the slide, and I just
want to say that wraps up the
walk through and demo of Podcast
Analytics for you, and I hope
that you were able to learn some
things about how this tool works
and maybe how it can help you
better understand your audience.
And Podcast Analytics is in
beta, and we are very interested
in your feedback.
So, if you submit your feedback
link in the top left corner,
right next to the beta badge, we
would love to hear what you
think about the, about Podcast
Analytics, and give us your
feedback.
Thank you.
Now, let's bring it back to
Anne.
[ Audience Applause ]
>> Thanks, Alec.
That was awesome!
So, that's a walk through
Podcast Analytics, which, again,
is available to anyone with a
podcast in our catalogue.
If you haven't used it, try it
out.
It'll also be at a lab
immediately following this
session.
So, come find us.
And that's where we are today,
but really we've just begun.
You guys are the reason that we
did this, and we love hearing
from you.
So, keep it coming.
For example, we know that many
of you want an API or a
subscription and download data
or geographic demographics with
more specificity.
We love those suggestions.
We hear you, and we're working
on improving the product in the
future.
We are also going to be asking
for some things rom you in the
future, including new
specifications around HTTPS,
pubDates, consistent GUIDs, and
cover art that meets minimum
required resolution.
So, for continued access to
Analytics in the future, you
will absolutely need to have
those four things.
We're also going to be releasing
some guidelines around best
practices for audio mastering to
give you a sense of the ranges
you should be using so that your
podcast sounds the best it can
on HomePod and within the
context of the Apple Podcast
catalogue.
We'll be sharing these
requirements with you all.
We will put them in Podcast
Resources at Help so you can
refer back to them anytime.
And speaking of which, last but
not least, I want to remind you
about our resources for
podcasters.
So, we have a Branding Guide for
podcasters.
We also have the Affiliate
Program, and then Podcast
Resources and Help, which you
might remember from the top
right corner of the purple
screen that Alec showed you.
That features the hosting
partner list that James
mentioned; some tutorial videos
for analytics, and of course,
the ability to contact us for
help anytime.
For more information, you can
also visit this link or, like I
said, come join us in the lab,
the Apple Podcast Lab,
immediately following this
session at 10 a.m.
Thanks so much.
[ Audience Applause ]