Speaker: I think it's just a
way to kind of speed run the
downfall of your business.
I've tested ugly paywalls too, and they,
i- in, in, in the right app, it just rips.
I don't know what it is about it.
You do want as many users
as possible, in my opinion.
I have seen some apps
experiment with this as well.
Speaker 2: Hey, Jonathan.
I'm super excited to have
you on the podcast today.
Speaker: Thanks so much for having me.
I know this has been in the works
for a little bit, trying to, uh,
get on and have a chat together.
Um, especially after, uh, you asked
me to write your blog post, uh, uh, or
guest post on Retention blog last year.
So excited to keep the conversation going.
Speaker 2: Yeah, the, the newsletter
post last year was awesome.
I think one of my favorite posts
and other people's favorite posts,
where you got really deep into,
you know, tactical paywall tips.
I had multiple people message
me and say, "Oh my God, I
tried this, it worked great."
Tellâ¦
You know, "Jonathan, thank you."
Uh, and so, you know, we knew we
had to have you on the podcast
too to kind of follow up, and
that was like a year ago now.
So I'm sure- Right ⦠you know,
since then you're now working
independently on your own business.
Uh, you've probably tested, you
know, hundreds of thousands of
more paywalls since then and
have tons more things to share.
And so, yeah, excited to, to,
uh, uh, learn more about and, and
share with others how we can, how
they can optimize their paywall.
So, you know, I, I think you shared
that you've designed more than 4,700,
nearly 5,000 paywalls, which probably
is, is more than nearly anybody.
Um, I, I would love toâ¦
Let's, let's take a trip back in time.
Take it back to the early ones.
What do you think you were getting
wrong- Uh, early on redesigning paywalls
that you only see now in hindsight.
Speaker: Yeah.
So, uh, early on, um, I was one of
the early employees at Superwall.
So as employee number two, uh,
the way that we were handling
paywall experimentations was just
manipulating strings and changing
the text content in these paywalls.
Uh, and so, uh, we made, like,
small changes to colors, we made
changes to the copy, um, and, like,
that's essentially how that, the
product started, you know, then.
And I thought, like, "Okay, if
you change a button, uh, uh,
color," like, that was A/B testing.
I didn't know that much about A/B testing.
Uh, and it ended up, uh,
becoming so much more.
You know, now we have the state-of-the-art
where you have multi-page paywalls, exit
offers, win-back campaigns, um, and we've
essentially taken all the things in, uh,
paywall testing for mobile that all of
the eCom bros essentially came up with and
all the affiliate marketers came up with.
So now, uh, it, it's interesting seeing
that diverge now into mobile apps.
Speaker 2: Yeah.
So just, not just the individual elements,
kind of the holistic presentation, design,
motivation, psychology where, you know,
really, um, learning more about what
motivates purchases and actually what-
Speaker: That's right
⦠Speaker 2: uh, uh, changes behavior.
Any, um, high level kind of specific
examples you can give of, of,
like, what you were doing before,
and then maybe some tests now?
Speaker: Yeah.
Back then, um, we were making paywalls
in Webflow, actually, uh, funny enough.
Mm-hmm.
Uh, and, uh, now I think more about
the holistic approach instead of just,
like, "Okay, this is the paywall.
I don't really look at anything else."
Now you have to look at the entire,
um, app experience to see where does a
paywall make sense, uh, and what kind of
placements will move the needle the most.
And even more now, now that it's so
easy to build products, um, what isâ¦
Do, do you have some kind of
user acquisition strategy that
maybe someone else may not have?
Speaker 2: Yeah.
Yeah, 100%.
One thing you, you, you've mentioned
that you've said you can often predict
a test result Uh, before you even start.
When you're doing that, what are you
looking for in terms of those, uh, tests?
Speaker: Yeah.
I'm trying to see what
the app category is.
And so based on the app category
and what you've done before, um,
just how many tests you've run,
uh, if you haven't really started
running tests, I can expect to see
kind of like potentially incremental
gains and maybe a clear winner.
But it's, uh, it, it depends
on the app category and mostly,
and less on the app itself.
Um, that, that's not to say, though,
when ⦠Sometimes I do see, uh, a
really well-polished, well-built app
and, you know, they're just kind of
getting started, and I know when I run
an experiment with them, because they've
put so much effort into the product
itself, um, and they have such a clear,
uh, solid user acquisition strategy, the
results are, y- you know, going to be
increased in like 5 to 10% even sometimes,
uh, in terms of like conversion rates.
So there's, uh, uh, th- there's apps
that just have a ton of possibility
out of the water, and I can tell just
based on the polish of the product
and how clear the ICP is and are they
targeting them as, uh, effectively.
Speaker 2: Got it.
One metaphor I like to use for
marketing is marketing is kind of
the loudspeaker for the product.
If you have a crap product, doesn't
matter how loud you shout about
it, you know, it's, it's still bad.
And, and so if you have a strong
product, ultimately the possibilities for
conversion increase are much, much higher.
And, and so you're able to
predict test results by- Like
the quality of the products.
Um, but then how-- Like explain
to me how you layer inâ¦
Where does the industry come into play?
Speaker: Uh, the industry comes into
play pretty much immediately because,
uh, you know, sh- referencing my own
data, um, and seeing where this app is
in particular to every other app in, in
the app category that's like doing well
or maybe perhaps underperforming, I can
get a feel for, um, if, if you're-- if
you have 8% conversion rate right now,
for example, and you are not converting
that well, um, and I look at your, uh,
what you've tested already or I look at
your product, I can say, "Oh, you know,
there's a clear, um, issue with this
particular part of either your onboarding,
uh, or, uh, in, in the paywall itself.
You know, the, the value
offer is not clear.
Um, you know, let's change some things
up, uh, either into onboarding, uh,
to make that value proposition clearer
and, and increase the possibility
of then, uh, having a conversion
event on the paywall itself."
Uh, so it's, it, it, uh, when I look
at things holistically like that,
it's easier to kind of see what the
results of the experiment will be.
Um, that's not to say
I'm, I'm always correct.
Uh, I have had instances where even,
you know, running an experiment with
four to five different treatments, uh,
that I thought would be the control, um,
that did not end up being the control.
But out of, uh, since I went independent
with my agency last year, I've only had
two instances of that, in which case
I've just worked for free, uh, until
I get it right, uh, for the customer.
So I try to, uh, uh, if, if, if I'm
wrong, uh, then I try to make sure
that I'm not, you know, 'cause then
I end up having to work for free.
Uh, so it, uh, just kind of an
intuitive skill that I've developed
over the years, and it's by looking
at things, uh, in a super holistic way
and, uh, seeing where the gaps are.
Speaker 2: Got it.
So the industry component is
kind of using that as a benchmark
and a sense of understanding.
Um, for-- You've worked
with a number of apps.
You have, uh, benchmarks and
performance across industries.
You can go, "Hey, for your
industry, you're underperforming.
Um, I think there's opportunity here.
We can probably drive this up."
Where, you know, maybe for, uh, a wellness
or fitness app, like the conversion
rates are just much higher than for
a potential, you know, music app or
media app where people are used to
paying things or there's more emotional
value or, or something like that.
It, it-- and, and then you can take kind
of what you've seen work in the past for
those industries a-and apply them there.
Is that right?
Speaker: Exactly, yeah.
And even, you know, s- you may
have a freemium app, for example,
uh, and you're like- Obviously
some, I'm doing something wrong.
My conversion rate's only at one to 2%.
You know, what can I do?
And I'm like, "Well, you can remove the
free plan and then your conversion rate
will go up to, you know, eight to 12%, uh,
but then you'll lose all of your traffic."
And they're like, "No, I
don't wanna do all that."
So I'm like, "Okay.
W- there's, there's not much you
could do, you know, with a freemium
plan just because you're gonna have
so many users, uh, kinda go in."
So yeah, it's, it's-- And then, and then
I may decide like, uh, I d- I'm not sure
if I can give them enough value for their,
you know, return on investment in me.
And so I may choose, uh, to not work
with a particular app just 'cause
I don't want them to pay money that
I'm not sure I can make them back.
Speaker 2: This is a good, uh, topic
of freemium versus hard paywall
versus, you know, uh, no trial.
Do you have, um, clearly, uh, a
hard paywall will always drive,
uh, higher conversion rates.
Um, how do you think about this
in terms of the sacrifice of
building kind of this, um, free user
ecosystem that can support growth?
Uh, how do you weigh that trade-off?
Speaker: If you can support the free
users and i-i-in the sense that it
doesn't cost you money to support
the free users, I think it's okay.
So if you do, if you have an app and
the entire app experience happens in
the app and you're not n- you don't
have to send network requests out, um,
to either an LLM, um, or stream some
kind of video, uh, where-- So if your
inference costs are non-existent or
pretty low, um, I would actually say,
yeah, let's, let's figure out a way to
have, uh, a freemium offering, um, so
that you can focus on growth, uh, and,
and getting that large subscription base.
Because once you have users using
your app, you can monetize them later.
So, uh, you do want as many
users as possible, in my opinion.
Uh, but there are some apps
that, uh, any time a user does
something, it costs you money.
And for indies, you know, you can't
necessarily support that long term, right?
You're just gonna be burning
money, and you wanna validate that
this is something that, uh, is
useful in the marketplace anyway.
And so, uh, in that case, I would strongly
suggest doing a hard paywall with three
to seven-day trial and, uh, seeing how the
market, um, behaves in accordance to that.
Speaker 2: Yeah, I think that makes sense.
The world of AI apps is a
completely different world, right?
Where software used to be
basically free, um, i-in terms
of, you know, additional users.
I also think there's-- I'm curious
your take here, um, there's a trade-off
between starting with a new product
and validating that it's a, uhâ¦
people will buy it at all, uh, a-and
kind of the simplicity of not having
any free, uh, uh, freemium offering,
where you're able to kind of test
the product and value proposition.
Start with this hard paywall, where
you're able to just drive traffic to it.
Will people buy this at all?
Um, because I agree with you.
Ultimately, if you can build a
product with a free user base, this
is a much higher ceiling, uh, but
it's harder to get signal early on.
Is this something that people
are willing to pay for?
You can work on the kind of conversion
funnel optimization and, and convince
a, enough percentage of people to pay if
you have something valuable in the end.
But i-is that how you see this too?
To kind of like, a, a hard paywall is
super valuable for getting started,
but like maybe you, you wanna try to
transition away from it long term?
Speaker: For some apps, yes.
Again, if, if a free user just costs you
too much and you're not really able to
recuperate your costs from them, um, I
don't necessarily think it's a great idea.
Um, but if you can create features for
those free users that doesn't really cost
you money, uh, then I would say go for it.
You know, an example being you could
have a Calori type of clone where the
inference costs happen the most on, you
know, taking the picture of the food and
then calculating the macros for that.
Uh, that costs you money, but maybe if
you're logging, you know, water intake
every day, and that's also part of the
app experience, maybe that part is free.
So you're, you're giving users,
uh, free options, and then the,
the part of the app that costs you
money, uh, is what you feature gate.
Uh, and then you, you can keep
it like that as the app scales.
Speaker 2: Yeah, this makes sense.
This makes sense.
Cool.
So I, I alsoâ¦
y-you were saying, uh, that, uh, you
should be testing design before the actual
price, and I, and I think this is also
kind of referring to kind of packaging
of the products versus individually
just up-- changing the price up or down.
Can you tell m-me more about
kind of how you, uh, prioritize
tests, uh, between design, price,
price and packaging on paywalls?
Speaker: Yeah.
So, uh, at Superwall, we had an
opinionated testing system, uh, for
how to start doing experimentation.
And design is always kind of the
easiest to start with just because,
uh, there are so many elements in a
paywall that you can swap in and out.
Uh, like will you have a comparison
table, uh, accolades, user
reviews, um, you know, videos?
And are they, uh, paywalls
that don't scroll?
Do they scroll and are a
little bit more long form?
Um, are they multi-page?
Uh, do you include exit offers and,
and, uh, uh, transaction abandon,
you know, elements in there as well?
So there, there's just so much to
test there to find, uh, a, a good
baseline before touching, you know,
other placements and trying to connect
everything together that, uhâ¦
Because of the opportunity and just
the amount of testing that you can do.
Um, and then even personalization.
You know, show a man, if earlier you
selected your sex at birth is male,
then showing a man to, you know, m-
connect with more, or f- or female
if your sex at birth is female.
And so, uh, there's so many
things that you can do, uh, before
price testing that that's why I
like to do design tests first.
When it comes to price packaging,
um, depending on what the rest of
the apps in the app category do, that
is, uh, my second l- uh, lever that
I like to pull because, uh, it'sâ¦
You, you see a lot more
variance there as well.
Uh, so after trying a bunch
of different designs, how you
present the product groups.
So are you pushing the yearly plan because
doing, um, checking out your charts
and analytics, maybe you've seen yearly
has the highest LTV and lowest churn.
So, okay, fair, I wanna run a long-term
business, so I'm going to be pushing
the yearly plan, uh, and hide everything
else behind a view all plans sheet.
Uh, or maybe you wanna test yearly
versus weekly, yearly versus monthly,
or all three at the same time.
Or a design your trial, uh, portion
where you have a paid intro offer with,
uh, the yearly price or a seven-day
trial with the yearly price, right?
So swapping things like that, uh, you,
you can start to see also, uh, wh- what
kind of a difference does this make.
Uh, and pricing is, is after you've done
all of those tests, um, if you want to
test pricing and you have a bunch of
other placements, uh, there, the solutions
for doing that is, um, either creating
user seeds, uh, and having users kind
of fall into those buckets, and then,
uh, managing those cohorts manually.
Um, some of the no-code platforms, uh, let
you create those cohorts and track them
automatically, but it's a little bit more
difficult to track Um, and so it's, uh,
it's, it's kind of the last thing I do.
Once I find a design winner,
uh, then I'll start doingâ¦
And, and price packaging winner,
I'll start doing those price tests by
slowly increasing pricing to see, um,
where the conversion starts to break.
Uh, but it's usually worth it
though, uh, even though it is a
little bit more difficult to do
because it, it could increase your
realized, uh, LTV per paying user.
And so, uh, you absolutely
want to, uh, test pricing.
Um, another thing, uh, as
well that's, uh, onceâ¦
Or how you can approach price testing
is through, um, like demand scores.
So if you capture the device attributes
of a user and you show, uh, this user's
on iPhone 17 Pro, they, uh, are on
Wi-Fi and 5G or, or Wi-Fi and, you know,
don't have a low battery indicator,
uh, show them a higher price, you
know, potentially they will convert.
Or if a user's on a lower, like iPhone
13, uh, on 3G and, uh, low battery,
they're most likely not going to convert.
Uh, and so, you know, you could try
showing them a low- a lower price
to then entice them to converting.
Uh, so those types of advanced tests
exist, and after you've squeezed out
as much as you can, uh, people do them.
Uh, and even, you know, age-based price
testing, which, uh, uh, I know is a
slippery slope, uh, to do, but if you
are over 30, you can show a higher price,
and under 30, show a different price.
I have seen some apps
experiment with this as well.
Um, but I, I think, uh, some apps are
getting in trouble for, uh, doing this
recently, so I'm not sure if I'm a, a huge
proponent of those kinds of tests anymore.
Speaker 2: Yeah.
Generally, um, if you operate
in other countries, I would, uh,
not do price tests based on age.
Um, generally this is, yeah, uh, illegal.
Um, if small apps get away with it because
they're, they're small and, and, and
enforcement is hard and hard to discover.
But yeah, generally with the price
test, you, you, you don't wanna use
any protected class information.
Um, age is a gray area by country, but
age, sex, gender, you know, Tinder had
a lawsuit against them for- Exactly
um, uh, discriminating against gender.
We actually, um, someone from
Tinder told me they, they weren't
doing this essentially, but it was
kinda correlation and, and based
on sampling it looked like it was.
But, um, this is, uh, uh,
another, another story.
But, but yeah, I think that, um, and
this is, you know, showing the right
offer to the right user is, is a huge
value prop of, of kind of Botsy in terms
of, you know, building kind of custom
models to determine, uh, uh, what is the
right offer, what is the right price.
And so as you said, like optimizing
price, um, is hard, right?
Analyzing it correctly is hard.
Um, but even just in the packaging
side, this, this is hard as well,
and so this is what we've developed.
But, um, do you, do you see in terms
of the, uh, so, so the, to, to repeat
back to you the priority, um, you
think design, like test a bunch of
design variants first, then move on to
packaging, and then try to tweak price?
Speaker: So let me, uh,
I think restate that.
I think in pre-AI, that is my workflow.
Post-AI, you know, knowing that
paywalls are a quantitative,
uh, experience, so y-y-you
will learn the most when you can
quickly iterate on paywalls and
experiment as much as you can, right?
You, you definitely need a human
in the loop because you can't trust
the AIs to understand, you know,
psychology, price packaging, best
practices, all of these things.
But like what to te- what to test
and when to test it, the, the AIs are
pretty good at crunching data and then
making suggestions based on that data.
I've actually been surprised, um, at
some of the, uh, uh, what-- if, if I
try to run a test and, uh, get a second
look from, you know, Fable or Sol, uh,
they're, they're pretty spot on actually.
Uh, and so, uh, post-AI, I think,
uh, using the data that they have
access to, uh, to be able to run price
testing and design testing at the
same time, um, is, is a good idea.
Um, but I think it's, um, trying to
over or optimize for price first.
I'm personally not sure if
that's the best approach because,
uh, if you increase pricing,
conversion rate is gonna decrease.
Uh, so I say it's better to find a design
winner first that a design winner could,
you know, be double your conversion rate.
Um, and then once you start increasing
your pricing You know, you could get back
to the conversion rate that you previously
had, but you've now doubled ARPU.
So, uh, it's, uh, that- that's why I think
it's better to start with design tests.
And also pricing, even with the LLMs,
I think is still a little difficult.
Um, I haven't really found a, a great
solution but, you know, that'sâ¦
You- you've seen the same thing
and that's why you built Botsy.
Uh, so yeah, it's, uh, uh, I actually
would love to chat about how you're
setting that up and, and, um, uh, what
that supports 'cause it's, it's a huge
pain in my behind and I'm sure it is
in a lot of people who test as well.
Speaker 2: Yeah.
I think this is what I've seen as well,
like if you're on the design side, if you
have a poor converting paywall, theâ¦
It's gonna be hard to, to understand
the impact of a price test.
People don't understand your product,
doesn't matter what price you have.
Uh, what is the ceiling?
You have no idea because you're
just not selling the product well.
Uh, but once you kind of have this
better converting paywall, better
converting onboarding flow, um, then
getting this offer right, it, it,
there's this huge, huge upside for.
Um, and yeah, so I, I generally,
yeah, that's what I've seen
as well and agree with that.
Um, on, on the paywall design side,
you mentioned in I think a, a Mobbin,
uh, podcast you did, uh, that's, you
know, one finding was like the ugly
paywall wins a lot of the time, uh, and
versus kind of a pretty well-designed.
Like do you get frustrated here where
we spent all this time designing
this beautiful paywall and then
this just like, this ugly one where
we just slap in a bunch of social
proof and, uh, logos and, you know,
press mentions, this is what wins?
Speaker: Yeah, absolutely.
Especially because I'm a trained
UX/UI designer, and so I like to make
things pretty, I like to make things
look nice and enticing to the user.
Uh, but the CRO game is
completely different, you know?
It'sâ¦
You can be in your face.
I've seen so many tweets and articles,
um, you know, by Peter Levels.
Uh, and, like, he, he, he looked at,
you know, different landing pages
too of, uh, uh, different websites.
And even, you know, closer to what m-
maybe some people have seen, icon.com,
their website was notorious for having
a ton of copy, you know, above the fold.
Uh, just so much going on.
But in their testing,
they found it's, uhâ¦
it converted way better than something
more minimalistic or what I would
consider, you know, better design.
So, uh, you can't really argue with the
performance of, uh, uh, one, you know,
versus the other, espec- And if it's
an ugly paywall, um, it's just, um, I'm
gonna, gonna have to accept that that's
the, the winner if that's the winner.
Uh, and, uh, I've, I've tested
ugly paywalls too, and theyâ¦
In, in, in the right app, it just rips.
I don't know what it is about it,
but, uh, specifically in the, um,
the companionship, uh, space for,
uh, either, like, your chatting,
like, character AI kind of space,
uh, it, it does really well.
Uh, so that's, uh, uh, something
interesting for those of you listening
that are, uh, are in that space.
Speaker 2: Yeah.
I, I'veâ¦
You know, thisâ¦
We've seen landing pages where this
is all started with, yeah, you see
these ugly landing pages and, you
know, this is what's, what's winning.
And people do these tear-downs, "Oh,
they should fix this, this, and this."
It's like, well, no, like, this, this
converts well and this works really well.
I, I think it's, you know, interesting
to think about the psychology of why,
where, well, ultimately, like, your user
doesn't really care about your design.
They care about kind of functionality.
And if, like, your ugly paywall is
motivating, tapping into user psychology,
touching upon the points, drawing
attention in the right places, uh, uh,
you know, th- this is kind of what wins.
And so, um, I, I think it's, it's, uhâ¦
Sometimes you just have no idea, though.
Sometimes it's, there's no hypothesis.
You run a test and I guess
it's a, this is our winner now.
And, you know, you constantly
try to beat this ugly paywall,
and you can never beat it.
And I, I think this is a, a very common
experience for people kind of doing this
in the conversion optimization world.
Speaker: Right.
And, you know, ClickFunnels
is still a thing, you know?
They have a very specific design
language, and you can tell if
you click on a Facebook ad that
you're in a ClickFunnels page.
Uh, and some of those are the ugliest,
you know, squeeze pages you've
ever seen, but they work so well.
And, and, uh, you know, you, you
just gotta accept what, uh, uh,
what, what you end up seeing.
Speaker 2: Yeah.
I think it'sâ¦
You know, there's a lot
of parallels like this.
You know, quiz onboarding flows
that you have in apps too.
Like, like- It's just a bunch
of multi-choice questions.
It's like, why does this work?
You know, and, and you dig in and
kind of touch upon user psychology.
People are kind of inferring the
value when they read these things.
You know, there's kind of sunk costs and
the investment and, and so, you know, um,
I, I think to, to, to the point of like,
if you really wanna succeed, you have to
understand the user psychology as well as
kind of the, the different tactics and,
and, uh, uh, design and pricing where,
where that's kind of the cool thing about
marketing, where it's like this holistic
field where you're, you're pulling from
kind of all these different disciplines.
Um, so I, I, I think, um, we've talked
about this, this first paywall a lot.
I think that, um- A lot of times
people, uh, try to get a user to
convert once, and they don't realize,
"Hey, there's still intent there."
Um, te-tell me about some exit flows
you build where, where someone,
you know, closes the paywall, uh,
instead of just kinda letting them
kinda leave, what do you suggest
a, a, a great exit flow looks like?
Speaker: Yeah.
So, um, whenever you have a user
that maybe they have the intent
to purchase, but they're not
subscribing for whatever reason.
Um, recently I've been playing a
lot more with those questionnaires.
After they exit the main paywall, show
a questionnaire, and then instead of
guessing what it could be that, uh,
you know, they, they, uh, they close
the paywall, present them with an offer
based on what they end up tapping itself.
Uh, and I came across, um, or I settled
on this particular strategy recently
because what I used to do was if there
was an exit offer, uh, either offer them
a, a sheet, you know, a sheet could come
up with instead of a three-day trial,
try a seven-day trial or a 14-day trial.
Maybe the trial was the que-
the, the answer, you know, to
why they weren't converting.
And then if they hit "No thanks,"
show them a one-time offer, uh,
where, you know, it's, it's a 33%
off, you know, offer on the yearly
plan if they subscribe right there.
Um, and, uh, what I realized is this was
kind of cannibalizing, uh, revenue that I
could have, uh, that I was missing out on.
Uh, and but by offering a, a questionnaire
the-- and letting the user go through
kind of their own flow, uh, you can
modify the paywalls that, like, really
speak to the user's, um, reason for not
subscribing, and then showing a different
paywall based on, uh, the reasoning.
So if, uh, they wanted to try longer, then
uh, uh, you-- I would give them a 14-day
trial or even a paid 30-day trial, right?
Experimenting with those offers.
Um, if the price was too much, uh,
then I could try to get, you know, 25%
off, and if they still didn't do that,
one final offer for 33% or 50% off.
Um, I try not to go super crazy
in discounts because you're still
trying to run a profitable business,
and if you're running ads, uh, you
know, ideally, your average revenue
per user is above, you know, what it
costs you to acquire that user, um,
and after what you gotta pay Apple.
Uh, and so, uh, doing it this way
has let me, like, be more intentional
in what kind of offers I show users.
Uh, so the questionnaire form I
think is something that I'm gonna see
more of just because I'm doing it.
And so when people are ripping up, you
know, large apps and they see, "Oh, they
have a questionnaire, you know, I wonder
why they, they did that," you know.
It's, it's, uh, uh, it's, it's because you
have more control over, um, what, uh, what
that experience can be like for the user.
So I expect more people to see that,
uh, since I'm working on, uh, those
for some pretty large apps right now.
Speaker 2: What are your typical
questions you ask in that questionnaire?
What are the typical options you present?
Speaker: The price is too high.
Um, theâ¦
Because I give them a reason, you
know, like, "You were almost there.
What was, what was the reason,
you know, you didn't continue?"
And, uh, price is too high.
Um, the, the trial length
wasn't long enough, you know,
to see if this is for me.
Um, if there is no trial, uh, so
some apps don't offer a trial at all.
Um, but I do want a paid
intro offer, you know.
So maybe like for NGL, they have a
paid intro offer that's, uh, $2 for the
first week and then rebill at $9.99.
Um, and so experimenting with
different prices for, uh, for that.
Um, and, uh, there's also, uh,
I don't pay for apps, you know.
And for those users that say, "I don't
pay for apps," maybe you do hit them
with like a super steep discount,
uh, and because they were probably
not gonna pay anyway, as long as it's
still, you know, reasonable for you
to acquire that user and serve them.
Um, or something else and, you know,
ideally, uh, they, they could write about
the reasons for, uh, not subscribing.
Uh, but then you have NPS data that
you can, you know, reference to
say, "Okay, maybe this is a product
problem, um, instead of a pricing
or like, uh, clarity problem."
Speaker 2: Very cool.
So, you know, starting off with
price, trial, I don't pay for
apps, and then free response.
Figure out how to actually not just
show the offer sheet, but actually
personalize the paywall based on that
intent, and so you can customize things
a little bit to kind of have theâ¦
Give you a better chance of
success and kind of appealing to,
to whatever they're telling you.
Um, and then- Exactly ⦠with free
response, we can run this for a little
bit, go analyze the data, figure out if
there's any other typical reasons, and
then we can add to our survey to then
create kind of more, more options there.
I, I think this, uh, umâ¦
Yeah, I, I've-- to, to be honest, I've
seen this, um, and, and tried this and
that we, we couldn't get it to work.
And, and, uh, I, I think that,
um, but it's cool to hear it
kind of working well for you.
Uh, I think this is, uh, overall
like a really smart approach.
Like why guess, right?
Just ask them what, what- Right
the issue is and, and, and then
try to, try to account for that.
Speaker: Right.
I'm surprised that, uh, it, it,
it hasn't been working for you.
For the, the last couple experiments
that I've done, I've tried this out
and it's, it's working fairly well.
So I wonder, uh, uh- It was a while
Speaker 2: ago
⦠Speaker: where the
Speaker 2: discrepancy was.
Yeah, yeah.
Okay.
It was a while ago.
Let's talk a little bit about, uh, uh,
lifecycle marketing and win-back and I,
I think that, you know, we all focus on
this initial offer where rightly so, this
is where a lot of the value is captured.
But, um, neglecting, uh, uh, lifecycle
marketing and win-back and offers
throughout the lifecycle, I, I think
means that, like, you're, you're
leaving some opportunity on the table.
Uh, and, and, and so, you know, tell
me more about kinda what you're doing
with onboarding and cancellation
surveys and recovery flows.
Uh, why do you think win-back gets,
like, so little attention and,
you know, what do you think, where
do you think the highest leverage
points for, um, for most apps are?
Speaker: Yeah.
Um, I've been working with a lot of,
uh, newer apps lately actually, and, you
know, they, they always ask me, um, am
I-- is it too early to work with you?
And while I do enjoy working with
large apps because you can get results
faster, uh, just with the amount of
downloads that they have, um, I also
wanna set newer apps up for success.
And so the first, uh, uh, round of
experiments, you know, that I might do
for a, uh, new app is onboarding paywall
and a feature gate paywall, right?
Those are pretty much what, uh, you're
used to when you think about a paywall,
um, either onboarding or, or feature gate.
Uh, but what about transaction abandon?
Uh, you know, there's a lot of users that
will look at the price and say, you know,
close that sheet, and then they're done.
So there's an opportunity there.
So I'll also include a transaction
abandon or exit offer type of paywall.
Um, and okay, let's say that,
um, you did subscribe or, or, uh,
continue and you start a free trial
for maybe three days or seven days.
Um, if a healthy app will have
fifty percent of those users cancel
their, their trial and, uh, you
know, they won't do a renewal.
And so there's an opportunity there
to capture some kind of conversion
from those fifty percent of
users that, uh, cancel the trial.
And so you can target users that
have, um, their-- they have an active
entitlement but, uh, they'reâ¦
a-and still in trial, uh,
but auto-renewal off, right?
So send them some kind of offer the
next time that they start the app.
Uh, and then finally, for users that
have expired, you know, welcome them
back whenever they open the app again.
Ideally, um, if you have some kind of
onboarding where you capture emails,
uh, or, or, and/or they've enabled push
notifications, there's an opportunity
there also to kind of like send them some
kind of offer and bring them back into
the app to capture more of that revenue.
So those are, uh, that's what I would say
is my, um, uh, i-initial set of tests that
I like to do 'cause it covers, I think,
the most amount of ground for being able
to, uh, start off with a successful app.
So, uh, onboarding for sure,
feature gate, uh, on, you know,
whatever you're asking them.
So onboarding is maybe
if it's soft, right?
If it's a soft, uh, uh, paywall,
meaning that they can still close the
paywall and play around with the app
after they've gone through onboarding.
Um, feature gate paywall,
usually a hard paywall there.
Uh, transaction abandon or exit
offer Um, active trial, uh, but no
renewals, and then, uh, expired,
uh, entitlement of some kind.
So those are, uh, what
I like to test with.
Speaker 2: And expired entitlement,
expired trial, but also expired
subscription where, where-
Right ⦠anything actually ends.
Um, and then but also when they cancel.
And I think, yeah, I think like
thinking about these moments, right?
This is, um- I think a very important
concept in the subscription world
where, like, e-commerce, it's
kind of this continuous thing.
You're just buying other
more transactional products.
You can buy whenever.
Um, but s- and, and there's different
moments interact with the user.
But in your subscription product, there
are specific precise moments in this life
cycle, and basically you have to capture
in on those and, and use the intent there.
And ultimately, like, this is where
all the value is in, in my experience,
and it seems like yours as well.
Like, this is where all, like, 90%
of your attention should go and,
and, and this is where, where,
where a lot of the opportunity is.
So l- let's talk through a little of,
uh, or, or at least a few kind of tactics
or examples of things that worked.
Tell me about, um, you know, I, I'veâ¦
I'm curious your experience.
Trial, uh, um, expiration
or trial cancellation.
This one has always, I think, been
tougher, uh, to, to kinda get wins here.
Have you seen things that
have worked well for you?
Speaker: Uh, yes, and if, if they cancel
a trial, it could be a price issue.
Um, it could be that they're not
seeing value in the product itself, but
still showing a, a- an offer, right?
Uh, a- and being a good offer, but
not your best offer yet because
you do, you do wanna save your best
offers for either Black Friday sale
or Cyber Monday or for those subscri-
uh, you know, users that have expired
to kind of like win them back.
Um, I, I, I see that, uh, y- you know,
offering them to come back at like a
lower price or a different price point,
uh, sometimes users, they'll see a
yearly plan and it's just way too much
for them, but they'll be okay with like,
uh, th- uh, subscribing for a month or
a week because maybe they just don't
need the product for the full year.
Um, so typically I'll actually see the
average conversion rate on those kinds
of campaigns, um, at around 5 to 6%.
Uh, so that's, that's huge
because you're capturing 5 to 6%
of users that already canceled.
Uh, so, uh, it adds, you know, 2.5
to 3% to your overall
conversion rate at that point.
Speaker 2: Yeah, and
that, that, that is huge.
Yeah.
And, and so that's why I wanted to
touch upon that first because, like,
the other ones, when you actually
look at the total percentage of their
user base, they're a little smaller.
So if you can get a win there, because
ultimately, um, for most apps, there's
many more people canceling their
trial than converting on their trial.
And so it's a huge base of users that
have already shown like some pretty
high intent by starting a trial.
And, and so if you're not-- If
you're neglecting those, it, it's
usually a, a, a big opportunity lost.
Um, and, and so you think like still
generally just like discounted plans
but maybe different subscription
lengths, shorter lengths c- can,
uh, win some users back there?
Speaker: Right.
Even, um, it-- there are some apps
that they'll have kind of like
an ultra plan or- Mm-hmm ⦠or
like a max plan or a pro plan.
And so if maybe you subscribe to
the max plan and you're like, "Yeah,
but this is maybe super expensive,
so I'll just try it out and see
if it's for me," cancel the trial.
Uh, in the, uh, offer then for
after you've canceled your trial
to kind of like keep you as a, a
subscriber, maybe you'll offer them
the p- the, the pro plan, right?
Mm-hmm.
Uh, so it's-- y-you're not getting
as many features, but you're
still, um, uh, c-coming in and, uh,
staying with the product, right?
So to try and keep you there.
Um, something that I've, uh,
that I've also seen instead ofâ¦
And I like this better from a brand
perspective, um, if, if you're okay
with just in general lower conversion
rates because your realized LTV per
user will be higher and that's, uh,
everyone should have around like the
same price that they pay for a product.
But you, uh, to win a customer back then,
uh, like for if, if, if they've expired,
um, or if, um, they're an active trial
but don't have auto renewal on, uh,
give them only 33% or 50% off for the
first year or for the first- Mm ⦠um,
you know, part of that subscription.
Um, and that, that it doesn't,
um, uh, cheapen, I think, the
brand because then you can still
rebuild at a higher price later.
And if a user is getting
value from your product, uh,
they'll stick around for that.
Speaker 2: Yeah.
Got it.
Got it.
That makes sense.
Um, and so, so what about, uh,
transaction abandon, where this can
be a high percentage of people too,
and they're really high intent.
Uh, uh, w- what, what tactics do you
usually use for transaction abandon?
I know you can just remind them
and say, "Hey, didn't go through.
Let's try again."
But, but is there anything
else you do there?
Speaker: Uh, in transaction abandon,
there's the, uh, there's not a lot
that has changed in terms of what
I've already previously talked about
recently, um, or in like the past year.
Uh, so you still offer some kind
of discount or extended trial,
um, in transaction abandon.
Um, but these are users that are getting
cold feet for some kind of reason, um,
which is why I like the questionnaire now.
So, and transaction abandon
is, is typically where I'll
actually have that questionnaire.
Mm-hmm.
Um, and, uh, uh, th- then I can
change my offer just based on that.
Uh, but either a longer trial or a,
like, cheaper price for the first,
um, for the first year or the design
your trial, uh, type of paywall.
Um, those are what I'm seeing are, areâ¦
They're, they're still really good.
Um, I have not seen any new, uh,
designs that I've been, like, wanting
to try, um, because I get a lot
of my, uh, ideas actually from the
affiliate marketers and the e-com guys.
Uh, so I'm, I'm just not seeing anything
kind of new there that I wanna try.
Uh, so I haven't, um, uh, changed it to,
uh, or implemented it in the app space
Speaker 2: Got it.
Got it.
Yeah, I think the app space, everybody
copies each other so quickly, right?
Where, where it's often hard
to come up with new ideas.
So going outside to, to other
industries I, I think isâ¦
makes a lot of sense.
Um, what about for, um, churn?
Uh, expired subscription, canceled
subscription, someone purchased,
converted, uh, but, but is now, uh,
you know, ha- has canceled their
subscription or has ended already.
H- how do you approach those,
Speaker: uh, points?
Uh, those you, you, um, those
are users that you may have had.
Um, and, like, I do like to reward,
um, new users, uh, just as much as
users that you may, um, have lost.
Uh, and so, like, for, for those
perhaps like a, uh, discount that,
um, uh, th- that's when I would do
kind of the most severe discount.
So Cali Eye, you know, infamous
for 80% off or whatever.
Uh, you know, showing like a big
number that you have percentage off.
Um, they liked to have it in the
beginning, like in onboarding.
Um, I personally would like to
kinda save those, uh, for users just
later, uh, s- in, in the, um, user
journey for, um, just lifecycle.
So that's when I would do, um,
like a, a big discount for, uh,
they're just-- this is like your
last shot at getting that user.
Uh, and if, if the reason why they
didn't stick around is product,
perhaps giving them, like, s- such
a g- sweet offer will entice them to
stick around while you fix product
Speaker 2: Got it.
Got it.
Um, do you-
Speaker: I'm curious if, if,
uh, if you've seen, uh, any,
uh, anything different as well.
Uh, I'd love to hear your thoughts
on, uh, the life cycle- Yeah ⦠of
paywalls as well you've tried.
Speaker 2: Subscription cancellation is
hard, um, to, to win them back because
it's usually in the middle of their
subscription, they still have access.
Um, this is, this is much trickier
unless you're, like, you go,
"Well, why would I pay right now?"
You've consciously made some decision
that, "Hey, I, I don't want this
anymore," uh, and, and you've
gone actively and canceled it.
And, and so I've found that, like,
responding to that in the moment,
it will get you a little bit.
Like, you should still try,
like, "Hey," um, butâ¦
A- and, and then with Apple, um,
Apple tiers and subscription groups,
it's like, it's often tricky.
So it's like, what are we doing?
Like, are, are we shifting
them to a new plan within the
existing subscription group?
And, and it's like, but
why would someone do that?
They already still have access.
And so unless it's, like, close to
the end of their subscription, um, you
know, for, for that one, I, I see that,
okay, let's try with a web purchase,
um, to kind of shift them over to web
and with an email offer or something.
But you have to have the infrastructure
in place to correctly manage the,
uh, um, the access controls and
entitlements for, for web versus app.
And, and, and, and so, so it gets a
little tricky where, where I think
you just have less options, uh, there.
For expiration, when your subscription's
actually expired, um, or subscription
cancellation when you're getting close
to the, um, uh, expiration, you know,
you, you, you, like 50% discounts
hammer home the loss aversion, hammer
home that their, their subscription is
ending, what they're losing access to.
Loss aversion and a, a sizable
discount, um, typically I, I've
seen, uh, uh, work pretty well.
Uh, a- and, and I think that, um,
this is what I would start with.
I think if that doesn't work, like-
Probably like y- you have some
product issues and, and understand
like what people are, are missing.
I think as part of that, you also
wanna have this cancellation survey.
And so when someone cancels, probably
the most valuable thing you can do is
send them that cancellation survey to
understand why they're canceling, and
then try to address that and fix that.
Look, like a lot of times y- you're not
gonna get a lot of lift, uh, from kind
of talking to people at cancellation.
Doesn't mean you shouldn't try,
but ultimately probably the more
valuable thing is to learn from
that and improve that for new users.
Uh, and, and so this is,
this is what I've seen.
Speaker: Yeah.
Totally agree.
Speaker 2: Yeah.
Um, cool.
And then in, in terms of like, uh, email,
push notifications, SMS, um, the, the
impact ha- uh, uh, ha- has-- is capped
here and there's always a ceiling.
How much time do you, do you focus
on kind of those areas and, and y-
how do you think about th- those
mar- other marketing channels?
Speaker: Um, I, I don't, um, when
I'm working with customers, uh,
typically, just because that's an
entirely, like, different set of
problems that we have to address.
Okay.
Uh, and so, um, I, I may bring it up
that, you know, we could do this, uh,
if, if we wanna try doing something here.
Um, but typically, uh, when I'm working,
especially with, with like large apps,
large apps are the only ones that have
this infrastructure in place, uh, anyway,
and they have, uh, their own team kind of
working on that part of, uh, life cycle.
So I personally don't ever really, uh, w-
work in that part with, with these teams.
Um, but I do bring it up because
I, I do think it's important.
It's important, um, to, uh, at least like
have some activations, um, in those other
channels Um, and like you said, if w-
you know, someone has made a conscious
decision to cancel, it's really hard to
get them back without offering something.
Perhaps a discount isn't enough.
Um, and subscription apps, if you
don't have more subscription apps
that you can perhaps cross-sell or
upsell, um, after checkout, then, uh,
it's, it's differ- it's more difficult
to like entice them with something.
So yeah, totally, uh, uh, y- you know,
get why I think a lot of people are
struggling to, um, uh, figure out like
this part of, uh, the customer journey.
Speaker 2: Yeah.
And push notifications
used to be super effective.
Um, they're just overdone by every app.
Apple has been constantly, uh, uh,
minimizing kind of their visibility.
Uh, rightly so because like, you
know, as marketers, we, you, you
abuse everything till, till it
no longer, uh, kinda works well.
And so what I see is that
like my, my general heuristic
is, um, look at your revenue.
Uh, if, um, you could probably increase
revenue from life cycle marketing by
five, 10, maybe 15%, um, over like
continued focus over a year or two,
like if you, if you keep pushing this.
And so like think about, okay, if
you're making a million dollars a year
What investments make sense if you can
improve revenue 10%, um, and kind of
base your, your effort based on that.
Mostly, in my experience, set up
triggered campaigns at all those
lifecycle moments we just talked about.
Exactly.
Uh, new user, transaction abandoned,
uh, trial canceled or expired,
subscription canceled or expired.
And until you get to a few million
in revenue, like leave it at that.
Like- Yeah ⦠unless you have something
inherently about your product where like
notifications are part of the product and,
and kind of build the product experience.
Um, so this is my, my quick advice
and, and my quick take there.
Um, so something else I wanna
talk about is, uh, weekly plans.
I, I would love your perspective on
when it's right to use weekly plans.
They've gotten hugely popular.
I think personally they're probably
overused because, uh, they-- y-
you see more revenue up front and
people are bad at estimating LTV.
Um, but, but there's
certainly a place for them.
Uh, what's, what, what's your take
about when the, the right time
and place is for weekly plans?
Speaker: Um, I'm, I'm also not
a huge fan of weekly plans.
Uh, people look at the ARPU,
you know, of weekly plans and
they say, "Oh, this is great."
Uh, but I, I think it's just a way
to kinda speed run, uh, the downfall
of your business, uh, just because
you're just-- your churn is gonna be
so massive, um, that I-I-it really just
seems like a money grab, you know, to
me if you just offer, uh, weekly plans.
Uh, so I'm not a, I'm not a huge fan.
I am a fan of them, uh, with either
paid intro offers, so you pay for like
a week's access and you're maybe paying
like a dollar or two dollars, uh, or
you have, um, some kind of product
that you know is either seasonal, um,
or the user is not going to use a lot.
So then you're like, "Okay, you know, uh,
for-- I have multiple ICPs in my product.
One of them, uh, is-" Only gonna use this
for a week, so that's why I'm going to
offer, you know, the weekly plan to those
users instead of, uh, monthly or yearly.
Uh, and in the hopes that,
okay, this weekly user, um, paid
for two weeks and that is 1.5
times, uh, or, or 50% more the
cost of like a monthly plan or
something, and so you're making
a little bit more money there.
Uh, and, uh, so that, that's
where I'm seeing, uh, maybe the
best use of having a weekly plan.
Uh, but only weekly plans, um, I have
seen this multiple times in the last,
you know, year or two where you're seeing
an app, they make a ton of money because
they have great user acquisition and
are going viral on TikTok, and they're
having millions, uh, tens of millions of
views per, uh, month and it's, you know,
eventually they just hit, uh, a wall.
So, um, it's interesting to watch.
Speaker 2: Yeah.
Yeah.
There's the, the cohorts
don't stack over time.
Um, okay, well this was awesome.
Our, our, uh, my last question,
this is the Price Power Podcast,
so we ask everyone, uh, what's the
biggest single pricing or packaging
win you, you've had or seen?
Speaker: Biggest win, um,
that I've personally seen
are the multi-page paywalls.
Uh, y- once we started, like, implementing
those, uh, y- and that's the, the
video with the, you know, we want you
to try Cali AI for free, whatever.
Uh, it, it, it really created
kind of a resurgence in, uh, uhâ¦
Or it kicked off a ton of apps,
you know, kind of doing the same
thing, and I think they kept doing
it because it just worked so much.
Uh, but it was just so crazy seeing
how well that paywall performed.
Uh, and I s- I still use it, you know?
So it's, uh, uh, a, a paywall and price
packaging design that I think works well.
It's just one price, big buttons,
uh, big text with a video.
It doesn't get any more clear than
that, I think, for a user, uh, that,
yeah, just press continue and, uh,
everything's gonna be great in the world.
Speaker 2: Yep, yep.
I've, I've seen those, those
be big winners as well.
Um, okay.
This is awesome, Jonathan.
I really appreciate y-
you joining the podcast.
So many great learnings to share.
Uh, and so go follow Jonathan on X.
He's got gr- lots of great, uh, tactics,
insights he, he shares on there.
We'll have the, his link to
the profile in the comments.
And, and then, uh, y- you run Tapas
Growth, and so if you're, you're
looking to have kind of an embedded
subscription and paywall growth expert
on your team, go reach out to him.
Uh, anything else you
wanna promote, Jonathan?
Speaker: Uh, no, just, uh, please, guys,
if you have any, uh, paywalls that you
see that are super cool, send them over
my way, uh, because I'm always, uh, s-
you know, looking to see kind of what's
new and, and working that I can try,
uh, since, you know, you may have a
question like, "Oh, could this work?"
But may not have the traffic
to, like, test it yourself.
Uh, I definitely have the traffic,
you know, to test your ideas if
you see something interesting.
So, uh, send over, uh, uh,
anything interesting that
you guys see out in the wild.
Speaker 2: Amazing.
Amazing.
Yeah, send it my way too.
I wanna see it also.
Um- Cool.
Well, this was so great, Jonathan.
I really appreciate, uh,
you joining the podcast.
Speaker: Uh, likewise.
Thanks so much for having me, Jacob.
Speaker 2: All right.
Thanks.
Bye.
Speaker: Thanks for listening.
Hope you enjoyed.
Please go to price power podcast.com
to see all the episodes.
Go to Spotify and YouTube and
give us a subscribe and follow
so you don't miss any episodes.
Alright, talk to you next time.
All.