Revolutionizing CX with AI-Enhanced Direct Mail
Dive into a fascinating session with Michael Epstein, a seasoned D2C entrepreneur now at the helm of Postpilot, as he navigates the evolution of customer acquisition from early e-commerce days to the present. Learn about Postpilot's inception, a platform that seamlessly merges traditional direct mail with modern e-commerce tech stacks like Shopify and Klaviyo, making direct mail campaigns as intuitive as email or social media campaigns. Delve into how this transformation was driven to address the clunky experience of traditional direct mail campaigns and how Postpilot has grown to cater not only to retention but also customer acquisition, providing a full-funnel approach.
(upbeat music)
- Michael Epstein here from Postpilot.
Michael, welcome to the AICX Virtual Summit.
Glad to have you here.
- It's like to be here, Brian, thanks.
- Nice, well, we were talking a little bit beforehand,
but we're just gonna dive right in.
I guess first and foremost, Michael,
can you give us a little background of yourself?
I think what was interesting to me was the amount
of years of experience you have in D2C.
- Yeah, I am an old-timer.
I started my first e-commerce brand in 2000.
Like Shopify didn't exist.
Google didn't.
Google AdWords didn't exist.
Like none of us existed.
And yes, go to it in 2013 to private equity
and gotten to private equity operating roles
with the notable companies for a while after that
and then Postpilot, here I am.
- All right, sorry, one question.
Customer acquisition at that time
compared to our customer acquisition now,
is it, was it easier, do you think,
is it easier to acquire customers back then
before Google Ads, or is it now so much easier
with so many different channels?
- It was different.
It was different.
It was just so many things that you see now
or sort of have their genesis from things
that you were doing before, but they weren't,
you didn't know what they were called at the time.
But just as a example, we blew up by partnering
with bloggers in specific categories
who talked, who reviewed our products
and talked about our products on their blogs.
These were like old school blogs
where people have websites around specific topics,
which would now be called like influencer marketing.
And things like that.
So it was different in that it was,
you didn't have as many ways to get your name out,
but there also wasn't nearly the same level of competition.
So yeah, it was just very different.
- What goes around comes around,
just basically the same nature,
they just renamed everything.
And now here we are.
- One way or another, they just evolve.
They just keep evolving.
- And now you are the co-founder of Postpilot.
So, and you told me just four years ago,
so tell us a little bit about Postpilot, Michael.
- Sure, so Postpilot is a direct mail platform
that natively integrates with Shopify, Clavio,
your e-com tech stack and makes sending direct mail
as easy as like email or Facebook campaigns.
And really the genesis of Postpilot was
we had used direct mail successfully as a channel
in my career, throughout my career,
working with different portfolio companies
and different brands.
But it was just the stereotypical clunky experience
of a bunch of spreadsheets and finding a printer
and then a bunch more spreadsheets
to try and figure out if it are a wide for you.
And it just sucked.
So the thesis was like,
somebody needs to build Clavio for direct mail.
And so that's what we set out to do.
And that's what we did.
And then it really started out as more of a retention platform,
like you're sending campaigns to existing customers,
but over time we really evolved it
to a full funnel platform that helps with acquisition as well.
- When you say Clavio for direct mail,
can you go deeper into like,
what about Clavio outside of, yeah,
they do email and SMS that you are bringing in to Postpilot?
- Sure.
So you think of your email program,
you think of Clavio or Sennlainer,
these other email service outliers
and you're creating segments.
You're defining a segment of customers,
like somebody, I want everybody that has bought for me
more than once and not in the last 90 days.
And you're sending, you can send campaigns to those,
one of all campaigns to those customers.
By, in your ESP, you're sending these email blasts,
think of like newsletter type announcements,
like new product launch or sale announcement
or things like that.
And then you're also creating automated flows
in your email service provider.
So triggered automations, think of your winback campaigns,
your abandoned cart campaigns, your cross-sell campaigns.
When somebody makes their fifth purchase,
you wanna send them a thank you message.
You can do all of that with Postpilot
the same way you would with an email,
with email, we're literally triggering
individually personalized cards to go out to people
based on their behavior in real time.
So Brian, you buy a 60 day supply of supplements.
You go 65 days without placing another order.
We tried emailing you to tell you it's time for refill.
You haven't come back on day 65 trigger a card
that says, "Hey Brian, it's time for refill.
"Get back on track.
"Here's 10% off your order if you come back now."
And we send you that card in the mail.
- So this is very aligned with the previous conversation
that I've had about first party data,
second party data, third party data.
And tell me if I've run it wrong on this.
But if you look at just direct mail,
I feel like it started with a lot of third party data
of like, okay, like, try to send to people that you think,
like I've worked with yourself and saying,
all right, try to find my target market
and send this stuff out.
And now I think the table is turn where it's all about
zero party data, first party data,
and utilizing as much of that in all of your direct mail
is that, do I have that right?
Would you kind of explain that the same way?
- Sure, so it's actually both.
So when you think about on the retention side
and even on the retargeting side,
it's using a lot of that zero party data.
It's using your existing customers and transactional data
to target people based on their purchase behavior.
Or if you're thinking about retargeting type campaigns,
we can actually take an email address
from somebody that subscribed to your list
but never bought anything from you
so you don't have their postal address,
match that to a postal address and retarget them
with a card.
We can match anonymous site visitors to a postal address
and retarget those folks with a card.
But what you're talking about is really on the acquisition side
and knowing that this is more of like an AI
sort of machine learning summit that we're in right now,
it's, this is where you start to really be able to apply
a lot of that.
There's a couple of ways that you can apply it to the channel.
But in prospecting, that's a key way
'cause we're building lookalike audiences
the same way that like Facebook is using AI machine learning
to build lookalike audiences for Facebook called prospecting.
Typically in the past,
you've relied on kind of like superficial attributes
to target people.
I just want all females 25 to 49 in these zip codes
to get a card.
Like that's not a great idea.
But what we're doing for brands is we have thousands
of different attributes across hundreds of millions
of US consumers.
And we can take your first party or zero party data
and use that to build lookalike models
and then model it against these large consumer data sets
that have thousands of different attributes.
So we're looking at things like,
okay, yes, you're, you might be a female 25 to 49,
but you make this household income.
And we know that you're, you have kids
that are between ages two to five.
And we know that you shop in Target twice a month.
And we know that you buy meat jerky products from Target
and you buy this type of beer and you drive this kind of car
and you are an avid outdoors person
and you are health enthusiast and you, you know,
you have Disney Plus subscription.
Like there's literally thousands of attributes.
And this is where again, that sort of AI learning
starts to really get interesting
because we can model your first party data
against all of these attributes and find people
that look very much like your core customer.
Super useful on the, the call of acquisition side.
- So it sounds to me like where the AI is super impactful
is just that you are thinking about so many different
attributes and have the availability to look
at so many different attributes
that you kind of need an AI tool to better understand that.
Do I have that right?
- Exactly.
There'd be no way to like model,
create these models that factor in all these thousands
of different attributes against huge data sets of customers
without that type of technology.
Be different if you were just saying,
I just want to target all females in this.
- Yeah.
So then if I'm thinking about this right,
like it's a formula of the TAM
and making sure the TAM is right on
and the message and the personalization of the message
and making sure that is right on,
is it like almost a 50/50 split of like,
listen, you have to give as much attention to the target,
the TAM as much as you is about the message
or would you say one comes first to them in next?
- Yeah, I think it's,
what we really advise brands is
you should really have a well-defined customer.
So think similar to Facebook,
there's a lot of correlation to like how Facebook
shows acquisition and prospecting audiences.
Like if your product is super generic
and your TAM is everyone in the world,
that's actually not a good fit for these types of technologies
where they want to be able to identify
what are the distinct attributes
that makes your customer, your customer.
There's an example, this isn't quite like AI-based,
but as a good example of this,
we work with a baby clothing brand
and they want to target expectant moms
in their third trimester.
Like we actually know that information
based on due dates that is available to us
from tens of thousands or hundreds of thousands
of expected mothers, we can back into like their due date
and where they are in their pregnancy.
But that is a very specific attribute
that you want to ensure is present
because you could look like an expected mom,
but if you aren't an expected mom,
that person is like not relevant to you in any way.
So you want to make sure that certain attributes
that are really important to you as a brand
can be identified so that we can really zero in
on those distinct things that make your customer
really relevant to your business.
- Is there a line?
Is there a line between personalization
and then intrusiveness?
Like do you talk through that with certain clients?
Does that come up at all or no?
- Yeah, the interesting thing is people do ask that
and direct mail tends to have,
like it's, you don't get like the cringe factor
that you do if you get like a retargeting ad
or even an email from a brand,
particularly that you don't know
or haven't done business with.
Because people have gotten cold catalogs
and cold direct mail pieces from companies
for like all of their existence.
So it's actually, we actually don't hear that really ever
in that people actually like receiving highly relevant
direct mail, they perceive it more like a gift
than an interruption, like a digital ad
or an email.
And so no, people actually respond to it really well
and you don't get that sort of yeah, like stalker factor
that sometimes people feel when they're getting
like a super personalized retargeting ad
after they visited a site
and they like told Alexa they're interested in something
or something.
- Well, really quick story, like personalization
is vastly improving and I know that for a fact
just because I look at the holiday, like the Amazon
and all these different toy stores, the magazines
and all the stuff that I'm getting now.
And it used to be my, I got two kids
and it used to be like they'd circle a couple things
that they want and now they're circling every single toy
that's on the page because it's very highly relevant
and it's clear that they've learned exactly what my kids
actually would enjoy and stuff like that.
And so like that personalization is getting to a point
where it's right on.
So let me kind of take the next step
which is that I know you were talking about
clavio integration, shopify integration.
So someone that is new to this technology like myself
that maybe some merchants that are listening,
how do you integrate with something like Shopify
as an example in the e-commerce platforms
and how is that fruitful for the merchant at the end of the day?
- Yeah, great question.
So super simple, we have a Shopify app.
You install it with one click the same way you install
other Shopify apps.
And once it's installed, like we can then create these segments,
these audience segments, the same way you would create them
like in an email platform or your SMS platform.
- Yeah. - Or you would sort of segment
your audience that way.
And then also because of that integration,
you can track performance of those campaigns
in real time dashboard.
This isn't like the old school way of direct mail of again,
60 days after a campaign goes out,
you're trying to put a bunch of spreadsheets together
to figure out who of these people do I see
in my customer database.
We are, we have a real time dashboard
because we're integrated with Shopify.
We know when a customer receives a card,
we know if they go on to make a subsequent purchase,
we know if they use coupon code on the card
and we attribute that back in real time with that.
- Oh, interesting.
Oh, I did not know that you were able to do that.
Oh, wow, that's very tightly targeted.
And I'm sure the merchants then,
I mean, that's gotta be enlightening for the merchants
they have an understanding of like,
okay, here's actually the results coming from these campaigns.
- This, it really is game changing.
It really opened up this new growth channel
to brands that historically would have never known
where to begin and also just don't,
it doesn't feel, it never felt familiar,
you know, it was just so different from how you would execute
the types of campaigns that brands were used to,
like a Facebook campaign, like an email campaign.
- Yeah.
- We really made it feel very familiar and easy.
- Now, do you have clients that,
most of this conversation has been centered around acquisition,
obviously retention and transactions, right?
Getting something to like come to the website again
and make that transaction.
Have you dealt with clients or even refer,
or recommended to clients that like,
well, like there are other messages that you can utilize
or these could just be another touch point
in the customer journey.
And what I'm thinking about is just like,
the celebration of customers is common theme right here.
Do you have those conversations or merchants
that utilize this, not just for that transaction,
but just to continue the relationship?
- Yeah, great question.
So yes, in a number of ways.
Interestingly, I just heard from a customer this morning
who was talking about how one of her best customers
had just gotten her Black Friday card from us
and called in to say, thank you so much
for thinking of me and sending me this card,
which we hear not immediately,
people feel like it's this tangible thing
coming from a brand.
And that does create a different type of relationship
and it's a different level of engagement with a brand
and just sending them an email or sort of a digital touch point.
But the other things that we can do to your point
around celebrations,
we can actually match your customers to their birthdays.
We have the ability to take your customer list.
And even if you've never collected their birthday,
like through a form or something else,
we can actually match that for you.
And then we can send them birthday cards
that are personalized.
And it's just like a great touch point from a brand.
And often brands will include something like,
and here's a gift from us,
take $20 off your next purchase.
And they were super well from an ROI perspective,
but they're also great from a general,
just how many people are sending you
like a birthday card in the mail?
Very few.
And so when it comes from a brand,
you remember who sent that to you.
You remember that brand.
And then we also have a handwriting technology.
(upbeat music)
(indistinct)
And writes with like all the nuance of a human hand on paper
and can write like a handwritten card to these folks,
whether it's a birthday card or a VIP, thank you card.
You just made your fifth purchase from us
or you just spent over a thousand dollars
all with us.
So I want to trigger a card that goes out
from my founder that's like,
just want to say thank you so much
for being such a loyal customer.
We appreciate you.
And just making them feel seen like that
and getting a handwritten note,
you know, to your point,
it's a really memorable touch point with a brand.
And that, you know,
that's what we're all really competing for
at the end of the day, right?
It's like a tension and creating a memorable experience
with our customers so that we're not having to constantly pay
to get in front of them again and re-acquire them all the time.
I think, I mean, you nailed it.
Like I just want to stop right there
because at this point,
and I'm just sharing conversations that I've had
with several different merchants,
like looking outside right now,
there's a difficult, you know,
we still have a difficult economy
and there's just a lot of hardships going on
around the world at this point.
And if you look back,
just at the last 10 or 15 or 20 emails
in your personal inbox, right?
I guarantee you,
there's not going to be many that are celebrating you
or things that are like,
hey, like nice work or whatever it is
that you're celebrating.
It's mostly about like,
hey, getting someone to buy something.
And so I do think that there is such an opportunity
that's like, just stop and like,
can you figure out ways to celebrate your customers
through a channel like Direct Mal,
which is highly unique?
Because those are what you just said, Michael,
memorable moments that make a difference
in the customer journey.
- I love you're absolutely right.
- So, okay, let's really quickly,
like we'll go into the next stage, which is,
you know, I know we're talking about like current state
and current state to me is also future state
because these are things that are still brand new
to the ecosystem.
But are there things that you are working towards
or you just see a lot of innovation outside of
what you're working on that you get excited about,
specifically around customer experience and AI?
And is there anything even with post pilot
that you're, post pilot that you're, you know,
looking forward to when it comes to some next
in the developments in this AI space?
- Yeah, I mean, the thing that I can say is that we are,
we are very interested in AI and how that plays a role
in what we're doing and makes things even more personalized
and more timely.
So the fact that we have all this transactional data,
there's, we can use AI to be even more precise
around the timing of how, when certain messages are triggered
and what messages are most relevant to somebody
at any given time.
So I think there's a lot of opportunity there
that we're excited to continue to tap in terms of,
and just getting even more precise.
And you're seeing it with various platforms
in terms of their predictive analytics
and predictive triggering around certain events.
I think that that's a place where this gets really exciting.
- Well, I would also say, well, triggering
is probably one of them, right?
Like where, right now it's humans making that decision,
obviously of when to send something.
But I could see in the future utilizing AI,
having all these moments, the customer, all this data
and say, when is the bright time
to send this message out, right?
- As CDPs continue to become, you know,
more relevant and more prevalent.
That's it.
I think that's where a lot of this is going,
is like being able to be more precise
and more personalized with what message at what time
and what channel to what person.
And, you know, I think we're excited
for what that opens up for us.
- That's awesome.
Well, right there, we'll end it, Michael.
I appreciate all your time on this
and learning a lot more about post-pilot
and kind of the future of what's to come.
But again, thank you so much for your time.
- Thanks for reminding me, Brian.
I appreciate it.
- No problem. Thanks, all.
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)
(upbeat music)