Listen to Fairing on Subscription Radio with Ben Fisher
Ben Fisher ,
Subscription Radio’s Host, and Fairing’s own Matt
Bahr discuss the importance
of attribution and how collecting data can help companies understand
their customers better. This episode focused mostly on how to use
customer surveys to build a competitive advantage for ecommerce brands.
Just a few of the discussion points:
- Survey data helps solve customer attribution
- Build your competitive advantage with customer feedback
- 5x the average survey completion rate of other companies
- Four types of question categories
- Surveys made simple
- Learn about your customers over time
If you’re not already subscribed to Subscription Radio, be sure to give
this episode a listen
here
.
Key Takeaways
Survey data helps solve customer attribution
1:53
At Fairing, essentially, we bring speed and scale to survey data is how
we position ourselves today, where we capture that data is like 99%, on
order confirmation pages. So we fit into that kind of post-purchase
survey category. Of course, what we're building beyond that is way is
way, is way greater. But today, that's where the majority of capture
happens.
And the reason why we built this, honestly, as a side project back in
2018, was really to solve attribution, or at least help provide another
data point into the attribution equation. When we initially did this, it
was with a Google Form, it was for a consulting client of ours, and the
data wasn't connected to the order data. And that's where the lightbulb
went off in our head around, oh, we could just like build a simple form
and connect this. But that was kind of square one. I think the zeitgeist
hadn't caught up to, or wasn't at this kind of
zero-party
, first-party data world yet. So we, I don't want to say patiently
waited, two or so years before things started to align with where we
thought it was going.
But yeah, as far as the data goes, like we again, we were founded on
attribution, we power about a half a million, “How did you hear about
us?” questions a day at this point. And … attribution is still that
wedge. It's what literally 100% of our new customers come in and that's
the first question they ask.
Build your competitive advantage with customer feedback
3:25
What we're really excited about is we're seeing customers add more
questions into their Question Stream™, which is like this stream of
questions that are our product can serve to end consumers.
And that's where things get really interesting for our customers,
because first they just had this attribution data. So they're asking
questions like, “How did you hear about us?”, for example, and let's
say, TikTok, it's obviously a very growing channel right now. There was
no way for them to understand that specific channel audience. So you
have your channel distribution, which like, if we were to look at a
typical client, we're looking at like Facebook, Instagram, Google,
TikTok, word of mouth, podcast, and TV.
As you scale up, it was very difficult to understand who the customer
was from each channel. So as we were chatting before, we kind of hit
record on this. Facebook Audience Insights used to be like the way you
upload a list of emails, and then you understand demographic data and
can understand who these specific cohorts of customers were. And until a
lot of these brands kind of graduate and start kind of enriching their
datasets with third party data, which again, is getting more and more
difficult to pull in. It's very hard for brands to know like, who is my
TikTok customer? On the retention side, it's like, okay, I have a TikTok
customer.
They are a very low churn subscription customer, like, who is that
customer? What's their age? What's their gender? What did they buy? How
would they classify themselves from a persona perspective? And that's
where our product has morphed into with this Question Stream™ is
allowing merchants to help solve attribution, but also understand
channels and like more deeply their customers on such another level than
they have been before.
I think the thing that's happening is like, it's no longer kind of money
in money out scenario where you just put money into most like literally
the Facebook platform for so long and you get money out, brands are
looking for competitive advantages, like there's, it's very difficult to
operate and scale without doing something proprietary at this point for
a lot of these brands. And that's where like we come in, we like
literally give them a data set their competitors have zero access to.
5x the average survey completion rate of other companies
7:15
Yeah, so it’s [survey completion rate] a little over 50% platform wide
on ours. So it's like a ton of data. Like, I think most people, when
they think of surveying, especially via email, they're expecting, like,
I think I'm on the bright side, maybe like 10%, let me a little a little
less, if they're providing an incentive, maybe 15-20%. And then you
start getting into some biases around surveys, if you have an incentive
bias, where people are just completing it for the incentive.
So we are like 5x, essentially, any expectation and that's, that's
definitely been a big factor in our growth. And it's also a reason why
we're not like jumping to go do email, SMS, and like landing page
surveys, is because our, our completion rate is so high right now, like
that's a core value prop of what we offer.
Four types of question categories
8:40
So there's, there's a few categories. So we we grouped them into four
categories. So we have attribution questions, and really the core two
are How did you hear about us? And when did you hear about us like, the
when helps you understand that window, like, if you have a very long
purchase consideration window, like you shouldn't move away from any
kind of direct response reporting, because it's not gonna be very
accurate. If you have a very short then like, you can almost look at GA
and help yourself scale. So attribution is the first one.
Personalization, and segmentation is the second. So these are questions
that like, think about if you're building any kind of flow in
Klaviyo
or any other ESP, these are questions that you could filter on. So it's
a question like, “How would you classify yourself?” Oh, you're a
hobbyist or a professional.
It's like, okay, this is a hobbyist flow, let's talk to this persona,
very personally, as if they're a hobbyist. They're not someone who's
using this product as a professional. That's a second category that
we're really harp on.
The third is CRO, which is a little bit like out of left field. Because
it's less of connecting the data to the customer, but it's more of like,
how is your shopping experience like given the point of where we live? A
really good feedback loop is to turn that question on for a week. sort
through all those open ended responses as people that are like I
couldn't find your return policy or like the most common one, frankly,
is like, I got email the code and it didn't work. So just ways to
improve the customer experience.
You can also ask, “Why did you buy?”, and a customer could say XYZ,
okay, now you go use that copy in your creative, like creating a really
good feedback loop to help increase conversions.
And then the fourth is just this category of general customer research.
And this is questions like maybe age, gender, that you're not going to
like filter necessarily in Klaviyo or use for personalization, but it's
really informative. And … my favorite question … is just “How would you
classify yourself?”, it's like the easiest thing to then create personas
that, honestly are quite difficult to build if you want to really action
on them. So anything that would help you there, we work with this, like
hardware company that literally has that example. They have Hobbyist,
Educators, and Professionals. And previous to asking that question, he
thought that like 80% of his customers were Professionals. After asking
that literally within like two days, he goes, yep, completely wrong,
it's like 20% Professionals. I have this whole cohort of Hobbyists who
are buying our product, and then they can then hey, let's maybe look at
Reddit, let's look at some other channels where the Hobbyist hang out.
On the customer research side, this is the one that we like to kind of
join with that attribution data. So questions like, “What brand are you
switching from?” on the CPG side, I know a toothbrush, toothpaste
company we work with, they were using that and all their investor decks
because they were like saying, like, look at this, like we are literally
taking market share from Crest and like all these core toothpaste
companies, which I guess was very informative, and then kind of raising
the next round of capital. And then also like, you start to join that
data with your survey data. And you're like, oh, we're taking a lot of
people from this channel who are this demographic and are switching from
this brand. Like, what can that tell us? Like, maybe we can do like a
viral TikTok campaign that like mentions this brand, and how so many
people are switching from it kind of thing.
So there's some sort of really good feedback loops to help optimize
creative, optimize ad copy, and just prove improve, like conversion and
CTR across all your channels.
Surveys made simple
13:56
Surveying’s obviously been around forever, like on the political side,
polling is obviously a huge business. And I think with a lot of CPG,
like B2C brands, like surveying in its historical sense, which is very
much research based like definitely was not like a popular thing to do
or thing that was definitely preached on doing and if you were going to
do it, let's say you were going to do it quarterly and send out a
Typeform or SurveyMonkey survey, like crafting that survey was always a
like, challenge, essentially, it was like, “How do I write all these
questions?” “Are these the right 10 questions or 20 questions?”, kind of
thing.
Where with our product, since we've decoupled questions, like that's the
whole concept of Question
Stream™ ,
let's serve them over time. They don't all have to be captured at once.
It reduces a lot of that, like not necessarily fear but like friction of
like getting started and asking questions to your consumers, where
historically like you literally had to spend maybe a week crafting
questions like if you're at a certain scale, you were probably working
with an agency to make sure you weren't like writing them in a very
biased way or whatnot. It was like this big project.
I know a company that spent like $100 grand on a quarterly survey with
like one of the big four consulting firms. And where we come in is like
removing a lot of that, like hesitation around like, I don't have to
build a question, literally just add a question like, you want to go ask
about where people are switching from, like, just go add it right now.
And like, you'll get data within the next half hour as long as you're
doing, like some volume. So it's definitely like, especially for that
brand in particular, it like started with “How did you hear about us?”
Then oh, let's just asked these other questions, and pretty quickly,
they realized that they were solving a ton of problems.
Learn about your customers over time
16:39
Let's say we want to ask you like 10 questions, for example. If you were
to answer three of them on your first order, those 10 questions, it's
not like a monolith where we're like, it's loaded on the page, and you
answer all them. If you were to answer three of those. And then let's
say you got distracted, someone texted you, you didn't answer questions,
four through six, the next time you place an order, it's going to pick
up right where you left off. And the rules could have changed in that
period of time. Like maybe the marketer was like, let's actually ask
this question before this one.
So essentially, like the way it works with our servers, like every time
you load a page, we tell you which question to ask. So this is where
we're gonna, like we're spending a ton of time here will be like,
getting more intelligent with when we're asking questions like, “Hey, we
know Ben never answers this question.” So like, how about we don't show
it to him? Because we know it may piss them off. That he’s like don't
ask me that question. So but that's what I mean by like, decoupled where
each question essentially has its own rules, and can be served over
time. And like you and I might get different questions based on our
purchase behavior.
18:00
I don't know if you're familiar with the term progressive profiling. But
that's the whole idea of like, okay, we're … trying to learn about this
customer. We obviously can't ask him, these 30 things [at once]. … And
so it's building that profile over time. And that's what we want to get
really good at where it's like, it's gonna be impossible for you to
build 100% of that profile on that first order.
But if a customer is a customer over the lifecycle of a few years, we
can get smart about building that profile, and like a really good
experience. So like, not overloading the customer with like, 15
questions on their first order kind of thing. And then get targeted with
hey, we don't need to ask him this question or this question because we
know it's not going to be overly helpful.