Latest / Inside Commerce: Ecommerce Strategy, CX and Technology Podcast / EP155: Personalised Product Discovery for Ecommerce Using Klaviyo & Octane AI, with Paase Digital
Transcript
- 0:02Thanks for tuning in against three platform.
- 0:04Podcast sponsored by ambulance and clavia listen to myself.
- 0:07James, skirt and my co-host. Paul Rodgers, how's it going
- 0:09today? May?
- 0:10Yeah, good. Thanks, how are you doing?
- 0:12Yeah, I'm good. I'm embracing another freezing.
- 0:14Cold Monday morning. But looking forward to this
- 0:16episode. So thanks for everybody tuning
- 0:19back in. And if it's your first time,
- 0:22welcome, we hope you enjoy it. If you subscribe, we send out
- 0:25new episode alerts every week. We'd also love you to give us a
- 0:27like, on YouTube Spotify. Apple, it makes Us feel good
- 0:30about our own inadequacies. So let me just set up the topic
- 0:33and I'll introduce you to our guests stay.
- 0:35So, we're talking about the impact of integrating play, Bo
- 0:38and Tate AI to create personalized Communications.
- 0:41Focusing on one specific business, Cole Smith.
- 0:45And the pretext of this is basically quizzes, have become
- 0:48really popular e-commerce, as businesses seek new ways to
- 0:50engage customers and products covering by richer online
- 0:52customer experience. But, however, as with any shiny
- 0:55new bit of Kit, success depends upon the quality of
- 0:57implementation. How tiny the quiz.
- 1:00So, each audience needs and also have tackles their pain points
- 1:03and decision processes for product Discovery.
- 1:05So, today we're talking to team a highly experienced agency.
- 1:08Whose focus is on using customer data to drive business growth by
- 1:11us CRM. Automation, we're going to cover
- 1:14using an evolving proposition. Recom, the impact of a quiz
- 1:17versus standard roots of What discovery and smarter ways to
- 1:20use data to drive personalization.
- 1:23So let's do interactive stop with Phil.
- 1:24Let's start with you. So Phil and Dave both work at
- 1:29pass. And fills the MD.
- 1:31So you want to give people a bit of an introduction and we had
- 1:33over to your brother. Yeah, no problem at all.
- 1:35So, hello everybody. So I set up passed back in the
- 1:39summer 2019 pre those heady days of covid really to build on my
- 1:46background of being Tech side and working with large
- 1:49Enterprise businesses to try and bring some of that skill set
- 1:51knowledge understanding down to the the SME market.
- 1:55So and I only let businesses at the time.
- 1:58What? Trying to do, and what we didn't
- 2:01quite successfully, is to translate that customer data
- 2:04into actionable marketing activity.
- 2:07So, rather than, and trying to move businesses along from just
- 2:10broadcasting, General newsletters all the time to
- 2:13really focusing on that. Customer Journey piece, we got
- 2:16involved with klaviyo very early on back in August 20, 19.
- 2:22And actually, we started working with curl Smith back in or
- 2:25September October 19 when they were on at the time, my watch.
- 2:29Should we help them migrate over to klaviyo bringing in the in
- 2:332020? So yeah, really our focus is as
- 2:36an organization is using Brands customer data to drive increase
- 2:39retention. Cool exit yet.
- 2:42It's amazing that how many still do Mass?
- 2:44Broadcasts are not, targeted Communications, it's Bonkers.
- 2:47But yeah, really important approach to have.
- 2:51And so Dave, and I know we talked about before we started
- 2:55recording this that your brother kindly Stitch you up and brought
- 2:57you in on this last minute on a Monday morning.
- 2:59So do you wanna give everyone on the podcast an introduction as
- 3:02well, please. Yes, I'm Dave.
- 3:05I'm Phyllis River. So the commercial commercial
- 3:07manager, side more sales LED now, but I did a lot of work
- 3:10with the car. It's About Time quiz so it's
- 3:13nice to be on the call to sort of discuss.
- 3:15What we did? Yeah.
- 3:16Wonderful. Thanks right?
- 3:17You ready for some questions gents?
- 3:20Let's go for it and that's it right, Paul.
- 3:22You can start as using my absolutely.
- 3:25And so the first question and I think, maybe I'll start of you
- 3:28feel. And if you want to add anything
- 3:29Dave, go for X, in terms of the kind of business challenge to
- 3:34you, typically solve for the brand that you're working way of
- 3:37around, kind of CRM and personalization.
- 3:39That what is that challenge? Like why did Colonel Smith
- 3:42originally come to pass? Okay so I think there's this,
- 3:46there's two things there I think Brands sometimes don't
- 3:52articulate or Define what they mean by personalization.
- 3:55So quite often someone has said they need, we need to do
- 3:57personalization, nobody really knows what it is and you've got
- 4:00some people thinking that's as simple as a first layer merge
- 4:03you've got other people thinking, I need to be a bespoke
- 4:06one-to-one messaging, service of you, is there somewhere in the
- 4:09middle? There's something commercial.
- 4:11Bible around brought person, broad-stroke personalization.
- 4:15Now, the biggest challenge we find when Brands want to start
- 4:18to delve into personalization, they haven't got the data points
- 4:21required to deliver that Journey.
- 4:23So quite often, you know, all of us are consumers around on the
- 4:26school. Will go into a website where new
- 4:29to that site and will give an email address or the sunny.
- 4:32We're expecting that brand to give us a personal Journey.
- 4:34Where is the only piece of data? We've shared is our email
- 4:36address. So actually Brands and start to
- 4:39collect more data, but it's historically.
- 4:41Eric Lee based on transaction. So Phil's been on to curl Smith,
- 4:44he's brought a product, whether that's the right product for
- 4:48Phil or not, the brand doesn't really know, but it will start
- 4:51to make some assumptions. Get RK feels brought product a,
- 4:54we're going to sing more content about product a, where we want
- 4:58to move businesses through to his having that two-way
- 5:00communication do then starts. They've actually fill, you don't
- 5:03need products. A, you need product, b, c, and d
- 5:06for these reasons. And thinking about how do you
- 5:08bring that kind of online shop assistant to?
- 5:11I saw you walk into the store and you are somebody say I've
- 5:15got this type of hair or I've got this problem and there's a
- 5:18recommendation May. So this is where we're using
- 5:21octane to start to really drive that personalized experience by
- 5:24listening to what people are saying taking that data in and
- 5:26then asking a different question based on that that answer
- 5:29basically. It sounds and I guess onto the
- 5:33quiz more specifically so after you've kind of set it up and let
- 5:38it run for a while. How is this evolved?
- 5:41A timeline of your kind of Taken learnings and adapt to quiz, and
- 5:45then what kind of things have you learned about the customer
- 5:49to stages tonight. So I think, for us, the quiz
- 5:52itself is been tweaked and an evolved.
- 5:54Get one of the key lines, we got quite quickly as is how to serve
- 5:57the quiz so you can serve it in a number of different ways.
- 6:00Yeah, the traditional on the, on a page on the site within the
- 6:03navigation, please confine to or obtained actually has its own
- 6:07kind of pop up Builder, so you could build it within a pop-up,
- 6:10what we found I am. Dave, can correct me on this if
- 6:13I'm wrong cuz I'm going back a little bit into the depths of my
- 6:16memory was actually when we hosted it on the pop, up the
- 6:19conversion rate was less than it was doing natively on the page.
- 6:23So actually we change back to a klaviyo pop up and use the
- 6:26pop-up to right people through to the online quiz within the
- 6:29host. So it's thinking, almost not
- 6:31only the questions you're asking about how you're serving the
- 6:33technology to make the best use of it.
- 6:37And then, you'll have to forgive me with the second part of the
- 6:39question, what have you? You learn about the customer
- 6:42from the quiz, okay? So I think, within seven months
- 6:46we gained a hundred thousand new subscribers brand-new, not just
- 6:51having an email address, but having profile information and
- 6:53product recommendations about those, those individuals and
- 6:56Dave, and I think was about 85,000 of the existing database
- 6:59within 7 months had gone back through re subscribed in effect
- 7:04by going through the process. But also added all that
- 7:07information about themselves. So we suddenly knew what type of
- 7:10curly hair? They had weather.
- 7:11With heavy curls, like curls, how thick the hair was.
- 7:15And the products that we would recommend, was that the cal
- 7:17Smith recommended for those individuals that in turn, then
- 7:21allows us to start to tweak the messaging that we're giving to
- 7:23the customers by using iCloud. Your product blocks to show in
- 7:27high different pieces of content based on those answers.
- 7:30So again, very quickly you're taking data and inputs from from
- 7:34the customer them directly and then serving the back yo abroad
- 7:39commercially viable level. Real strong product
- 7:42recommendations, that make sense.
- 7:46And then I guess, just go and I went for the quest the other day
- 7:49and it's obviously gray. And I can see how you could
- 7:51learn a lot of information around potential customers, it's
- 7:54pretty long. And I was going to ask how
- 7:57you've seen this before on whether you've tested it versus
- 7:59a shorter version, I've seen others in the past, talk about
- 8:03longer kind of quizzes or processes being more like a skin
- 8:06in the game mentality in maybe getting more kind of subscribers
- 8:10at the final point. But yeah, what have you got any
- 8:12kind of learnings from the kind of length of the quiz?
- 8:17I think from from, from my son I think is probably one of the
- 8:19longer quizzes. We built and done nor we've sort
- 8:23of had with our clients is engaging my.
- 8:25So people are going through it people on dropping off.
- 8:28So the numbers suggest that actually the length is, the
- 8:31length is right for them. Is the question is, will people
- 8:35drop off the question 1? Question 2?
- 8:37What does that engagement? Like if people gets bored, then
- 8:40becomes an issue but people know that going through that quiz,
- 8:42they're getting a group of recommend recommended products.
- 8:46It also says in the quiz, that actually only takes a two
- 8:48minutes. I think it says on the course,
- 8:50so having that sort of time frame in there, that, that is a
- 8:53key to people. Go as you know what, it's only
- 8:55two minutes and I know what I'm going to get at the end of it.
- 8:59Yeah, that's an interesting point.
- 9:00I've seen this, when I've been doing work people on customer
- 9:02Insight surveys, which often gets sent out by email, because
- 9:05that's the main Optimist and telling people explicitly on the
- 9:08starting page of a survey. This this takes less than 5
- 9:12minutes to complete. If you leave it blank, you find
- 9:15fewer, people get past the first page.
- 9:17When they see multiple step. There you go.
- 9:19Our can't be asked because we're all quite lazy human.
- 9:21So, yeah, interesting that you've done that and found that
- 9:24I also like to the point you made earlier for, I think, to
- 9:27really point It was around this personalization because seeing
- 9:30those clients when, when you say the word personalization to a
- 9:33business, that doesn't have a big data tame.
- 9:35The interesting thing is I'll cry.
- 9:37So I need a really expensive tool or it's impossible.
- 9:39We don't have the resources because its massive and then
- 9:42breaking it down and pull it out specific capabilities and say
- 9:45you could get here within six to eight weeks and we can measure
- 9:49and we think the uplift will be eggs and Things by far the best
- 9:52way. And I've seen that before people
- 9:54on their email comes where they just picked, right?
- 9:56Okay. We're going to take first time
- 9:57buyers. A warming, welcome series and
- 10:00add a level of personalized communication.
- 10:02They wouldn't otherwise have got and then you suddenly get the
- 10:05interest people out, all we can do this and then we move on to
- 10:08the next step. I think quite often Brands think
- 10:12about the outputs rather than the inputs to drive the
- 10:15personalization, so that it will be a, we want to be able to do
- 10:18this and they haven't got the data points to be able to
- 10:22trigger that and actually backfilled those trigger points.
- 10:24I also think we talked about commercially viable, a lot.
- 10:28Of well actually if you've got a segment of I don't know 10
- 10:32people that are brought the white T-shirt in size medium.
- 10:36Well actually, if you then send a follow-up email to those 10
- 10:38people but time, you've got a, you know, a 50 percent open rate
- 10:40that's down to 5. You get 10 percent click-through
- 10:43to one or two people, as a three percent conversion Webster
- 10:47suddenly go, is it actually worth the effort?
- 10:49And it's how do you how do you get those businesses to go?
- 10:51Well, if you've got ten thousand people, maybe split it in half
- 10:54and 50/50, if you've got a hundred thousand that's way,
- 10:57maybe you can start Going into, you know, 10% Parts because
- 11:00you've got that relevancy, but start Broad and take it on a
- 11:03journey. I think the brands that struggle
- 11:05to do that are the ones that try to go from Zero to Hero
- 11:08overnight. And it's almost impossible.
- 11:10You'll spend your all your all your time, almost creating the
- 11:13crew, the creative variants, that you will need to drive that
- 11:16personalization versus actually, the value of just getting an
- 11:19email. That's broadly.
- 11:20Personalized out the door. Yeah. 100%, I've done work in
- 11:24previous within people in training programs where it's
- 11:26right again. Yes, you can do it but you're
- 11:29trying to do too much. Exactly.
- 11:30That is, you don't have the data, you don't have the
- 11:32analysts, you can't possibly do the analysis to know what's
- 11:35working and what's not to pick pick a segment where you have a
- 11:38decent, say, like, you say, and it often like first people have
- 11:41bought ones but haven't bought a second time and using them and
- 11:44do it segmented based on the catcher a ball or the type of
- 11:47product depending on what vertical you're in?
- 11:49Yeah, the nice way of framing this I think.
- 11:51And it feels if you're a new business and you you're not
- 11:55quite sure if your retention rates of How the product, you
- 11:59know, repurchase frequency is easier.
- 12:01Maybe if there's a subscription style business way, you're
- 12:03selling a product that's going to last 30 days.
- 12:06But if you're by, I do selling T-shirts or trainers, people
- 12:10will buy a different Cadence and it's actually starting to work
- 12:13out what they are. You know, you can have a really
- 12:15defined, beautiful personalized Journey, but if I've just made a
- 12:19purchase I'm unlikely to buy that t-shirt, 50 pound.
- 12:22Again that week, it's about trying to understand which leave
- 12:25is to pull up which times so God, I've got an additional
- 12:29question to throw it, you guys, I'm interesting because I've
- 12:32seen I know Paul's done a lot of work with quizzes, things like
- 12:35octane. I've worked a lot of business is
- 12:37we have not implemented it. I've been I've worked with them
- 12:40to help them understand how they could use them and the
- 12:42commercial benefits and I've seen in Industries Beauty
- 12:46skincare, specifically worked very well because there is a
- 12:50specific personal influence to the purchase skincare for
- 12:55example, with any spin specific skin issues, You got, I can see
- 12:59where the hair like, you know, depending on what type of hair
- 13:01you've got if you've got really dense.
- 13:03Yeah, not for here. For example, if the hair care
- 13:06treatment you need is totally different to somebody with
- 13:08thinner finer hair. So therefore, the more data you
- 13:11give and the more info you get back, the greater value, you're
- 13:13getting as a customer, do you find that there are you know,
- 13:17there's just certain types of products that don't suit the
- 13:19quizzes well, or do you think it can generally be adapted to any
- 13:22product set? My personal View and again,
- 13:26people made it easy. It can be adapted.
- 13:28So we know we've set it up for dog food manufacturers.
- 13:32We've got it for companies with trainers and it's about
- 13:36thinking, well, actually, what's this, you know, how do we better
- 13:39serve and personalize the future comes to using it, you know, not
- 13:42from a data acquisition perspective.
- 13:43So so well, actually we know Flynn is, I know size 10 shoes,
- 13:47he's in the UK in this market. It's like, there's so much value
- 13:50there. And actually, if the value
- 13:52exchange to the customers that we're going to make some
- 13:54Translations to you, the back of this by giving this information.
- 13:57So I think it can work. I think there are certain
- 14:00sectors, maybe the gifting space where it's going to be.
- 14:03You're not buying for yourself, you're buying for somebody else
- 14:05that there may be a challenge there.
- 14:07But generally I think I liking it too, you know, the concierge
- 14:11in a shop or the shop assistant, you're coming in, you're asking
- 14:14a question, you're asking for a recommendation and that
- 14:17customers use it. Well, so that's shop assistant
- 14:20is using you answers, you're giving it to make a
- 14:22recommendation. It's that same process.
- 14:24Yes. Yeah, that's an interesting
- 14:26point about the gift one. I saw.
- 14:28I've seen it done very badly and guess what?
- 14:29Big space illegally. Everyone gets the same gift
- 14:31recommendations this but yeah. And you're basically on a
- 14:33category page filtering by price and it's not the point of that.
- 14:38There's a way to do gifting correctly and I think you can
- 14:42build it the right way. So we one of our, one of our
- 14:45coffee coffee companies, we work with do, do a gift in section.
- 14:48But the question is, it Taylor's Taylor's that are you buying for
- 14:51yourself or you might have somebody else that he literally
- 14:53goes off of different? Because you can type that you
- 14:55can no flavor profiles. For example, you can know what
- 14:57somebody, like some, uh, some he dislikes as part of that gift in
- 15:00process, therefore, you can start to get recommended
- 15:03products, but you have to design in a way that klaviyo know, that
- 15:06you're buying the person who's a gift to this, not your profile
- 15:09properties. So, if you're then going to do
- 15:10stuff with in the email, that you're actually dictating the
- 15:13right email to the right person at the right time.
- 15:15Yeah, I think I'd be older. Sorry.
- 15:19Connor, can you say the art of design?
- 15:21I think the principles. The same as survey design you.
- 15:24You can, you can get a very poorly designed survey and have
- 15:27a really crap response rate, if you don't think about the
- 15:29audience and the value to them. Yeah.
- 15:31I mean, the lovely thing about octane is the fact that you can
- 15:34run multiple quizzes almost nested.
- 15:36So you can start to say, well, actually, the first question is,
- 15:39are you buying for yourself or, you know, for somebody else and
- 15:42then split them in two different quizzes, based on those profiles
- 15:44on those answers. There's nothing to say that you
- 15:48do a, you know, an almost a new customer quiz and then do a
- 15:51follow-up to Legacy customers. The Patrick session.
- 15:54So you can run multiple quizzes within the same functionality,
- 15:57which is really helpful. I know it's a leap year leap on
- 16:02from when I started doing stuff like as years ago before these
- 16:05tools listed and things like surveygizmo because the users
- 16:07face isn't quite as attractive as using an old tapes now.
- 16:11And I think there's other pieces out there that can do kind of
- 16:16customer feedback, but it's that the first, the power is how you
- 16:20get that data into klaviyo and then use it as actual data.
- 16:24Rather than look a retrospective look back of.
- 16:26Oh yeah, that was buying, you know, someone was buying for a
- 16:28gear. In fact, is we know that we can
- 16:29action that from day, 0 onwards, what is a lian's in a word, it's
- 16:36Freedom, the freedom to build a digital experience as Limitless
- 16:40as your vision create preview schedule, and manage, all your
- 16:44content in one easy place. Find out more at a lien.
- 16:49Start calm and cleans experience freedom.
- 16:55Yeah, exactly. And in terms of Jake's we've
- 16:56been talked about data 0 first-party data is becoming a
- 17:00massive landgraf Recon teams because all the problems around
- 17:02third-party data and privacy and what you can and cannot use what
- 17:08percentage of users who start a quiz, do you typically see are
- 17:12submitting and get into the end of it versus those that drop
- 17:15out? Dave's meet silence this for him
- 17:19on his behalf, the wonders of online.
- 17:21So we're about 99 percent completion rate on the cal Smith
- 17:25because so it is very high. And again I think it's about two
- 17:28pulls Point earlier. Yeah, making sure you're not
- 17:31asking for too much but you're providing that value in that
- 17:34context of why you're asking. So I think the col Smith quiz is
- 17:38a fantastic one but it's a complicated or more complicated
- 17:42area than I ever believed possible for.
- 17:43Yeah, ma'am for sure. They're just once a buyer wash
- 17:47and go and go the complexities are there.
- 17:50So the question there are relevant and I think that
- 17:52refinement is there you take a dog food quiz it?
- 17:56Maybe have you got an allergy or you a puppy or an adult dog or a
- 17:59senior dog, you can you can do do the same results in two or
- 18:02three questions. So it's, you know, it's about
- 18:05tailoring it and not and also thinking about what you're going
- 18:08to use that data for, I don't believe in taking data just for
- 18:11the sake of taking data, I think there's got to be a context and
- 18:13a reason behind, we're going to have to action this day.
- 18:16To moving forwards. And this is why we're taking it.
- 18:20And now just expanded on that. So the next question is, those
- 18:23who do abandoned and maybe not just in the context of curl
- 18:26Smurf, but maybe with other brands as well.
- 18:28Do you ever try and recover those users or customers?
- 18:32And what level of impact is that?
- 18:35So if we know we've got email address on site and they're
- 18:38they're clear your cookies. He did don't go through the
- 18:40standard sort of about him and series messaging that we have in
- 18:44there. If I don't, if they don't put
- 18:48their emerges, we don't have any light at all time data.
- 18:50So it's hard to sort of personalizing it.
- 18:52Yeah. We could we could retarget them.
- 18:54I think if someone's dropped out of a quiz it's always like if
- 18:57someone's unsubscribe from an e-mailed you then retarget on
- 18:59social media, for me it's like they dropped out for a reason.
- 19:03Is there any point retargeting if they're going gone to buy
- 19:06then potentially we look at potentially really pushed it
- 19:09repression and back for the quiz, after us for me a first
- 19:11time, buyer is not a customer there, try list and they're
- 19:14really only just tried in the and so it's about and get them
- 19:16into a second one. So if they haven't brought the
- 19:18right product because I didn't complete the quiz then
- 19:20potentially we would look at that kind of Branch out.
- 19:23Be since I'm just giving them a slightly different message.
- 19:25I think I think Paul raises an interesting point here as well
- 19:28because there's a, you know, a thought process and a potential
- 19:31test about, where do you ask for the email address?
- 19:34So you can go through, you know, a quiz and an affect the results
- 19:39page is gated by an email. So actually do you ask for the
- 19:43quiz you know email address up front in which Case you could
- 19:45then potentially retarget people versus the convenience of
- 19:49getting through and go. Okay, just one more question.
- 19:51Okay, enter my email address so that data depends on where
- 19:54you're asking for the data. Now, I can save what's right or
- 19:58wrong. I think my personal preference
- 19:59is to ask for that email address kind of at the end especially
- 20:02with an incentive of will give you X percent off your first
- 20:05order when you complete the quiz type thing.
- 20:07So there's no matter the first-party cookies there, but
- 20:09it might not be tied to an email address at that point.
- 20:13Yeah, that makes complete sense. And next question.
- 20:17So have you like what? From a purely kind of conversion
- 20:21and transactional perspective? What is the impact like of
- 20:26sending a user for a quiz versus, you know, standard
- 20:29landing page or collection? For example, does in your
- 20:33experience does quiz typically perform better.
- 20:36I know I've definitely seen an increase in people doing
- 20:40Prospect and activity via social and sending People directly
- 20:43through to quizzes. Like what's your view on that?
- 20:45It's a tough one and I can again comes back to Brands quite often
- 20:51just doing a quiz and not doing that kind of a be testing of you
- 20:55know quiz versus no quiz. I think it's about the value
- 20:59that data. And also because Smith where
- 21:02what works really well for their quiz is that they did some
- 21:05social media marketing to push people to the quiz.
- 21:09That's where the real benefits are you.
- 21:11You're driving traffic to To it versus just hoping traffic will,
- 21:15you know, the trip over the site on their usual, browsing
- 21:18activity will. So I think there's a couple of
- 21:21ways here. I know, Dave's got we had some
- 21:23really strong results with the email that's fed off the back of
- 21:27the curl Smith quiz. So you complete the quiz, you
- 21:29may go on and transact but you're going to get a results
- 21:31email, and Dave think that was about one and a half percent of
- 21:35their total digital Revenue with liver, just from that one, email
- 21:38alone to not even email, total digital revenues.
- 21:41So very, very Oil. You could argue that some of
- 21:45those would have gone on to complete their transaction.
- 21:47Anyway, it's not kind of ring fence with a an a/b test, but
- 21:51that one and a half percent was actually slightly smaller number
- 21:55than as you know, that there's actually three and a half
- 21:57percent. But we would have assumed that
- 21:59we were actually taking some people away from the site back
- 22:00to the email. So I kind of justify what we
- 22:03would have lost our other, so the welcome journey and those
- 22:05kind of things. So one and a half was a was a
- 22:08decent number. That makes sense and how have
- 22:13your client to have carlsmith bound work and octane.
- 22:16Like they found when I first worked with October's quite a
- 22:21while ago. And number the merchandising
- 22:23side was quite time consuming. And, you know, you had to build
- 22:26out different sets of results. In some of the templating was a
- 22:30little bit inflexible in places, like, how have they found
- 22:32working with? It is like how those areas come
- 22:35along since it was first introduced.
- 22:37So I think there's two sides to this.
- 22:40One is Part of the biggest challenge with Chris's is
- 22:45generating the I've got a bit of feedback story is almost
- 22:49starting at the bottom. So with your results will what
- 22:52you trying to hit how you try to segment the customers to give
- 22:54them the results otherwise you can, you know, you can be too
- 22:57complicated. I think that's one of the almost
- 23:00the pre-sales element of working with Brands is that they want to
- 23:03try and get too much out of it from day day one.
- 23:05It's about actually, how do you do those broader brush steps and
- 23:09then moving forward, The platform itself, you know,
- 23:12continues to evolve. I do think they've introduced a
- 23:16custom CSS elements to it versus what it was originally where it
- 23:20was purely linking into the Shopify product Library.
- 23:25We do a lot of work with them, you know, as a partner of
- 23:28theirs, to develop those pages in the quizzes, on their behalf
- 23:31so that they need heavily on our skin.
- 23:33Thats that sense. I think with all of these,
- 23:36there're if a brand is highly, highly School of Design, and as
- 23:45a bespoke site, I think there's definitely challenges or, you
- 23:48know, in these scoping properly to really get it integrated into
- 23:52a platform, you know, to a custom site really well, a
- 23:55simple. Your Shopify theme store is a
- 23:57little bit easier because again, the brand the guidelines meet
- 24:00their, the other side of that is that, you know, another change.
- 24:03We've seen recently is the the actual product merchandising
- 24:06itself. If it's not set up and
- 24:07configured correctly, trying to find those recommended products,
- 24:10it will Ties back to that unsexy data there underneath.
- 24:14If that's not set up correctly, it can bring out the wrong
- 24:17results or, you know, you've then got to customize it more
- 24:19and more heavily to get it to where you need it to be.
- 24:25And it's all times got the ability funds to create a path
- 24:28submission data into the customer property, klaviyo
- 24:32loved. Some people might know what that
- 24:34means. Other people might be new to
- 24:36think about. What does that mean?
- 24:37What's the benefit? Is this something you'll do?
- 24:39And if so, what is the benefit to the business of having that
- 24:42data passed through how can you use it?
- 24:44So the answer is, yes, we are doing it and this is the key
- 24:47thing for us. So I was looking at a site
- 24:52recently that had a an honest Got a Christmas gift quiz and
- 24:57they haven't asked for an email address to purely because
- 24:59they're looking at it from a website, conversion perspective.
- 25:01We want not to put a barrier in place for a customer to go
- 25:04through the quiz. But suddenly that cus that
- 25:07individual browsing. The site is given the fact that
- 25:09they're buying for their wife or, you know, they've got X to
- 25:13spend there. They've got a preference on, I
- 25:15know, sent or whatever it may be.
- 25:17All that data is are they given that data, but it's only
- 25:19available within that, that store or that session at the
- 25:23time. What we can do with the klaviyo.
- 25:27Sorry. Claudia octane quizzes and
- 25:29integration is start say well, okay, Phil's gone through this
- 25:32site, he's saying he's buying for a gift.
- 25:34It's for his partner and he wants to spend ex and his
- 25:39partner has a preference towards, you know, floral smells
- 25:43all that information is then stored in clay videos, custom
- 25:46properties. So, the lovely thing about Craig
- 25:49is how quick it is to action and get going, but it is a little
- 25:53bit Limited. In terms of the data sometimes
- 25:55because you've got a customer profile, email address.
- 25:59First name last name, you've then got the transactional data
- 26:02that's stored against it, and you can do elements of that
- 26:04date. If you then want to start
- 26:07collecting that preference data or that personalization data,
- 26:10you then need to kind of build that through.
- 26:12So, you can use clothes and klaviyo to start to store
- 26:15information about the fact that I know, Dave has previously
- 26:18purchased, you know, immense white T-shirt.
- 26:21But you've got to manage to do that where the octane klaviyo
- 26:25integration comes in as if they've gone on to a quiz and
- 26:28he's put in that he's interested in men's t-shirts, color, white,
- 26:33all that profile information is then stored directly into clavia
- 26:35as a custom property, so it's generating that on the Fly and
- 26:39allowing us to then use that data to try to trigger
- 26:42personalized comes with in emails or natively, new emails.
- 26:46So, the fact is we can use that to power personalization in
- 26:49campaigns moving forward. So like Earl Smith have
- 26:55obviously got a lot of growth through CRM and their focus on
- 26:57area, apart from the quiz that you've worked with them on which
- 27:01a lot of success. What other things have you been
- 27:04doing to help them achieve growth?
- 27:07Okay, so we've got a long-standing relationship with
- 27:11cal Smith. So when I first got introduced
- 27:13them back in back in 229 to they were on MailChimp at the time I
- 27:18think pool remember this, you know, Shopify and MailChimp when
- 27:21talking to You had to go through a third party platform to do
- 27:24anything involving data. We what we have done as a
- 27:29business working with them, is migrating to klaviyo.
- 27:32We set up a number of the key functions within that process
- 27:36around the customer experience of the customer Journey.
- 27:39So welcome emails and people sign up and people get different
- 27:42messaging based on where they signed up or how they signed up.
- 27:45So the pop-up email filter, the quiz and the checkout they all
- 27:50get slightly different messaging based on On that activity.
- 27:53You've then got your retargeting messages so you know your top of
- 27:56funnel, so collection, abandoned browser, abandonment, add to
- 27:59cart basket, abandonment. So, all of those people come
- 28:02through the funnel, but there's a lot of work on the data as
- 28:06well, so behind the scenes. So they've got a large database,
- 28:08not all of it is engage with email at any one time.
- 28:11So thinking about that email frequency management of well,
- 28:14actually this this this cohort people are more engaged with
- 28:17email so actually the frequency may be slightly hard and the
- 28:20cohort that I'm in last Black Friday and haven't really opened
- 28:23an email since. So we're starting to think about
- 28:26the data to it, to enable us to manage deliverability, customer
- 28:29experience and ensure that we're not that we don't want Our
- 28:33Brands to be three times a week discount, heavy to everybody all
- 28:36the time. It's about managing that each
- 28:39individual cohort correctly with the right frequency and the
- 28:43right messaging. Last two questions.
- 28:51So my last part of my second to last question.
- 28:55What are the best? What are your favorite?
- 28:57Implementations of quizzes? Have been e-comm sites and
- 29:00octane generally. Oh that's a tough one.
- 29:04Dave, you got any any any thoughts from the one you've
- 29:07just talked about for ha. Ha ha.
- 29:08Yeah, we've done one for a midwife company and this is a
- 29:17good one because it's basically Four separate quizzes within one
- 29:20but it allows us to live a scale out those four quizzes on two
- 29:22different stages or both of that guy pregnancy Journey.
- 29:26So I quite a quite like that one because that was a lot of
- 29:28legwork in terms of working out the logic of wait, those bits
- 29:32octane of got a new AI tool which looks really cool.
- 29:34I've never managed to be able to use it yet and so that looks
- 29:37like a really interesting one to come in to come into play.
- 29:43Great. I think yeah to that.
- 29:45I really like someone is may even if you send that line, M,
- 29:49AE V, and that's a pet food quiz.
- 29:51But it like personalizes as you go, like, change your site, some
- 29:54of the content and imagery and stuff.
- 29:56And then I think both really good from InDesign and kind of
- 30:00engineer perspective. And if you could change the curl
- 30:05Smith implementation, like maybe beyond the capabilities of
- 30:08octane or yet maybe beyond the limitations.
- 30:13Have the client. Like what would you change?
- 30:15Like, what do you think? Could be improved on the
- 30:17implementation? I think it's there's always room
- 30:24for improvement, I think some of the data integration, you know,
- 30:28we'd like to get more data into into klaviyo.
- 30:32You will know this pull from your time looking at it as well.
- 30:35It's very, very good and it is similar to way to claim you in
- 30:39that actually, it does 90% of everything you needed to do.
- 30:43I, you can just get in, get going and you can go from sort
- 30:46of Zero to Hero very easily with the right sort of structure and
- 30:50process. It's when you want to get into
- 30:52that, 10% of that customization, I think like, with all
- 30:55platforms, it's done to a level where it does just about what
- 30:59you need. But when those temps an edge
- 31:02cases that you really want to push the boundaries, that's
- 31:05where I think the platform. Yeah.
- 31:07Starts to not starts to fail, but starts to you start to hit
- 31:11the buffers and what's possible. And that's where you may stop.
- 31:13Think we'll actually do. I go for a more custom setup
- 31:16with you know, get my developers to build something.
- 31:19Now, we have this conversation. Often with our customers of well
- 31:23I can get our developers to build something versus paying a
- 31:26monthly fee. Nine times out of ten, our view
- 31:30is actually using a tool that can do it front end.
- 31:34A marketeer can manage themselves.
- 31:36Is generally better than getting the developers to build it as a
- 31:40custom activity, purely from an ongoing management perspective.
- 31:44IE, you've done a development project to build a quiz but you
- 31:47then want to change the third question to something else.
- 31:50Well, actually, there, You've got to rebook it in, with the
- 31:52developers to get that question, change versus the front end of
- 31:55Octavian. Have to go in there and actually
- 31:57do it yourself. And that's where I think.
- 31:59Sometimes it's understanding what the opportunity in the
- 32:02scope of the project is up front.
- 32:03So is octane the right fit and it will be things like that.
- 32:07The other side is, you know, the zero party cookies, that you're
- 32:09then capturing as well, which is really helpful.
- 32:11So, I think for me, Octavius. The design be able to upload
- 32:17more custom CSS, is being able to maybe ideally with their AI
- 32:21to almost develop the question. You know, the cut some basic
- 32:24questions on its own. Once it's at your data and
- 32:26generate that, that could be the future.
- 32:28But I think generally, you know, we can talk about the bells and
- 32:32whistles but 90% of it is there and Brands can get on and use it
- 32:36quite quickly now. Yeah, that makes sense.
- 32:40I think yeah, we've seen a massive uplift.
- 32:43People just been able to roll out and I think before attained
- 32:47came on the market or when it was still a chatbot back in the
- 32:51day. I remember all the different
- 32:53tools that did this took months to roll out and cost, you know,
- 32:56thousands and thousands of pounds a month.
- 32:58So yeah, I think it's really good.
- 32:59That it's like a bit of a commodity now.
- 33:01Initially by. Yeah, I mean you could almost
- 33:04look at it in a way of you know what bit?
- 33:06I've been in the email space for well, over a decade now.
- 33:09And used to go into Brands and talk to him about putting the
- 33:13platform in place with basket banner and developers go.
- 33:17Well we can build a basket Banner we can see that but again
- 33:20it comes down to then it's an IT lead process versus a marketing
- 33:23lead process and then someone's got to manage that process and I
- 33:27think sometimes that convenience and speed to Market and be able
- 33:30to almost test quickly and fail quickly is more important than
- 33:34saving a couple hundred dollars a month in fees by building it
- 33:37yourself. I think it's also, if you're
- 33:41going to develop your own one, you always want to prove a
- 33:43concept that the quiz is going to work.
- 33:46So this op octane, Bill gives you that proof of concept before
- 33:49you even spend money in building the developers one because do
- 33:52better this time, obviously cost more than being able to do it.
- 33:55So if you are, if you get to the point of limitation of it, at
- 33:58least, you've got a proof of concept of it works.
- 34:03As peanut I found that really interesting.
- 34:05Thanks gents. Yeah, I really like the idea of
- 34:09prison. I've seen some good
- 34:09implementation. I think you'll your the key
- 34:12takeaways from this is just thinking through that user
- 34:15journey and thinking through what data do you want to clip?
- 34:19Also, what the value to the customers?
- 34:20Because I think that's the key, isn't it?
- 34:21If there's no perceived value to the customer, that's when you
- 34:24get the high band and that's because they give up.
- 34:25They can't see why this they're carrying this on.
- 34:28Yeah. And this is where it comes back
- 34:30to that personalization thing as a A whole piece around if you're
- 34:35you know, we talked about earlier to.
- 34:37So if you're a larger organization with a data Team
- 34:40without a good-sized CRM team with a designer and you've got
- 34:43the bandwidth to be able to create the variance, you need to
- 34:46drive, you know, bespoke personalization on a regular
- 34:49basis great. If you're, you know, starting
- 34:53out you're a little bit further into that Journey, but you're
- 34:55still a small team, creating those variants sometimes, just
- 34:59doesn't add the value. So yes, it would be nice to do,
- 35:01but there's Bleep. Better ways to spend your time
- 35:04and money on. And this is where something like
- 35:07an octane can help because suddenly you go.
- 35:09Well, actually from an automation perspective, we can
- 35:11have Pre-K and Designs but just plug in play.
- 35:14Based on those parameters, we can use elements of those quiz
- 35:17data to start to drive more relevancy and columns, or
- 35:20frequency, and columns. And it's just, I mean, I think
- 35:23the takeaway put the customer at the heart of everything you're
- 35:26doing is a brand. So, if you're asking the
- 35:28question, why, you know, what you're going to use it for, and
- 35:32how You deliver that honor, you know, commercially viable basis.
- 35:36Yep. Sensible advice and if people
- 35:39are listening. Yeah, people might be thinking,
- 35:41I don't, I don't believe I can make this course work for my
- 35:43brand and my products or they might be thinking, well,
- 35:45actually I've got a few technical questions about how
- 35:47to, how I can do this properly and whatever I need, how do they
- 35:50reach out who they connect with? So either reach out to Dave or I
- 35:54directly at Passover website is www.hs e dot k.
- 36:01Dot U, k-- email address is a day that or fill out and fill
- 36:04with two L's because I'm A bit awkward in that regard or
- 36:08simply. Hello it passed.
- 36:09Someone of the team will be able to connect and help fantastic.
- 36:13And thanks as always for everyone for listening, we hope
- 36:15you enjoyed it. Keep your ears open for a next
- 36:18episode. We drop them every Tuesday, do
- 36:20let us know if there's any specific topics or Technologies.
- 36:22You like us to feature, do subscribe if you haven't
- 36:25already. And we'd love that rating on
- 36:26Apple Spotify or YouTube until next week, everybody.
- 36:29Thank you. More information on this topic,
- 36:38head over to re-platform dot. F m, for our audio podcasts to
- 36:42discuss a project, or if you like to chat about any of the
- 36:44topics covered in this episode in more detail.
- 36:46Please reach out to myself. James curd and my co-host, pull
- 36:49Rogers, via LinkedIn, and Twitter.
- 36:50Thanks again for listening and keep your ears peeled for the
- 36:53next episode. More information on this topic,
- 36:38head over to re-platform dot. F m, for our audio podcasts to
- 36:42discuss a project, or if you like to chat about any of the
- 36:44topics covered in this episode in more detail.
- 36:46Please reach out to myself. James curd and my co-host, pull
- 36:49Rogers, via LinkedIn, and Twitter.
- 36:50Thanks again for listening and keep your ears peeled for the
- 36:53next episode.