Episode 214

Split-Test Everything with Kurt Elster

Kurt Elster - Ethercycle
November 9, 2022
SUBSCRIBE: iTunesStitcher

Kurt is a legend in the eComm space and hosts one of the most popular podcasts in our industry.

His show, The Unofficial Shopify Podcast, just crossed the 2 million downloads milestone.

Kurt and his company are all about one thing - helping Shopify store owners make more money. And the best way to do that is by split-testing everything. 

An important first step of split testing is questioning your assumptions. (We all know what happens when we assume.)

Here are a few recent tests Kurt and company have run. We discuss the results on the show...yes, that's a teaser. Listen to find out what test won! 

  • Does free shipping always result in better performance (and is it even worth testing)?
  • Should I put a slider on my home page? 
  • What about hero images and banners on my category pages?
  • Do add-to-cart buttons on your collections pages help or hurt conversions?
  • What about stuffy old early 2000s breadcrumbs - are they helpful or not?
  • Plus, favorite tools, tips, and tricks that Kurt uses often.

Mentioned In This Episode:

Kurt Elster
- LinkedIn
- Ethercycle
- The Unofficial Shopify Podcast


Jay Leno’s Garage
ShipScout
Intelligence
Google Optimize
Hotjar
Overtone

Transcript:

Brett:

Well, hello and welcome to another edition of the E-Commerce Evolution podcast. I'm your host, Brett Curry, CEO O M OMG Commerce, and today I have a longtime friend, a legend, someone that you have undoubtedly heard of. You probably follow this guy on the regular. We were just researching prior to hitting record, we were wondering how long have we actually known each other? And turns out we first met in October of 2016. That's forever ago in internet years for sure. So my guest today is the man himself, Kurt Elster, co-founder of Ether Cycle, and the host of the unofficial Shopify podcast, One of the best podcast on there. Oh, I heard that, I heard that. That was the your, what was that sound effect, Kurt. Let's dial it up one more time. Go ahead.

Kurt:

The air horn.

Brett:

Yeah, man. So I think you, in fact, I'm confident you were the only guest I've ever had on this show that brings their own sound effects. And that's just one of the benefits of being a podcast host yourself.

Kurt:

Yeah, it's a dubious honor. But you got, I have fun with it

Brett:

<laugh>. That is awesome. And so we have a really close mutual friend, Ezra Firestone. And actually we, you and I kind of were able to sync up again, I guess back in when was that? We were in Miami together. We both spoke at Ezra's event in Miami. And I heard your presentation. I was like, Dude, you gotta come back on the podcast. It's been too long. But you've got the best clip, the best sound effect from Ezra Firestone that I hear in a number of your episodes. But dial that up if you would, and then explain to folks what that means.

Kurt:

Tech nasty. That's Ezra <laugh>. Like

Brett:

Te nasty.

Kurt:

Two years ago he would call me Tech nasty. And then two years ago outta nowhere, he writes a freestyle Rap records it. That's what he does, sends it to me, in which he refers to me as tech nasty. And so I was like, well that's great. And I play it on my show. But the one time, but the tech nasty sound bite I took. And now every time I introduce myself, I use that like a sound signature <laugh>. And it, it's a fun thing, was like, I'm your host Kurt Elster Tech Na. And then I just moved past it. I don't address it.

Brett:

People are like, That was cool. I don't know what that was, but that was cool. And I embrace it

Kurt:

By the eighth time. They're like, Yeah, that's tech nasty

Brett:

Te nasty, nasty. So Ezra has that skill set. He will freestyle rap for you. He did that on a podcast episode with me one time. I need to go back and listen to that. But I would love to hear the te nasty freestyle rap sounds, I think sounds amazing.

Kurt:

Skill one, can you freestyle rap? That's hard enough on its own, I'm

Brett:

Sure. Yeah. Can you even do it? You

Kurt:

Could really practice it like fake it once. Step two. Do you have the insane level of confidence required to just start freestyle wrapping as a 40 year old white man? Just

Brett:

Do it. Just do it As you're recording, just lay it out there and lay it out there with some, some bravado. And he does that, which is awesome. So hey man, I'm super excited about this. So we're gonna talk about split testing everything. So assume nothing split test everything. So how do we improve our performance? How do we make more money with our Shopify podcast? But I'm sorry, with our Shopify websites, Shopify podcast, you run one,

Kurt:

You just hit

Brett:

<laugh>. You guys just hit a milestone not too long ago, didn't you? With the shop unofficial Shopify podcast?

Kurt:

We did. We just hit 2 million downloads.

Brett:

2 million downloads. I want to be as good as you one day, Kurt Ster. When I grow up, I would like to be a podcast host. All

Kurt:

Right, here's the trick. You could brute force it by just never stopping. I'm, I'm never stopping. I'm over 400 episodes every week. We never miss one

Brett:

400 episodes, one episode a week. You never miss it. That's awesome. I've been doing this now since 2017, this show. I did another show prior to that, but I haven't been as consistent. I went through a period of time where I was like once or twice a month, you know, kinda lose momentum. So consistency,

Kurt:

The trick is key. Get yourself some sponsors now it's a client deliverable. Yes. And if you miss it, a you gotta send that refund. That hurts.

Brett:

That does hurt. Yeah. Make it like a client where it's like, I can't let a client down. I can let myself down or Right. I can push it to the side if it's just for us, but yeah, for a client. Gotta make it happen. So a couple things. You guys work with some really impressive brands, including gans and a few other automotive brands. And maybe my favorite, most interesting in my point of view, Jay Leno's garage, right? Which is cool. So I want you to talk about that. But before we do, I would like to talk about Kurt Ester's garage because oh, you and I are friends on Facebook, we get to hang out and chat a little bit, but you've got a bit of an eclectic collection of cars. Maybe not to the size or scope of Jay Leno. A few people do, but you's got some cool rides, man. So talk about what is in Kurt ER's garage.

Kurt:

All right, so the daily driver, the crazy one, I got a 2017 model S 100 D, that was the quick one, The big sedan. But it was boring silver. I don't want silver. So

Brett:

This is a Tesla. Those that may not not know the Yeah, it's Tesla. Yeah,

Kurt:

Yeah, yeah. The

Brett:

So zero to 60 in what? What's your zero to 60 time

Kurt:

3.9 on this one.

Brett:

Nice.

Kurt:

Yeah, it's quick and it just goes, it makes jet's noises goes and you're just do it with 60. That's crazy. But we vinyl wrapped it this satin yellow gold color. It's the same as this microphone if you've got the video version, it's like that's the vinyl we used and that is the, I put wheels on it in tense. And that's the coolest thing. I love that car.

Brett:

It's a beauty man. I've seen you got some stylized photos of that thing. It is a beauty. Oh,

Kurt:

I got a client's product photographer to come shoot photos of it. It was really cool. <laugh> awesome. But my wife's the cool one. She has the really cool cars. She always wanted a Volkswagen Beetle. I said I could figure that out. So she bought a 79 Super Beetle the last year they made in the US and it had been repaid as good shape, but it lots of deferred maintenance. So I got that back on the road running strong. I love that car. It's got 60 horsepower, slowest thing in the world. And that is pure driving experience. It's

Brett:

Such a good juxtaposition, right? So it's such the opposite of the Tesla. Yes, it's it's so fun that you go digital then you go analog. Yeah,

Kurt:

They're like, we gave you just enough what you need to not die <laugh>. And then the one that the car she daily drives, she also, she found negotiated, bought and had shipped here and then I discovered it had a ton of problems and we had to pull the motor and 84 Ford Bronco.

Brett:

Dude, that's my favorite. The 80 for all

Kurt:

Original. It's just that's my favorite too. Yeah,

Brett:

Yeah.

Kurt:

Oh we love the thing. Yeah, it's like 10 miles per gallon

Brett:

<laugh>. Which isn't the best right now, but hey, come on again. You balance it with the Tesla ex. Yeah. But yeah, it's so fun. You and I live online, we're always doing digital stuff. We're thinking AI and machine learning and algorithms and all this stuff. And sometimes it's fun to just have something that has no internet connectivity. It's still just that's you smell the exhaust and you hear the sound of the engine as you rev up. But it's a beautiful

Kurt:

Thing. It's like driving a lazyboy recliner. It's great. <laugh>

Brett:

Just comfortable. So that's awesome. Well hey for some of you that was a blast. For others you're like up with, I don't care about what's in your garage. So we get, it will dive into some content. So like I said, we were both speaking at this event in Miami, heard your chat thought, okay, this is brilliant, let's talk about it. So we're gonna dive into split testing everything. And first of all, I wanna ask you, is that a real thing? Should we be split testing everything or what should our mindset or approach be as we look at split testing?

Kurt:

Okay, so the reality and practicality of split testing every single individual element on a website, <laugh> just not realistic. You can't do it. But the phrase split test everything it really mean, you gotta question your assumptions because there's so much info out there where someone just goes so just posts and goes, this is the way to do it. This is the best I have declared it and here's like a screenshot of the 7 million. I made it 25 seconds this morning and it's all bs. You don't know what you're looking at. But that's how a lot of the common sense wisdom in e-commerce, cuz it's such a young industry comes to be. It's just like whoever was the loudest voice with the best presentation in their tweet becomes our best practices. And I, I'm guilty of doing stuff like that too. I think we all are.

Once you have an audience online. And so 2020 I went, I'm gonna really figure out split testing and really question this stuff and figure out, put my money where my mouth is. It gives some legit advice and it took some efforts of playing and I ended up two years later really mastering Google Optimize. I feel very comfortable with it. And so I've amazing the other. So number one, mastery Google optimize to the point where I could figure out, all right, this is a realistic split test to run. We could figure this out and we don't have to guess, well will this perform better? Does X perform better than Y? I don't need to guess anymore. I could just go have Google tell me, just test it. And so that's really freeing. But what also blew my mind was the number of times I was wrong about what the winner would be.

Brett:

Yeah, cuz you've been doing this a long time. I've been doing this a long time. We talk about this with ad copy a lot, right? Pros is we do day in and day out, we're spending tens of millions a month, we know our stuff. But you still gotta test cuz sometimes your gut, we trust it a lot. Sometimes your gut is just wrong. You need to lean on data. That's the better way to go for sure. So one quick note on tools to optimize with, and we'll dive into the first couple points, but you mentioned Google Optimize. What's the quick pitch or quick thought on why that tool versus some of the other optimization tools that are out there or split testing tools?

Kurt:

I really think Google Optimize has become the standard here because it's free and it's already plugged, it's halfway integrated. If you're using GTM Google Analytics, you already have most of your Google Optimize set up done.

Brett:

Great. Makes sense. Which really, essentially all the clients we work with, they've got Google Tag Manager, they're using Google Analytics anyway, so yeah, makes sense at that. That all just plugs it in. Okay, cool. So let's dive in. Let's, let's go through some interesting split tests, which I would say have some kind of surprising results, which I think will be a ton of fun for people to listen to. So first one, this is something you guys recently tested, I want you to talk about and the results you saw, but should price appear on the collection slash grid pages. So first of all, describe what that means. There's always difficult in a podcast setting, but what is that test? And then what did you guys find out?

Kurt:

All right, so you're shopping on a jewelry store and you go to, you're, I wanna shop product by type. I'm looking for tennis bracelets. I don't know who still wears a tennis bracelet these days.

Brett:

I think they're pretty popular.

Kurt:

Yeah, click through to tennis bracelet. You are now in Shopify, this would be a collection page and the rest of e-commerce universe, this is a category page on that page, which is a listing of products. Do you include price which will perform better with price or without? I don't know.

Brett:

Dude, that's a good question, right? Cuz you think, hey, you want people just clicking on stuff they feel comfortable buying that they feel like fits their budget or whatever. So yeah, maybe you think you want price to be visible on that collection page, but what say you, or more importantly, what does the data say?

Kurt:

And the answer is maybe you do, maybe you don't. It depends so much. So you have to test this yourself. But the first time we ran this test, it increased revenue procession. So this revenue procession, good way to combine conversion rate and average order value, increase revenue procession 23.6%. And this was in an apparel store with 97% confidence. So pretty good. This one, this test had a sample size of 10,000 or more really solid results. So we ran it again and we did it on mobile and then we did it on desktop and we did it on new and we did it on returning customer, same result every time. Then we ran it at some other stores. Consistent result. Ran it a few more times and we found it really depends on the category of product. It is either going, this is one I just wouldn't blindly do.

You have to test this yourself, but I suspect what's going on here is when someone is on, this is my hypothesis, I'm on the product page, I see the item with no price, then my brain is going to decide, yeah, the category page, my brain is gonna decide a value for that price. Whether I think about it or not, value subjective, I click through, I'm on the product page, I see the actual price. Is that below or above what I thought it was gonna be? If they're consistently less than what people thought. So you had this really great presentation and your prices are a good value. I think it's going to, I think that's when conversion rate goes up. When it's the other way around, I think that's when it goes down. But in testing it, I think my guess is two thirds of stores fall into the former category where this is a win.

Brett:

Interesting. But I wonder, let's say there's a good value, let's say you, you've got this value in your mind or maybe you don't, but when you view that category page with the price, if it's a good price, if it's a good value, I think they would still have a benefit, right? You're still looking at it, you're thinking, oh that's great. Let me just let dig in a little bit more, right? So someone may have a mental win or get excited, even just seeing it on the collection page, but regardless, interesting, that's what you

Kurt:

Can test.

Brett:

Gotta test it. So worked in most cases, but not all. I like that. So here's a good one, and this is one that I wouldn't have thought to test necessarily, but it makes sense. How important is font size? So the size of the text on your pages, does it matter? Does it not matter? What did you find out and why? Why did you test this by the way?

Kurt:

So I'm a big believer in, I'm a big typography nerd. I like type fonts,

Brett:

Which I notice looking@ethercycle.com, whichever to go check it out just to see. It's a super fun well-designed website. You can tell you typo typography, beautiful and interesting fonts. Yeah,

Kurt:

That one's got this real retro look. But I thought easier to read will perform better, we'll convert better if a site's hard to read, I'm not gonna fight my way through it, I'm just gonna leave. I don't think there's really any dispute there. So I wanted to prove this with data. And so I set out and I started testing different sizes of font. I'm gonna split test my way to determine the exact right font size. Well it turns out you a test with a really subtle small change like that, it's so hard to get to statistical significance. Yeah. And so I couldn't, no matter which way I set it up, no matter how dramatic that change was, I could not get a statistically significant answer out of it. But I knew font size and was important. But really it turns out readability is what's important.

I was looking too granular. The real picture here is readability. So this is one ends up being a trick question you would have to do in different layouts and split test those. And so we didn't do it, but around the same time Bayard Institute came out with a usability study and they do like these, they call them large scale usability studies. And they figured out here is guidelines a starting point for ideal readability on fonts. And if you think about magazines and newspapers, they have those real narrow columns. That's what makes stuff easier to read is you wanna follow print type setting that we have at this point used for centuries ever since the Gutenberg Bible. Yes. It's a line length is 60 to, you want a line length of 60, 80 characters. And so if you go on your website and you put it full screen and the text just goes end to end, all right, we're not optimized here.

Brett:

Not, it's not inviting to it, it feels like it feels work. It feels really difficult to get to the meaning of what you're putting on the page.

Kurt:

I remember at school you used to double space your paper so that you could fill out more. Well

Brett:

It turns out you just promise.

Kurt:

Yeah, 1.5 is an ideal readable line height. And then you can also real mess with paragraph spacing, word spacing, letter spacing. They have those. I tried just putting them blindly into a few websites. It didn't look quite right. So I think it's a starting point. You have to tweak it from there. But if it's, it's really just that how many characters are on a line? You got two sentences, they're not gonna read it. You got five words, perfect. Easy to read. And so I like those narrow product descriptions. I think that performs well.

Brett:

Yeah, I love it. And just to illustrate this point, I don't know if you guys are this way as well, but just to share a little bit of my weirdness the friend of mine got me a leadership book, gave me this leadership book that he raved about and I believed him, I believed him. It was a good leadership book. But when you open up the pages, there are certain pages that are one paragraph. So the entire page is a paragraph and then wall to wall text, right? Is the least inviting page ever. And I'm like, I still just have, I've not read it for that reason, which is silly. There's probably some great content there and I'm not reading it because it's so uninviting to read. So

Kurt:

Yeah, no, so if you format it poorly, no matter how good the content is and you look at it and it's just wall of text, Yeah, it's a turnoff. People aren't gonna bother

Brett:

My back. Use

Kurt:

The inner. You put some line breaks in there, man.

Brett:

Exactly, exactly. Okay, so readability is greater than font size or readability is what really matters. Font size, you could kinda go one way or the other, but the goal,

Kurt:

Yeah, I would say just as a rule of thumb for body font, really 16 is probably the minimum. If it's a really chunky font, maybe 15. But I'd say for most sites it's like 16, 17, 18 pixel.

Brett:

Nice, nice. Okay. Good stuff, good. Easy win there and helps us focus in the right way. Let's talk about one that I know is a real hot button. This is something that, and honestly this may be something that people have a four gone conclusion on. They believe in one way very strongly but that's free shipping, right? Free shipping is a must, right?

Kurt:

Free shipping, oh my gosh. Free shipping is a must have. We've been told forever. Forever that you have to have three

Brett:

Years

Kurt:

That the number one cause of abandoned checkouts is unexpected shipping expense. My god, these monsters charging for shipping <laugh>.

Brett:

That's,

Kurt:

That's a good firm service.

Brett:

What is this? Early two thousands with this website charging for shipping.

Kurt:

Ugh, disgusting. So is this the case? It turns out split testing, free shipping not the easiest thing to do, but at least in the shop in the Shopify universe you could do it with dedicated apps. And there, I'm sure there's several, but the two we use successfully ship scout and intelligence will let you do split test shipping rates. And so we tried this at several stores with, they got different results. But I'll illustrate one here. The effects were always very similar. And so we were able to track checkout conversion rate when we offered no free shipping at all. Everybody just paid 20 free shipping at $25 and free shipping at $75. And what was interesting is the checkout conversion rate for all of those things were within a few points of each other. Our checkout conversion rate, meaning people who get started, the checkout for no free shipping was 72% for free shipping at $25. That was the highest. Or we got the lowest, but it was still under 80%. And then free shipping at $75 right in the middle, right at 70

Brett:

Five's.

Kurt:

Interesting.

Brett:

But 72 was with no free shipping at all.

Kurt:

So I'm really not gaining, I'm gaining very little, it's just marginal gains with these free shipping. And so I think we should reward our whale v i p clients. If your average order value is 50 bucks, then it should be a no brainer that someone who spends $200 should get free shipping. Fine. I'm not gonna dispute that one. But it's more as customer service really. And so I think it gets interesting when you start to consider profit per order, profit per checkout because especially since shipping keeps getting more and more expensive. Totally. And people are, so many people switch to e-commerce more than they ever had before in 2020 that I think as consumers get more sophisticated, they're also more willing to pay for shipping and be understanding about it depending on the item. And so in this particular test, we knew our average fulfillment cost was seven 50 in order and when we ran so we could calculate profit per visitor, our $25 rate was where we performed the best. That had a $12 profit per visitor. But when we didn't offer free shipping at all, which we only lost a little bit of conversion, that had a almost $20 profit per visit.

Crazy. I sacrificed a little bit of conversion, but then I gained considerable profit like 40%. And so that's why it really pays to take a hard look at what does shipping cost us and what's free shipping cost us in profit as a cost center.

Brett:

I love that. And I think, again, this just underscores the need to split test things because a lot of us just believe, hey, because Amazon's been free shipping forever, we have to be free shipping with our products on our stores. And the reality is you need to test it cuz maybe it's not boosting your conversion rates that much and maybe it's just eating in to profits. So you gotta test it. Love

Kurt:

That one. Well, and what I like too is when you're not offering free shipping just all the time as a threshold, then you can use that as a promo instead of having to do the traditional like, oh we discounted our product. Yeah, hey boom, we're offered free shipping this weekend. Only go.

Brett:

Yep, yep. Super easy. Then built in promotions and now it's legitimate. Right now it's real because you don't offer it the rest of the time you are offering now. So built in promos. Love it.

Kurt:

That urgency. Yep.

Brett:

Yep. That's awesome. Cool. So let's go back to collection pages. So you wanted to split test hero images. Do we put hero images on our collection slash category pages? What is that? And explain what is a hero image and then what did you find with that split test?

Kurt:

All right, so I'm shopping on funky t-shirts dot com, I click men's funky t-shirts, boom at the top there's like three dudes and robo cop and a banner image. By

Brett:

The way, do you have a tech nasty t-shirt? Cuz if not, you should probably have one at time.

Kurt:

Nasty. I don't but I gotta work on that. Yeah, <laugh>

Brett:

Funky.

Kurt:

So yeah, men's funky and there's like three dudes of Robocop T-shirts looking funky but it pushes all my products down the page. Uhoh like I've got this big banner at the top, I gotta wait for it to load <affirmative>, but it looks great. So I'm like pro hero image. Cause I think it makes a difference when you've got a really nice well done image.

Brett:

That's the image at the top of the category page kind of frames. What is this category or collection? All that's

Kurt:

That banner. And it's always a lifestyle image almost always. Or an action shot. So this is pretty cool, I love these, but it does this matter. Should we test it? So we did. And this was in a store where they look great, they had put in the effort and it cost us conversions. Having it there, it performed way better without it. Crazy revenue procession up 16%, 92% confidence. Well maybe that's just cuz on desktop it doesn't look right or on mobile it doesn't look right. So we tested that, we ran this one again, same store, mobile versus desktop new versus returning. Every single one of 'em performed better without the darn banner image. And my thinking is it's scrolled it. There's like, I got this banner at the top, but I'm already shopping so all you've done is presented me with a thing I have to load and scroll past just to get to what I wanted to do. If they're at that category page, they're already shopping, stop selling, They've already started shopping.

Brett:

I like that. And I would agree with you too, if you look at a good collection page, good category page, it's got that brilliant here, hero image. You're like, Yeah, I like this. You show me two pages. One with one without the hero image is gonna look better. But you make a very good point. If someone lands on the homepage, they get to the collection or category page, they're already in shopping moat, just show me the freaking products at that point and let me shop. Or if they land, the only time you really land on a collection or category page is if you are doing non-brand search or something like that where it's a very specific search that again is very product focused, otherwise they're gonna be landing on something else first and then getting to a collection page. So yes, this makes a lot of sense. And with a 16% increase in revenue procession yes please, I'll save that load time, pull that hero image and make more money there. So you would, and I love the way you framed this, that this is kind of a great way to look at both conversion rate and AOV at the same time. But talk about revenue procession and how you guys landed on that. Has that always been the key metric for you guys And no, Yeah, you guys land on that?

Kurt:

No, I mean, so as far as my e-commerce KPIs go, if I'm just yanking on levers to scale businesses, average order value is my favorite because you don't have to get new customers, which is the hardest thing. And you don't have to convert more customers, which is tough. You could just sell not even a ton more, just a little bit more a on average to every customer and you make significantly more money. I love AOV as this, but in conversion rate optimization, it's in the title, we would always look at conversion rate as our main kpi, but it's certainly, it's not the only one. It's not perfect if the further your test is away from that checkout where the conversion happens, the more noise that's getting introduced to you,

Brett:

What really raised the conversion right there if it's not there at that checkout. Yeah.

Kurt:

And certainly there is so much statistical noise and questionability in split tests and you just can't post a split test on the internet without a whole bunch of people coming outta the woodwork to be like, here's why you're wrong,

Brett:

Here's why you're

Kurt:

Like, okay, well wrong. We all know it's cuz this disagreed with what you assumed it would be, right

Brett:

<laugh>

Kurt:

But no. So in with another, this always turns into a pitch for Google Optimize. I swear I have no association, it's free.

Brett:

It's a free tool. People, yeah,

Kurt:

In Google Optimize, it'll ask you what's your primary objective and transactions is what they call conversions in. And you could do revenue, which now we're revenue procession and we're combining really conversion in average order value. I like that. But then a whole bunch of others. But you can choose multiple so you don't just have to commit to one. And so when you see consistent results across conversion and revenue procession, that's usually what I'm looking for. But you could also do add to cart, bounce rate page views. Sometimes we'll use those if it's more engagement focused. But no, you don't just have to use one. I'll usually set two or three and then that way I feel more confident when I see that results are fairly consistent across multiple KPIs.

Brett:

I love it, but I love that it's a good blend there because what's the benefit in increasing conversion rate if our AOV goes down, which can happen sometimes. That's where shifting that free shipping threshold can really have an impact. Yes, conversion rate went up but AOV went down. So in the end we made less money. But that's where that revenue procession really kind of helps combine those two and clarify things, which is great. So awesome. Let's move on to the next test. I got a couple more here. These are super fascinating. So should you include recently viewed products on the product detail page? So here we are, we're at the product detail page, we're looking at this funky. Should we also there show, hey remember you also looked at these other funky, is that a distraction or does that actually help with conversions? What did you find?

Kurt:

So alright, as the designer in my heart just wants to get rid of extraneous elements on a site, that's the easy way to conversion rate optimization is just try and keep people captive. Declutter by getting rid of distractions, declutter. And a recently viewed products widget on a cart. Who is this for? I have a history button, a back, I got history, a back button. I

Brett:

Not remember what I looked at five seconds ago.

Kurt:

I'm gonna get emails, browse abandonment, I'm gonna get marketing ads. And so this was the thing I wanna justify really just for a client, I wanna justify getting rid of a recently viewed items widget <laugh>.

And they're like, Yeah, you could do whatever you want as long as you show us the data first. All right, fine, smart challenge accepted. And so I tested this thing and it was net improvement. Let's see. Yeah, it was rev. Everything increased when it was present versus when it was removed. And this annoyed me, I'm like, now I'm getting rid of this thing. So I run the test again, mobile versus desktop, same result. Darn it. Run the test again. New versus returning. This is where it got interesting for returning visitors. It's increasing good version rate by 33%. And we had this thing on the cart page and the product page new visitors, it must have just been weird for them to see it decreased conversion rate by 9%

Brett:

<laugh>. So decreased conversion rates for

Kurt:

Them. It was for,

Brett:

Yeah. Huh.

Kurt:

So in Google Optimize you could deploy, it's called deploy personalization. So when I ran the test that said, hey, hide this for new people and that's a win. I could just have Google optimize do that. So you could use it to run personalization. So if it sees a new visitor, don't it hides that recently viewed item widget and if it returning visitor then it doesn't do anything.

Brett:

But returning visitor increased conversion rate by 33%. That's crazy. But I would tend to agree with you, or at least I could make the case there that yeah, it's just a distraction, why have it there? But your client was wise and said nope, show me with the data and the data said otherwise. So you have the recently viewed products, you keep it on the product detail page at this point. So brilliant test. Love that. What about, now this one's super interesting to me. Add to cart from the collection page, yay or nay. So I'm on that collection page or that category page, there's that quick add button or add to cart right there. Or do you want to view it or do you wanna do quick view or whatever? So what happened there? Do we put the add to cart on the collection page or not?

Kurt:

It depends, but I think the answer is yes.

Brett:

<affirmative>. Okay.

Kurt:

And so we ran this test again, I tried to do, at least for pulling these examples with data, I tried to do apparel stores for everything cuz it's very relatable, it's very general and that's also one of the biggest categories in e-commerce. But in this case, adding it increased revenue per visitor 15% with 85% confidence. So that's right at the minimum for where we're still statistically feeling this is statistically significant. But having the add to cart on a collection or category page aspe is really convenient. Especially if you're on a slower connection, you're making multiple purchases and items. And so I think it's definitely gonna depend on the store's catalog and what the products are. If I'm buying body jewelry, if I'm buying nuts and bolts, if I was just buying buttons for an arcade machine and I was like, I swear I wish this site would let me add to cart cuz I need two 16 buttons in different colors and I have to go back and forth, back and forth. And right away I knew that's the use case. And so if you have these less spec driven, lower price, similar items add to cart for sure. For sure. If it's a MacBook, real high price technical items, I don't need to add five

Brett:

Different, We're always gonna go to the product detail page before you purchase. Right?

Kurt:

Yeah. So certainly some common sense needs to be used here. And Bay Mar Institute, they did a usability study recently. They felt they made a similar argument but they liked quick view where it opens a window with more info. Right, right. That keeps you on the collection page. Yeah, I don't wanna have to go back and forth, which on a phone could be annoying for sure.

Brett:

Yeah. So then you just look at maybe turning that off for the mobile experience and keeping it running for desktop or is that an option?

Kurt:

I think on mobile you do. Oh for sure. I think increasingly we need these mobile specific and desktop specific optimizations. I think they help a lot. But no, I really think it's more product and category dependent. If it 50 bucks and less, definitely consider this. And if it's a lot of similar stuff, definitely consider it.

Brett:

Totally makes sense. But if it's a high ticket item and you're gonna be digging into that product detail page every time we add to cart, then it's less important. Yeah,

Kurt:

You could also do it if you wanna get fancy, do it per category or collection. Yeah, let's say

Brett:

Other, Yeah,

Kurt:

I'm selling drones. All right. I probably don't on the drones page, I don't want it. But then on the accessories I'm buying SD cards

Brett:

And batteries parts, the replacement parts

Kurt:

Page or whatever, then I could use it.

Brett:

Yep. Yeah, that's interesting. So maybe go collection by collection or category by category. Super interesting.

Kurt:

See what

Brett:

The price split test to me. I like it. So last one that we're gonna last split test result. We're gonna dive into them. We got some general tips which are super helpful as we wrap up. So what about breadcrumbs? Do we need the breadcrumbs? And in the early days, in my early e-commerce days, we supported a Magento agency and so we ran all the marketing for Magento agency and all their clients and super fun. But I know breadcrumbs were a big part of Magento, at least back in the day. But so first of all though, yeah,

Kurt:

Magento went hard on really making these gigantic faceted breadcrumbs.

Brett:

They loved the breadcrumbs. So for those that maybe don't know, explain what breadcrumbs are and then what'd you find with the test results?

Kurt:

So a breadcrumb, it's always in, there's a consistency to it. It's always like 14 point font in the upper left and it tells you starting from the homepage where you're at on the site. So really it's like it's a drill down, it's home department product or subcategory then product. And this just seems like a thing in the way I'm on the site, I I'm on the product page, I can just click the back button. So I thought, I'm like, this is just, what am I doing with this? I don't want this thing. And so again, I'm like it's extraneous, I'm gonna remove it. And so I gotta split test it. And it turned out on that when I was wildly wrong on the product page increased revenue per visitor by 53% with 92% confidence. That's insane. Cause what's going

Brett:

On? People love their freaking breadcrumbs. But yeah, why?

Kurt:

This is another one where when you get rid of the breadcrumbs, if I had just by straight conversion, it improves conversions. But at what cost? Because it's like they get to the product page, they go cool, add to cart, and then they leave. Whereas when I had the breadcrumbs, and you could see this in heat maps and screen recordings, they're using it as navigation. So they go back and they shop more. And so when I got rid of the breadcrumbs, all I did was sacrifice people buying multiple items. Nice. Cause I made it a pain.

Brett:

It's crazy. So people were adding to cart and or purchasing. So conversion rate went up. And again, that's why you gotta look at it holistically. Conversion rate isn't the only metric, right? People are buying less, they're spending less. So conversion rate only in this case actually hurts. So yeah, they're adding that item to the cart, but then they've got the breadcrumbs there where they go back to the category or two categories go whatever they can, they now easily continue to shop. So

Kurt:

That's sell more than one item. You probably have to have the bread crumbs. There's like only there'd be a rare handful of scenarios where you didn't want the bread crumbs. And so now I'm like, all right, we gotta make breadcrumbs work, look good, be even better. Gotta optimize my breadcrumbs.

Brett:

Now gotta, So we wanted to get rid of 'em stupid breadcrumbs but they make a difference now we gotta make 'em smarter better. So which by the way, how do you optimize breadcrumbs? Do you have any thoughts or theories there?

Kurt:

So in shop by the breadcrumbs are a little limited. They normally look at like, hey, what's the collection url? They look at the URL to fix it. And so we rewrote it, what it act one where it can use, if it doesn't know how the person, if they just landed on the product page, how do you develop the breadcrumb? And so adding some logic to backfill that in either you can have it check what was the last collection they look at with a cookie and add that in. And then if that's not present, okay, let's look at the products, maybe the vendor or the product type and then use that to try and backfill collection. This. If you, you're really getting deep in the weeds by the time you're doing this one, you're

Brett:

Getting deep. This essentially, I mean this happens. So we run a lot of Google shopping. Traffic's one of our core channels, but a lot of people then parachute in on that product detail page. So if I land,

Kurt:

So in that use case, yeah, this is a win

Brett:

For sure. Having those breadcrumbs there, they're gonna buy more items for sure. Okay, love it. So we have it. There were seven unique, interesting split tests, several of which surprised you and would've surprised me if you hadn't already told me the results. So super interesting there. What about, let's close out, which is some general tips. What tips would you give people as they're going on this journey to increase conversions and AOV and all that good stuff?

Kurt:

So for sure there's some good advice here. I tried to keep it to things that will be generally helpful, but you gotta question everything. And that's like when you see people sharing their wins and the results and their split tests, this is what I did. Okay, that's what they did not, that doesn't mean it's what you should do. And

Brett:

You know how they did it. What if they didn't even get a statistically significant result? What if they just got the result they wanted and they wanna share it?

Kurt:

And even if they did get it statistically significant, it was significant for their audience, their catalog offer, et cetera. It's not the same for you. And so what people are like, oh, they'll like, I'll post something and they'll start questioning the, not often, but they'll question the methodology cuz they're trying to get to, well here's why this is invalid. And to those I'd say like, look, I'm sorry that it disagreed with what you thought, right? Cause that's really what they're saying. Or

Brett:

You don't like the truth, it's okay. Yeah,

Kurt:

Yeah. But also it doesn't matter for you just go, you run that same test on your store and see what happens. And so I think question everything about other so-called experts, there's just being in this space, there's a fire hose of great info. So many. So who do you listen to?

Brett:

So many experts. Yes.

Kurt:

And certainly part of the problem here. And then also question your own assumptions. And so there's all kinds of elements on your homepage, on your website. Why are they there? What are they doing? Until you've split system, you really have no idea if they help or hurt. You're

Brett:

Just, I think I heard you and Paul talking about this on a podcast episode where we all, were in love with the sliders on the homepage, right? We all want sliders at top our homepage. Yeah, because that's what you do. That's what you've done since 2000 or the late nineties or whatever. But what, what have you guys found there? And I know we kind of finished the split test, but you found some interesting stuff there, I believe recently.

Kurt:

So that collection, homepage test or that collection test where it's like, does the hero help her hurt? Yeah, we did the same darn thing on the homepages and same result, it always performed better without the homepage hero image. And the reason real is what's going on is because the second thing is always featured promos and a product grid, a featured collection on a site. And so you're just getting them shopping faster. Again, it's the same premise. That one really blows people away. They're like, but we have to have them. I dunno, have you been to google.com lately? Where's their slider? Right? You don't have to have a slider.

Brett:

Yeah. Yeah. That's awesome. So other tools you had mentioned, so you mentioned Google Optimize, love it. What other tools would you suggest people check out?

Kurt:

Well, I really like Hot Jar you know, also wanna combine this with a heat map tool screen recording tool. So you can see, get a macro view of how people are using the site and micro views. But then I think ultimately the real deal, number one conversion rate optimization hack is talking to your customers. So few people wanna talk to their customers and that's where you get the really great test ideas is when you talk to people and you go, Oh, you know that because every customer will be like, well they'll speak for everyone. I want this and that and that. Well maybe they're right. And so when you go test those things, that's what we've had the biggest wins is taking a customer suggested idea from a phone interview and put that onto a site and then test it.

Brett:

Super interesting. So you actually, let me get this straight. You're actually calling customers on the phone, talking to them with words, not typing or something like that, but you're getting feedback from the customer and then you're implementing those into split

Kurt:

Tests in rare occurrence. Yes. Yeah, <laugh>, that is the most powerful thing you can do is pick up the phone. No one wants to use phone.app, but that's the one that prints money.

Brett:

But it works, man. It absolutely works that. That's awesome. Any final tips? And then I wanna point people to you so they can check out more of your brilliance as they go. But any final tips? I know we could talk about this for hours and

Kurt:

Hours. I mentioned a few tools. Use the tools you, I don't care what tools you use, so long as you're doing the work, whatever you're comfortable with, that's great. Perfect. You don't have to justify it to anybody. We mentioned it earlier, but I think it's a mistake to just solely and laser in on conversion rate, especially Google Optimize, where you can have secondary goals and I think revenue is a really good one to work with. And then certainly I think the other mistake is lumping everyone together. And so I would do consider segmentation, especially mobile versus desktop and new versus returning

Brett:

Mobile versus desktop to very different experiences and different modes, different mindsets people have. And then, yeah, it's another simple segmentation, new versus returning. Cuz like you showed with several of your tests actually that new visitors behave differently than return visitors. Which makes a lot of sense actually if you think about it. So Kurt, this has been awesome. So go listen, check out the unofficial Shopify podcast, one of my all time favorites. You gotta listen. But if you listen to this and they're like, Dude, I wanna hire Kurt and his team to work for me. Tell us a little bit about Ethercycle. Who do you work with and how can someone reach out to you?

Kurt:

Ethercycle.com, you could find us. And we work exclusively with Shopify store owners. We've been since 2014, I think. And that's awesome. We do really, we're a theme shop. We build a lot of custom themes. We do do store migrations but we also do maintenance programs and conversion rate optimization work.

Brett:

Yeah, and you guys did, did all the work recently for Overtone, which is as a Firestones company. We worked on that together. So you guys did a lot of the build side. We're doing the Google and YouTube and Amazon side. But yeah, what what'd you do for overtone specifically?

Kurt:

Overtone? That one was interesting. They were on Shopify, but it was headless. And so we built an entirely custom theme and that designed to develop to custom theme and then migrated back to Shopify. Got rid of the headless solution entirely, which is simplified it and admin for them made life a lot easier. And some of these split tests

Brett:

We mentioned too, didn't it?

Kurt:

Some of these split tests were run in that store.

Brett:

Yeah, yeah, yeah. It's done quite well. I think it was an improvement. I know kind of the, there's this rage for headless, which is a topic for a different day and there there's scenarios where it makes sense. But for, in this case, simplify, run some of these tests. It was definitely a win for overtone for sure. So awesome. Kurt Elster, ladies and gentlemen, hit that tech nasty button one more time. <laugh> Tech message. Awesome. Kurt, this has been super fun, man. Thank you so much. And we will have to, It's been too long. I think the last time you were on this show was like four years ago so we will four

Kurt:

Years ago. That is too long.

Brett:

I know. It's no good. So we'll have to just get it on the calendar, get you back as a regular guest. So much appreciated, my man. This was awesome. Please do. Looking for to next time.

Kurt:

Thank you so much.

Brett:

Absolutely. And as always, thank you for tuning in. Could not do this show without you. In fact, it'd be pretty lonely. It'd be pretty pointless to do this show without you. And hey, if you're listening to the show and you think, man, this is enriching my life, this is making me more money, this is improving my e-commerce experience, then share it with somebody else. We would love that. And leave us that review on iTunes if you think it deserves it. And with that, until next time, thank you for listening.

Have questions or requests? Contact us today!

Thank you for reaching out! We'll be in touch soon.
Oops! Something went wrong!