Drive and Convert (Ep. 077): Optimizing Conversions with Low Traffic (Live)
Live from the Logical Position Summit, Jon and Ryan explain that low traffic doesn’t make CRO impossible. There are alternative types of validation to help optimize your site for conversions.
Listen to this episode:
About This Episode:
Conversion rate optimization is important for any ecommerce brand, but it’s markedly more difficult for brands with lower traffic. However, there are still ways how you can increase website conversions with low traffic.
In this week’s episode, Jon and Ryan highlight the significance of leaning into rapid testing in the absence of data to validate statistical significance. They also discuss other validation tactics, such as surveys, user testing, heatmaps, and many more.
Listen to the full episode if you want to learn:
- How to use the confidence/tolerance relationship to your advantage
- The benefits of rapid testing
- Validation tactics beyond testing
- How social listening and session recordings work
- Why eye tracking is simple yet effective
If you have questions, ideas, or feedback to share, hit us up on Twitter. We’re @jonmacdonald and @ryangarrow.
Subscribe To The Show:
- Apple Podcasts
- Google Podcasts
- Pocket Casts
- Subscribe via RSS
Narrator: [00:00:00] You’re listening to Drive and Convert a podcast about helping online brands to build a better e-commerce growth engine with Jon McDonald and Ryan Garrow.
Ryan: We’re talking about small sites and how do you improve convers on those sites. You’re limited on traffic, which helps a lot of that. So Jon, at the. , obviously I’ve known Jon for many years and you’re constantly talking about, yes, we should test that. Yeah, we should. How do we do that ? Because, uh, when you have small sites, it’s not always about AV testing cause you can’t get the data behind that.
And so if you don’t have statistical significance, which generally speaking, Jon’s, you know, back of napkin math says 50,000 visitors a month or more of like real traffic, not bot traffic. Cause that’ll skew Dave, but 50,000 visitors a month or more is how you run ab tests. Vast majority of our clients in logical position and in on the internet don’t have that.
So you can’t run true [00:01:00] c o as it would be defined by you. And so the goal today is to discuss, you know, the different types of things you can do and, and kind of a term we’ve come up with kind of jokingly, but probably is more appropriate. C i I still haven’t yet to trademark that. We probably need to do that.
Unfortunately we steals it. But talk about how these low sites, low traffic sites, can improve their conversions and still have some good decisions being made even if you don’t have full data. High confidence level is a must when it comes to validating. So when you think of it as in, in terms of high confidence and how, how much you tolerate that confidence when making your decision.
Jon, what is our tallest from being wrong? Like what you don’t wanna do is say, yeah, we saw the data and you only had 5,000 visitors, but you should do this.
Jon: Well, and that’s exactly goes back to your conversation around needing to have enough traffic, right? So under 50,000 visitors, what do you do? How do you.
And I think part of the way to look at that is that if you were reducing the traffic in tests, you are just [00:02:00] reducing your confidence, right? Because you’re gonna have less of a sample size. So statistics, right? So the thing to be thinking about here is, are we saving lives in e-commerce? unlikely, right? If you, if you’re building airplanes, you need a 99.99% confidence level, right?
Because you don’t want that bolt to break and then a plane to crash. In E-commerce, if your site goes down for a couple minutes, the worst thing that happens, you lose some money. It’s horrible. Don’t get me wrong, that’s not what we want, but the reality is nobody’s gonna die. Right? So one thing I want everybody to be thinking about as we talk about this today is that really all you’re looking to do is.
Your confidence level a little bit by saying, Hey, less traffic means it’s not gonna be statistically significant, right? We’re not gonna be able to, to save a hundred percent accuracy. This is the winner. But you can still optimize. And there’s a lot of ways to do that. And we need to be okay with saying, Hey, I have less traffic.
Maybe I’m not as confident, but this is still way better than [00:03:00] taking the wild. And you’re probably not gonna get, like, uh, one of our mutual clients who will be here today have one test that generated an extra 1.3 million on month. Well, look, you’re probably not gonna get that. You need to have enough traffic to be able to get that.
And enough revenue, obviously, right? Yep. To where you can see that type of return on investment. But even for small brands, You can see thousands of dollars of, of a additional revenue and, and one of the reasons that we partner up so well between the good and logical position is because you guys drive qualified traffic.
We help make sure that it converts once they get to the site, that the deal with that is increases roas. Right. It just, it makes you guys, your team look better. And so that’s really why we’ve partnered up all these years.
Ryan: Yeah. And I, I liken it just too a great book called The Slight Edge. Like all you wanna do is like small little steps up to the right every time.
If you can make that on a small website, you will go to a big website. Yeah. At some point.
Jon: Well, that’s exactly it. We actually have t-shirts at the go, though I didn’t bring any today, but they say you get 1% better every day, and that’s really the goal. You should just be looking for [00:04:00] that, that slight edge every single day to make your site better and then, uh, you’ll win.
Ryan: Got it. Okay, so we know we’ve got small sites can’t utilize official AB testing, right? So what are the different tactics we’re gonna be looking at? Validate some of these tests or ideas versions.
Jon: Yeah, great question. I actually, I, I have brought notes cuz there’s so many of ’em here.
Jon: The first thing, the thing about it is since we can’t do quantitative feedback right around actual AB testing, analytics, et cetera, cause we don’t have the data number of traffic to do that, let’s focus on qualitative, right?
So there’s a whole bunch of qualitative feedback options. So we could do what we call rapid testing and I’ll dive into that in a second. We could do survey. Right. Talk to a handful of folks. User testing. Who here knows what user testing is? A couple of people. Okay, so this is where you would send people who match the brand’s ideal customer profile to the site.
And we’re gonna ask those people to do some tasks on the site while we record their webcam, their screen, and their [00:05:00] audio. What that allows us to do is ask these people to complete tasks, and then they’re talking out loud about the experience that they’re having. So they’re telling us, Hey, you know, I’m, you told me to go find a t-shirt my size, but there’s no size filter for.
So it’s gonna help you kind of get outside of that jar. Ryan’s heard me say this a hundred times, but it’s really hard to read the label from inside the jar. You’re so close to your brand and you use your website every day. You know your products. It’s how to get outside of that jar, and that’s what user testing helps you do, helps you understand.
The experience for that new defile customer that your traffic is saying,
Ryan: Is that easy to get? I mean, I wouldn’t even know how to do that. That’s why I was like, tell my sister, Hey, uh, can you go buy something on this site and tell me what it is
Jon: You can? So there’s a really simple way to do this. Go to your local Starbucks, take your website on a laptop, say, so the people waiting in line, I’ll buy your coffee if you can gimme feedback on this website and just takes five minutes.
You just ask them to click through and tell you what they’re thinking. Talk out loud about what. [00:06:00] That’s the cheap and easy way to do this. We also, the Good recently purchased a brand called User Input. So it’s user input.io, and that allows you to do user testing that is geared for SMBs. So it is a price point that every brand should be able to afford and get a good return on that investment.
This is something Ryan has really been pushing us to do is, hey, we have all thousands of clients in these smaller size. We really need you to have an offering for them. So, uh, user input is our solution for that.
Ryan: I asked y’all. He’s been great. It’s great.
Jon: All right, so, uh, we have other options too besides user testing.
We can do reviews and social listening. So reviews is a goal. Mine. If your site reviews go in there and hear what people are saying, what information are they surfacing that isn’t on. So we have often found that one of the ways that consumers shop, and you probably all do this too, is if a product detail page doesn’t have the information you’re looking for, you go down to the reviews to look for it, and you generally are gonna trust reviews [00:07:00] more than the content that’s on a pdp.
Otherwise, because it’s not from the brand, it’s from other consumers. And so, , that’s what visitors are doing on the site. They’re going to these reviews and you can mine the data as well by looking through that saying, what are people saying this size? Larger than true size. Well, maybe that’s content that you should be putting on the pdb, surfacing that up for all consumers so that that people don’t have to go down to the reviews to get that info.
Ryan: If you don’t have reviews, we might have somebody that can help you with that.
Jon: There you go. Um, social listening, same thing, but go on Twitter, Facebook, et cetera. Search to the brand, hear what people are saying.
Narrator: You’re listening to drive and convert The podcast focused on e-commerce growth. Your hosts are Jon McDonald, founder of the Good, A Conversion Rate Optimization agency that works with e-commerce brands to help convert more of their visitors into buyers.
And Ryan Garrow of Logical Position, the digital marketing agency offering PaperClick management, search [00:08:00] engine optimization, and website design services to brands of all sizes. If you find this podcast helpful, please help us out by leaving a review on Apple Podcasts and sharing it with a friend or colleague.
Jon: Session recording. This is one where you can actually, at any traffic level, Do a great recording. So really with session recording, what you can do is watch how people are engaging with the site. You can do this in an anonymous fashion. Great session recorders, almost all of ’em are California privacy and Europe privacy compliant.
They block out any information that would be personally identifiable information and what you’re able to do is see where is their cursor going on the screen? What are they clicking on? You just can’t see what they, what text the input in the fields.
Ryan: How is that different than heat mapping?
Jon: General heat mapping follows the mouse and it creates a heat map of where the mouse movement is happening the most.
Right? And that is another data set. You really don’t need thousands of visitors to get a couple [00:09:00] hundred, and you’re gonna have decent information about where people are lingering on the site, what they’re clicking on. But the difference is a session recording is gonna show you one user and what they’ve done and where their mouse is moving on the page, what they’re clicking on, right?
So you can go through each of those. And look at how people are engaging, where they’re scrolling, you’re watching a replay of their interaction on a page.
Ryan: Does it still work on mobile devices? You just can’t see the mouse?
Jon: Yes. It’ll just show the screen and I’ll, and some of ’em even will show you where they tapped, uh, if it’s on a clickable object.
Ryan: Okay. So those are some of the options you have as a small brand to start collecting some of this data or validate right changes.
Ryan: As a brand though. So my mind goes, how, how do I execute some of that? Like, yes, I can do some of these. , but where does the idea for a test come from and like, do I decide like the button chippy blue or red?
Jon: Ryan’s triggering me with that question. He knows button color testing is not optimization. I’ll just say that right now. , he knows that I have my point of view on that, but okay, [00:10:00] so let’s talk about tactics on how you can do some of these. Right? Okay. So preference testing is where you would put up two different images and ask what people prefer just in the name.
Pretty easy. You can also do. With ads. So for instance, maybe you really have eight or 10 ads in your designs in messaging, and you’re not sure which one you need to filter down so that you can launch a few campaigns. You can do a preference test where you are testing all 10 of these ads against each other and seeing which ones come out on top.
And then you know you have some data to go in that you’re gonna put money behind and you’re gonna feel a lot more confident about that. Really, all of these are just. to help me increase your confidence.
Ryan: Okay, got it. So I, if I’m a small business, I don’t have a bunch of people to put it as in front of necessarily.
Mm-hmm. , like I could ask my employees, but they’re in the jar. Right? That doesn’t help. So I would say I wanna run some display app tests with these three images to see which one gets a higher click through rate.
Ryan: To indicate which image they like [00:11:00] better.
Jon: You can do it live. Which is what you’re talking about, load ’em up into Google and see which ones do better.
Or you can avoid spending the money in, in terms of giving it to Google. And you can do preference testing a lot of different ways. There’s a bunch of tools out there, user input. We’re actually developing a tool to do this as well, but there’s the simple way, who here knows about Amazon mechanical term?
Oh, okay. If you don’t look this up. So Amazon Mechanical Turk allows you to pay 10 cents per test to get somebody to complete a simple. and you can segment the visitors. Now again, it’s not that great of segmentation. You’re gonna get a lot of international folks. Probably not very close to the ideal customer profile, but if you’re trying to test high level messaging, high level designs, It works pretty well.
So Amazon Mechanical Turk allows you to put up and do simple tasks for 3D cheap. So that’s a cool hack. Something you should look into. And what you could do is just create a PDF with your, uh, each page has a different image on it of your ad. Put it up in a mechanical Turk and say, if you like options [00:12:00] A, B, or C, give them something to help them, right?
Like, Hey, you’re looking at buying a T-shirt. Which of these would get you to click through more? Right? So again, we’re not flying airplanes. So you don’t need to be super scientific with this. You’re looking to get more confidence than what you have otherwise.
Ryan: Got it. So I would help with a product image, like, Hey, I’ve got this product image where it’s, Hey, this way, and one more, take this way. Which do you like better?
Jon: Right, exactly.
Ryan: Okay. What are some other types of tests that can execute some of these tactics for flow traffic?
Jon: We talked a little bit about ads. Uh, you could also do this with headlines. That’s a big one. Up on onsite when you’re looking at headlines and copy. Often we’ll do this type of testing with copy and headlines because it helps understand what is capturing attention.
What’s getting engagement? That’s really preference. Testing, rapid testing. Also, there’s other tactics here. You could look at what’s called tree testing. Tree testing is usually done around navigations, so the idea behind this is you’re gonna have a navigation that’s generic. You’re not gonna design [00:13:00] it or anything of that sort.
You just need the text. And you’re gonna allow people to click through, make it look like an accordion, accordion navigation, and give them some heading, Hey, I need you to find all of the black t-shirts. And then give them the navigation, and then ask them to click through and find that. Now what you’re gonna do is get really good understanding of where the consumer wants to go versus what your navigation is allowing them to do.
Right? So pretty simple, but it. Really good impact on navigation, uh, redesigns and some of you here have some clients with navigation that needs to be blown up, period.
Ryan: You found some sites with atrocious navigation, like if there are 20 links at the top on your menu, be done. No, sorry, five to seven. Five to seven.
There are some best practices that you can implement even before you get to the point of rapid. , which is important, and Jon’s site has a ton of content on that. Yeah, you probably just send your client to that. Say, Hey, your site needs some help and maybe you don’t have the money to do it all yet, but just look at some of [00:14:00] this and implement some changes because it’s starting bad and we can make it better, and then get into some of this rapid testing.
Jon: Yeah, everything I’m talking about today, we have articles up on our site about we, we have an encyclopedia of optimization up on our site and just go to the good.com, click on insights, and there’s a search bar. Just enter whatever you’re looking for and I guarantee you we have content. We’ve been publishing content for 15 years up there, so there’s a ton.
So tree testing, I was talking about a couple things that it can help you measure is how long it takes somebody to find something into. , right? That could be interesting how likely they are to discover something. So again, I said, Hey, find black t-shirts, et cetera. And then really the idea behind this is that you already have a navigation you’re looking to optimize or some ideas of how to improve it.
This isn’t something where you can just put your current navigation up. Unless you really want to understand how bad it’s, that’s generally what that’s gonna be useful for. Let’s talk about eye trackers. Cause I want to do that next with the tear down.
Ryan: Yeah. It’s, I think it’s an, it’s one of the easier ways to get [00:15:00] jumpstarted.
Jon: It’s a really easy way, mainly because it doesn’t require any traffic at all, because what you can do anymore is artificial intelligence, eye tracking. We actually have a tool that we’ve used that has been seated with thousands of eye tracking studies, and it can now with about a 98% accuracy rate, give us an eye tracking heat map if we just upload the image or send it to the website.
So that’s what we’re gonna take a look at here in a minute. As.
Ryan: Good because I think whenever you can use AI to give you the starting point, it will jumpstart a client into improving their website. Just getting the first step I think is always the most important for a lot of our small clients, just to say, you need to improve your site cuz it’s their baby too, right?
You’re telling your client you have another baby. Fortunately it’s another thing, but that first step is like, let’s te let’s do you to test something to see what it’s going to do to improve it. And once they get that first one, I think it’s gonna snowball into more and more and more like, oh my. So it’s not hard to do it, but the perception of conversion rate optimization is expensive [00:16:00] for a small business.
If you’re gonna talk to Jon about cro, 10 grand a month is not in inexpensive for some of the small business. It pays for itself, but they’re not usually gonna take that step. So it’s like how do we get them to that first step? Baby steps, and the eye tracking is well, and
Jon: The nice thing about eye tracking is, especially when you’re going to tell someone their baby’s ugly, is Ryan.
You want data behind that, right? You don’t want, it’s not an opinion. It should be a data based opinion at that point, right? There’s a pretty baby. Here’s your baby, right? Like, yeah, exactly. Like, you know, I like your baby, but you should see all these consumers, they’re not even looking at your baby, right?
They’re looking the other way. So something’s going on here. So I’m gonna give you some examples of that today, because I think that’s helpful for having difficult conversations, which is why use that data for tear downs. It does help disarm the client because it’s not you. Look, this is what people are looking at.
Ryan: And that’s my cue to get outta your way, right? Cause you’re can actually do some,
Jon: Yeah sure we can do that. We’re about up at time, so I wanna make sure that we get through the Teardown. So Whyt, we do that. That’s more exciting question. That’s what they’re here for. All. Well, [00:17:00] thank you Ryan.
Narrator: Thanks. Thanks for listening to Drive Convert with Jon McDonald and Ryan Garrow.
To keep up to date with new episodes, you can email@example.com.
About the Author
Caroline Appert is the Director of Marketing at The Good. She has proven success in crafting marketing strategies and executing revenue-boosting campaigns for companies in a diverse set of industries.