In e-commerce, scaling happens outside the ad account. In this article, we are sharing an interview with Carl Weische, who talks about Conversion Rate Optimization, how to do it, what are the best tests to try, and everything else in between.
Carl Weische owns a conversion rate optimization agency known as the Accelerated Agency. They do CRO for many big names, like David Fogarty, Brandon Nguyen, Drippy Amsterdam, and more.
Q: Let us give the audience a quick overview of why they should be doing CRO, what it is, and how that can help them.
The best introduction I can give to CRO is understanding what your customers need and then giving it to them on your website. You can think of CRO as optimizing the website's performance. CRO increases your conversion rate and average order value, so the traffic you push to the website will generate more revenue and profit. So I like to dissect this into ad performance and website performance.
Q: If I own a brand, when should I be looking at optimizing my off-ads metrics? When should I be looking at optimizing CRO?
It would be beneficial if you began with the fundamentals of CRO because there is a lot of stuff that requires 20% of the effort but yields 80% of the results. I am talking about product market fit, offer, and messaging. Those are like the 20% at the beginning. Still, you do not need to make crazy UI or copywriting changes to your website; just the basics will get you great results. Then, as soon as you hit $100 000 - $150 000 in revenue, you should start moving forward with the higher-level CRO stuff and a/b testing macro and micro changes to your page. Moreover, that is what you should be doing from the beginning.
Q: One thing I remember from the speeches you have given is that the perception I previously had with CRO was that you would test two completely different landing pages or, let us say, compare your product page to, I do not know, like a custom-built landing page or an advertorial to see which would perform better. But a lot of the changes you make are relatively minor, but they have significant impacts. You could run us through it. I own an e-commerce store and make $150 000 per month. What should be the easy tests I can do on my own, and where can I improve my conversion rate where I should be looking?
Great question. Initially, I like to lay out a strategy individually for each store. This all starts with understanding your customer — who you are selling to, how they interact with your store, and what motivates them. You can apply a lot of what you're doing in marketing, such as understanding their pain points, beliefs, desires, and motors, to the store. I see a lot of great marketers who drive insanely high-quality traffic to their stores. They have ads that are great at grabbing attention because they understand the target audience, but then they need to do the same stuff in the store. They put less effort into the messaging, the offer, the design, and everything else to convert the attention they are getting from the marketing into people buying from them.
So this is what I like to do in the beginning: set up tools like Google Analytics and Hotshot to visually see where they are clicking, where they are scrolling, where they are spending time on the page, and what the customer journey is throughout the pages. You may be pushing traffic to the landing page. They are going to the product page but then dropping off to the home page because they need more information about you as the brand and the story behind it, and then they will go and buy. So you can individually understand and analyze the customer journey that your target audience has while purchasing from your store. The next step would be to jump into it and optimize all the different stages of the customer journey. So this is what we like to do in the beginning to analyze because each store is different.
Analyze it for one store to understand why they buy, what motivates them, what their pain points are, what their beliefs are, and what they want to achieve from purchasing this product. Then we will think about tests we can run that are tied to this. Much of the stuff that you can do is mainly tied into increasing the trust that they have in your store — something like a trust badge or something like store benefits, something like social proof testimonials — everything that's playing into you creating more trust as a huge brand and a quick win. What I also see is that you need to improve your messaging copywriting and video proposition.
This is a lot of stuff we like to do at the start, so dive into message-mining customer service, analyze reviews, and then take exactly what they say and incorporate it into your value proposition. So I tell people all the time to remember this. It is easy to export all your reviews, and all the customer feedback and testimonials are put into a massive spreadsheet. Then you have 10 000 reviews and 10,000 answers. Then you can just read through the first couple and see what things like comfort, high quality, and great sleep are. Then you pick the first words. Then you search in this document to see how many times people have been searching for this, and then you take the words people use the most to describe why they are buying a product. And then you are just taking these words and putting them on the landing page, the product page, and everywhere, and then you will optimize the messaging. You will also increase how this messaging resonates with your customers because now you are using precisely the words they would use to describe their feelings and thoughts. Then it resonates with so much fun that you are much more likely to convert them than when you would be talking about it from your perspective.
So the basics, as I said at the beginning, are offered: product-market fit and messaging. The first tests I would run are conceptualized around increasing motivation, trust, and the value proposition of copywriting. Furthermore, after having done all of this, you move on to the small and nitty-gritty stuff. A quick example is reducing friction in the checkout process by giving them minor dopamine hits, showing green check marks whenever they fill out a form, and other really, really small nitty-gritty things that, at the end of the day, we give them a notch. We will give them a push to go through the checkout. That is like high-level, nitty-gritty stuff at the end of your CRO journey. So, at first, we concentrate on the big needle movers.
Q: How evergreen is the CRO in terms of changes? You run a test to improve the conversion rate on a specific product page. You'll get to know the winner of the test. Is it going to stay that way forever, or are there some retests you have to do after a while?
So the thing with CRO a/b testing and the real statistical significance behind it is that, to break this concept down quickly, whenever you test, you, for example, test something for four weeks. Then you get data on 100 000 users and 5 000 orders, so you have a massive amount of data. And then, you can apply statistics to it and calculate how significant this result is. And then, for example, you would have a statistical significance of 99. This means that with a probability of 99% when we now take this change and implement it in the store for the next six to twelve months, it will perform better with the result that it did while we tested it. And then you have a 1% chance of failure. Then, after six to twelve months, the target audience gets used to seeing it, and the effect depletes a little bit. That is similar to the lifetime of changes, which is used to store how long and valid these changes are. The higher the statistical significance that you are going for, the higher the probability that it is working for longer and with minimal potential for failure, but that is like the time frames that you are looking at in testing.
You can think of it as the creatives and ad sets. We are A/B testing creatives, and we have two different ad sets. Then we compare the KPIs, which is what you are doing on the website. When you do specific creatives, and then you may run an ad set for 6 or 12 months at some point, respecting the platforms, Tick Tock has a faster, more significant fatigue than Facebook or YouTube. You will have a specific time frame in which you can run a particularly creative project; at some point, the results will gradually deplete. Furthermore, the same is true for things you apply and implement in the store.
Furthermore, we see this with stores; they believe that whatever they have a scale of $500 000 a month, and then we do many changes with them, and they say, "OK, now we should have done enough changes."However, they scale, and then, for example, the traffic composition changes from 80% Facebook and 20% Google to 50% Facebook, 20% Google, and 30% Tiktok. So you have a completely different target audience coming onto the store now. Also, the store is now at a $1 million per month scale. Hence, everything changes because the traffic is different, how they interact with the store and never get through it, and how you talk to them at certain stages.
When you reach the scale where it makes sense to put the effort, time, and resources into getting out of these other landing pages, I recommend using different landing pages for your various traffic sources because, for example, Tiktok traffic requires a lot more easily digestible, short-form copy, really visual, and full of imagery. For the Google retargeting traffic, they are purchase-ready when they put in the keyword. So you can hit them with longer copy, less imagery, fewer visuals, and crazy animations because they have a higher attention span, so it is different from what kind of traffic channels you are pushing to the page now, what demographic, and how the audience may change.
Q: When you are doing conversion rate optimization, I also like my assumption about this. It is not just looking at the camera conversion rate and increasing that; there is also AOV that comes into play, and possibly sometimes LTV. LTV is more critical for subscription brands. So please keep it simple. I want to figure out how you balance, uh, I will call it a "triangle of CRO" between increasing the Nov and increasing the like click-through rate. Are there sometimes cases where you prioritize AOV but lower, you know, the conversion rate on the site, or vice versa?
Yes, this is also based on the brand strategy. So the three most important metrics we like to look at are our conversion rate, average order value, and revenue per user. Revenue per user is also an exciting metric because it means how profitable the traffic you are pushing to the store is. So we are not increasing the revenue but making the traffic more profitable. So this is based on the brand strategy. Right now, we have some brands that want to stay at the same revenue level rather than scale up and get more revenue. Still, they want to increase traffic and margins to increase profits. So what to focus on is entirely dependent on the brand strategy.
For example, we just tested it using different bundles for a product. We chose the most expensive bundle and a/b tested it. The variant had a 1% lower conversion rate but an uplift in average order value of 20%. so you have fewer orders. Still, all the orders you are doing have a higher average order value and are more profitable for you as the brand, so this is hard to generalize but rather individual to the brand strategy. Still, to answer your question, there are cases where we prioritize the slight decrease in conversion rate. Still, we have a higher average order value or higher revenue per user in all the tests where we may have a different revenue per user and average order value but a slight increase in conversion rate.
Q: Do you have much access to brands that you know have tested a variety of landing pages? One of your suggestions was not from a conversion rate level but from the way you structure your business for each traffic channel. You can have a different landing page. What other significant macro-changes could the brand make to improve its sales?
So the most significant micro change is understanding what funnel works for what traffic channel. So we test various funnels, such as from the landing page to the product page, from the landing page to the checkout page, or from the editorial to the landing page to check out right away. Hence, there are a lot of different funnels. We can be pushing traffic to, or you can be driving traffic directly to, the collection page or the product page. So I would test out all of these for the different traffic channels and then go with the one that's performing the best and then start to put more effort into this because, in the beginning, we have to understand which funnel is working the best. Maybe your listicle to landing page funnel has a 5% conversion rate. Then the advertorial to the landing page funnel has a 3% conversion rate. So I don't necessarily want to put effort into the 3% one and try to raise it to 5%, but rather to start with the 5% one, direct all traffic there, and then begin increasing this conversion rate and directing all efforts toward this funnel.
So the first macrostep would be testing out the different funnels. Then the second step would be to go with the funnel that's working the best and then start making some minor changes that are best for the listicle. I taught this to JB Coffee. For example, you have these five reasons, just as the proposition begins by going into your customer research and digging through all of the answers to understand what grounds they have for purchasing the product, so you are not talking about five benefits. You are giving that away, but you are taking their words and putting them into these five benefits, and then you can A/B test this.
This, for example, would be an excellent test for a listicle. Try out five reasons and base them on customer research or message money. Another big thing could be testing various offers, right? So other bundles, different subscription models, and pricing would also be huge. And then, obviously, from the listicle, the next step would be the product page. Test the above photo of the product page. So in the first step, you want to increase the click-through rate from the listicle page to the product page and then the conversion rate. You could do all the offer testing again with a little copywriting. Still, those are the big picture ones and differ from the small ones like trust badges, store benefits, announcement bars, or whatever. It is testing different styles of landing pages, various offers, and product photos and doing research with the customer to understand the terms and words they use so that you can integrate them within the landing pages.
Q: What are the biggest CRO mistakes you see brands make?
Many brands and agencies do this when they believe they are a/b testing something but need more data before proceeding. Consider this: you have two ad sets, each with 100 clicks, and the first one has 5 sales, and the second one has 7 sales, and you decide to completely scale up the one with the 7 sales because it had more, and this is what a lot of brands and agencies even do for clients or on the store that they have tests running for a week and then they have whatever 1 000 users and like 100 orders, and this is not statistically significant at all.
The second most significant mistake is trying to copy other brands or other tests or changes that other brands are making because your brand sells skin care products for women. Then you are copying a high-converting store that is selling shirts for men right now. Then you are trying to adopt some of their changes because they look good or you think they are working well, like trust badges or whatever. Still, it must be taught to the female audience trying to buy skin care now. So even though something looks good, or competitors are doing it, or it is working well for competitors, or they had a great result, I suggest against just implementing it in your store. You can take the test and test it; if you are getting the same data, go ahead and implement it.
Still, many brands cherry-pick what they see on the market or from competitors and apply it to your store. Then they are sitting there wondering why they have a 1% conversion rate. So this is the type of broad stuff that I see many brands doing. Also, many stores look the same. Then it is not about having a good-looking store or really lovely design in-store but rather about aligning your store with the various beliefs and pain points of your target audience using precisely the words that they are using in the copywriting and resonating with them with the copywriting and emotionally, and that is how you convert them.
Q: In terms of, like, conversions, what is the golden standard? Like, what is the percentage that is accepted as being very good?
This is based on your industry and your average order value because it is easy to have an 8% conversion rate with a $20 product. Still, it is hard to have an 8% conversion rate with a $200 AOV. Let us throw some cases at you, and then you can give us the range. Yeah, I could. I could also give you some examples of clients without naming them. So, for example, we have a $1 million store. Also, the scale is significant, right? So they added $1 million and have a $150 average value. Then they have a conversion rate that's between 5 and 7%. We have another client with a 1% conversion rate at a scale of $300 000 per month with an average order value of $1 500. Then we have other clients with a 5% conversion rate at the $500k scale with a $110 AOV.
So it is based on the niches, based on the above, and based on, like, what scale you had. However, we see that many AOVs fall between $80 and $120. So with this, you can get your conversion rate to 5-7% percent at scale.
Q: What would be your advice? There may be something on your mind that we still need to bring up or ask you about. Well, what do you think our audience should know about CRO?
Many people overlook it or need help understanding it because it applies psychology to e-commerce and consumer psychology. Your design is influenced by how your target audience behaves and what motivates them to buy. So it is not about taking trust badges and putting them on your page, but instead, on a psychological level, understanding, for example, what trigger is it affecting now? Are we using social proof here? Are we increasing the motivation, reducing the anxiety or fears, giving them more incentives, or improving the value proposition? So this is the starting point for CRO when you realize it is all based on consumer psychology and emotional intelligence. You are going to level up your CRO game. Then, all the good and great marketers, know this with marketing when they start looking outside of the ad manager and into books like scientific advertising or consumer psychology or human biases, or something else based on psychology. That is what the top 1% of marketers do right.
Q: So, if you jumped in a time machine and traveled vented in the future when you were 30 years old, what is one piece of advice you would give your 30-year-old self? What is something you want to say to them? What is something you want to remind them of? Like advice, how could it be more successful?
The advice is to let you say, "Tell them to stay hungry, stay foolish, enjoy themselves, and never stop going." I could believe that when you are like 30 years old, you have more success and may take a step back. The advice I give, and it applies to any age group, is to basically pinch yourself every time and get back to work and crush it.
Q: Before we wrap up, where can people find you?
The best place to find me is on Twitter, so just my first and last name, Carl Weische. Just put it on Twitter, am most active there, and I push a lot of content. You can reach me anytime, and I will respond.
We thank Carl and the team at AgencyJR for setting up this interview. Carl came up with some great concepts that we believe you the readers can take home and put into practice. If you thought this episode was valuable, share it with a friend who might find this helpful information and leave a comment so we can reach more influential people and bring lessons back to you now.