Why good copy performs badly

Presented live on Tuesday, July 23, 2019

Register for our live tutorials

So you write this siiick landing page… or this brand-new website…


And a month later, you meet with your client for the retrospective. 

You’re all excited. You can’t wait to see the numbers.

So your client pulls open her Google Analytics dashboard… shares her screen… and launches into these hard, hard questions about the thing you dread most:

“Any idea why your copy isn’t increasing our conversion rates?”


Sometimes the problem IS the copy. Or the strategy. Or the offer. 

Other times… it’s none of those things. 

Instead, it can SOMETIMES be a problem with how the data is being reported and/or interpreted. 

Side note: If this topic sounds dull, wait until you nearly go into cardiac arrest during a post-mortem when GA appears on the screen. Then you’ll be glad you watched this Tutorial Tuesday, in which Joanna walks you through 3 reasons the data is making your great copy look like sh*t copy. So you can turn awkward client convos into even more reason for the client to trust and believe in you.


Joanna Wiebe: We are talking about data today. Now we’re not going to dig into any data because copywriters, if you want to specialize or learn more about data and analytics, then you should go and take some training on that. That’s really good to know, but you don’t have to be able to get through Google Analytics to be a really good conversion copywriter. You just need to know to ask good questions so that you are not caught off guard by not having asked the question and having somebody else ask you a question that you cannot answer.

I’m going to walk you through. Now, this was, this is the kind of thing that as a conversion copywriter, you should be thinking about data and not just about, “Oh, I want to A/B test my copy.” That’s great. That’s one part of what we do. Again, you don’t always have to do it because you always can’t do it. But you do need people to speak and think in the data side of things. And that doesn’t mean that you, again, have to be an expert at it, but you have to know what’s going on.

Let me share my screen with you and I’m going to walk you through this, and then any questions you have, throw them into Q&A. This is a bit of a background. I know, the whole time I’ve been running Copyhackers, we’ve been very focused on of course copywriting. But prior to that, I was a consultant with Conversion Rate Experts, and prior to that I was with Intuit, which is very data driven. I was still copywriting there, but in a conversion focused sort of capacity. So very familiar with all of the ins and outs of data and analytics and all of that great stuff, but not an expert at it. Again, you don’t need to be. But you do need to think about this stuff.

I’m going to share my slides. So the real question today is like, okay, what happens when your client runs your copy and then says a month later like, “Oh, let’s look at the numbers,” and asks you hard questions like, “Why isn’t your copy working?” The goal here is to better understand things that might be making your informed smart data-driven copy look terrible, look like the worst, performing maybe worse than what they had before, et cetera, et cetera.

This happens to all of us. I’m going to get into what’s going on. So the client of course asks you, “What do we do? Your new copy is tanking our conversion rate.” And they bring up Google Analytics and you look at it and you’re like, “Oh down, the numbers are like not lying here. It really looks bad.” Is it that your copy is actually tanking your conversion rate?

Now, when you’re calm and cool and you’re like, “No way, there’s no way I could do that,” then you could easily start challenging and asking questions of the data. But when you’re in the heat of the moment, it can be a very intimidating sort of thing. You’re looking at Google Analytics. One or more members of the client’s team are kind of staring at you, asking you these questions, and you have to come up with some sort of response. It doesn’t have to be an answer, but it has to be a response. So this is kind of the problem set up in that moment. First things first, obviously you want to keep your cool.

This is part of, you’re not here to be a dancing monkey who puts on a show of like, “Oh, look what I did to your conversion rate.” You’re an expert, you’re a pro of what you do, and you followed a process. So take comfort in that process. Sometimes scientists go through trying things again and again with phenomenal hypotheses where you should be like, “Yep, this is going to work perfectly,” and it doesn’t. Same thing goes here. We don’t always know the answer, but we have a process and we have to take comfort in that. Also, remember that the client signed off on the copy too. So they believed in the copy just like you did. They’re probably as confused as you are.

So everybody’s actually on the same page here. You all just want some answers and you’re on the same team. They’re not trying to find reasons to what? Ask for their money back? They’re trying to understand what’s going on here and so are you. So cool. Start with that. Remember, this is the process. We went through this in our last tutorial. It is live today I think through the Conversion Copywriting Process, everything that goes into it and the process itself, so you know that.

Second, be open minded. So we want to avoid confirmation bias. We often find that when people put copy up there, they’re looking just for ways to prove that it was right, that it was better than the control. But sometimes it’s not. So let’s keep in mind that everything you did was right. Doesn’t mean that it ended up being the perfect solution, but you did everything right. It’s okay. Try not to look for it to be. Try not to look at the data and say, “There’s no way this can be true. Let me look for a way to look at these numbers and find out that my copy was fantastic.”

That’s okay. Sometimes we lose. As long as you had a good hypothesis going into it, or a research question, now you can come out and say like, “Okay, that wasn’t great. We thought this one thing. Our assumption was wrong. Let’s move on,” because losses are of course a good signal too. As long as you’re dealing with clients who are not assuming that in hiring you, they’re hiring somebody who can put on the show of increasing conversion rates. They have to be down for always optimizing. If they’re not, then they’re probably going to be looking hard with numbers in the wrong way. So this is why we want to of course as we teach in [inaudible 00:05:39] copywriter work with a better class of client who is more interested in long-term growth and iterative wins and things like that.

Okay. All of that said, what might actually be going on though? So once you’re like, “Okay, I’m keeping my cool, and okay, I’m not looking for reasons to prove that I was right.” What might be going on? Like what if your copy actually isn’t performing, or what if it is? Once we actually get into the data, what are the four things that … There are more than four of course, but this is, these are four really kind of obvious ones to jump up to.

The first one. One of the culprits might be that your copy isn’t actually working. That’s part of it. Okay? So you have to accept that that is one of the options here. Your copy, even though it’s informed by data, it might’ve been colored by the wrong things, might’ve been looking at the wrong data sets, whatever that is. You were looking at optimizing for one segment when in fact the other segment, what’s going there, God only knows what it is. But your copy isn’t actually working. That’s one possible culprit. Yep, your copy might not be working. Okay.

Two. This can be a more interesting question to start asking your client. So you’re sitting there, you’re looking at Google Analytics and they’re like, “Okay, look, Joanna, we launched this landing page in June,” or, “We launched this landing page in July, and look at it compared to the numbers in June,” where the numbers in June are like 30% conversion rate and the number in July is a 20% conversion rate. What is going on there?

Now really quick, a question that a conversion consultant would ask is, “Oh, okay. So maybe there’s seasonality at play,” and ask you to instead of looking at month to month, month versus month, go year over year. So say, “Okay, that’s June versus July. But how about July of 2018 versus July of 2019,” because it could be that every July conversion rates naturally tank. Because of the summer. People go on vacation. God only knows, again, why that might be. But this actually happened recently for me with a client where we launched something new for them, and a month later we looked at the numbers and there was the June versus July, that conversion rate for our variation was down by 4% and that was, that was what they were looking at. And that’s a perfectly normal thing to look at.

But then there’s the other thing of like, “Well, hold on. Cool. Is this the normal? Is it normal for conversion rates to drop from June to July?” And so you have to ask that going into and looking at last year. So tell me now 2018 versus 2019, and when we looked at that, the conversion rate was actually up by 19% versus last year.

So if you look at the numbers wrong, then you can be very misled by that data. So June versus July might not be the best comparison. Year over year sometimes might not be the best comparison too. But it’s important to ask both of those questions. If someone’s only looking year over year, you also want to look month versus month. And if they’re only looking month versus month, you also want to look at year versus year. So that’s an early question to ask, especially when your client has Google Analytics open and is already doing those comparisons. Ask about seasonality. Look in the data to see because it might be, “Nope, conversion rate is still down.” Okay, fine. That’s one more question answered though. Now we can start to really move forward to another one.

They’re assuming this really common problem, which is this whole last click caused the conversion sort of challenge that people have in setting like, this is really true for emails as well. A lot of emails can nurture over time and people are on various devices and all sorts of different things that they’re doing. And all we do is measure that final moment, the moment at the point of conversion where we looked through maybe a customer journey for people in that moment when it’s like, well, they always hit this landing page and then hit this page and then go to pricing. And that’s when they buy. But we have been measuring across different devices because people are looking on different devices and tracking, can’t really track that well.

So an important thing. If your client brings up a conversion rate problem, is to make sure you’re getting a really good understanding of the full funnel. People like to zoom in at the only part that you worked on. But each part of the funnel depends on everything else too. So if you only wrote emails, or you only wrote a landing page, you need to make sure that they’re not making the wrong assumptions about where the conversion actually happens just because they had a narrow view of that. Sometimes this won’t come up, but sometimes this will.

And that brings us to the fourth and final one. And this is a really big one too. These are like the four most common things of they’re reading their data too soon, too often, or both, for too few people. Statistical competence is a real thing. It’s an important thing. If you’re going to look at something like month versus month and you had 300 visitors in one month and 200 visitors in the other and your conversion rate is 2% and you have … That’s not enough data. There’s just not enough to work with.

So you have to be okay with talking to your clients and saying like, “I understand what you’re seeing. I understand how looking at this it looks like it’s down.” But there’s so much room for error in how low these numbers are, that we need to develop better ways of measuring the success of this copy, remembering that we all went into this believing strongly in what we were putting out there, in everything that drove us to this point. So we need to come up with a way to better assess, and that might be a timeline, to better assess how this copy is actually doing.

Clients want to see that the copy you wrote is converting. And when they look at the data, and it looks like it’s not, they immediately rush, very often rush to, “Oh, your copy is not working.” But so many other things could have changed. Or they might be looking at too little data, or they’re looking a week after it launched and you need way more time to understand how this copy is going to perform, especially when you’re not even A/B testing. You’re just looking at like before and afters or they’re looking at too few people.

Those are the four things to keep in mind. Now, what do you do to make sure that you don’t end up in a situation where you’re just dealing with all of the aftermath of clients saying your copy is not working? How do you first make sure that you set yourself up to avoid that?

First, we want to ask questions around like, “Okay, have we, or how will we collect enough data to help us make decision?” Then once they ask, once they say, “Okay, look, here’s the data,” then we can ask how we collected enough of it that would actually help us make a decision and enough from enough sources as well. Are we believing enough of this? Have we looked at the right data sets to help us make a decision? Are we looking at the right segments of traffic? Do we need to segment out any weird traffic that came in from some PR thing that was happening or some event that you hosted where you would have had a lot of Lucky Lukes coming by your site and negatively impacting the quality of the traffic compared to what they were looking at before?

So we want to ask these questions, and then make sure that we’re building up before we start a project with the right questions out of the gate, to make sure you’re aligned with your client. This is important. I know you’ll want to skip this. When you skip this, it hurts you every time.

So when you first start working with a client, ask that client, when you’re working with them on web copy in particular, like sales pages, landing pages, and actual websites, so not as much for email of course, but ask the client to walk you through Google Analytics for the page or pages you’re going to be optimizing. This could be part of that initial kickoff call. Just have them open up Google Analytics and share their screen. And you start asking them questions about it. Like, just show me your funnel. Show me the pages that we’ll be optimizing, et cetera, et cetera.

And if you’re creating something brand new, ask them about pages that are similar to the one that you are looking to write. So if you’re going to write a product page for a new product, and they already have product pages for old products, if they have product pages for old products that are similar to what you’re going to be writing, maybe it’s for the same audience is what you’re going to be writing for, or the same price point, et cetera, et cetera. Then I have them walk you through Google Analytics for those. Again, you don’t have to be a Google Analytics pro. You have to be competent asking questions.

The more that you write copy and the more you have clients challenge you on your copy, the better you’ll be able to start asking these questions. And that’s why we do these tutorials because you might not have gone down this path before where a client has actually asked you anything and shown you Google Analytics. But once they do, you’re going to want to make sure that you’re in a good place to talk to them and ask them good questions.

Asked to be walked through the whole funnel your work is going to be involved in. So if you are writing emails, ask them in Google Analytics to show you the funnel, so you can really see. And then document it. Any writer that I work with who doesn’t have a pen in their hand all the time, makes me crazy. Like you should be writing down everything all the time. Document it. Document it. Document it. It will help you write better proposals. If this is in the early stages, it will help you go back and do better postmortems or retrospectives as well.

So make sure that you ask, you walk through the funnel, that your work is going to be part of for the last month, the last quarter, and the last year. Then when you look at that, you write down the numbers, and you ask them about the traffic sources and what they believe was going on at those times, et cetera. That’s really important for you when it comes time to action, if somebody say your copy isn’t working and have you be able to have a discussion where maybe it’s not, but maybe it is, and they’re looking at things wrong.

Ask about potential sources of variation. Again, is there seasonality? If you work with finance, if you work with taxes, if you work in government or education, there’s definitely seasonality with all of these things, and also with their audience. If they’re selling a very low priced product or they’re selling to a market that doesn’t have great cashflow, then that market or market segment is going to have more money on paydays. So they’re going to have more money at the start of a month and on the 15th. And that’s just the way it is.

So if you’re running your test or you’re putting your copy out there from the 17th to the 27th, and the last time they ran it, they did from the 14th to the 24th or whatever, that would’ve had probably more people paying upfront during when they got paid on the 15th. So make sure that you are really asking these important questions so that your copy is understood best afterward, and you can write the best copy possible and even optimize the offer.

Agree up front on what measuring success is going to look like and what it won’t look like. So be careful to get into … Sometimes people don’t want to wait for statistical significance and sometimes you can’t because there’s not going to be enough traffic to a page. But people will still want to make a call on how the copy is working, even without enough data to really say anything.

So you want to make sure that you say, “Okay, we’re not going to ever make the call on how this page is converting based on the data. What we will use statistically not significant data for is X, Y, and Z.” And that might be, I don’t know, something else where you feel like, “Okay, I can get behind that.” But when it comes to conversion rate or to actually saying something about the copy, you do not want to say anything about your copy if there’s not actually good data to support that. And they shouldn’t want to either because they could either eliminate your great copy with not enough data, or make it sound like it’s great copy, even though you didn’t have enough data to really say anything about it.

And then of course set expectations for when you’ll check in on performance and what you’ll be looking for. And that’s things like we’re going to look in Google Analytics at month over month and year over year and quarter over quarter, et cetera, et cetera.

Okay. That is it. So it’s a bit of a crash course in how to ask questions of the data that is criticizing your copy. Hopefully you have found that valuable. That wraps it up for today’s Tutorial Tuesday. Thanks guys for your participation on this data e-topic.

Speaker 2: It’s

[inaudible 00:17:59]

Get our new articles, videos and event info.

Join 89,000+ fine folks. Stay as long as you'd like. Unsubscribe anytime.