Ecommerce Copywriting: How to Conduct Copy Research to Boost Sales

Ecommerce copywriting comes in many forms, but is rarely discussed outside of product descriptions.

Value propositions, microcopy, Facebook ads, promotional emails—all of these rely on copy to capture attention and convert visitors into buyers. Yet, many ecommerce sites are “just winging it” when it comes to copywriting. 

It's a shame because more persuasive copy is the key to boosting sales without investing more in acquisition.

The good news is that there’s a 4-step process that professional copywriters use to craft persuasive copy and improve conversions—one that you can steal and use for yourself.

Step 1: Define your audience and segments

Before you dive into copy research, you need to define your parameters. Start by asking yourself what you want to learn. The answer will help you decide how to define your audience and segments for stronger insights.

Here are some common segments you might want to explore: 

• Abandoned Carts: you’ll identify pre-conversion friction (anxieties, fears, frustrations, etc.) that prevent visitors from buying. Remember, cart abandonment isn’t normal, just normalized. People don’t leave full carts for no reason.
• New Customers: you’ll identify more of that pre-conversion friction. What almost prevented them from buying? Why did they choose you over competitors? What was frustrating during checkout? Plus, you’ll learn about product quality and understand how well you deliver on your value proposition.
• Repeat Customers: you’ll get to the heart of what products pair well together, how long the buying cycle is, what the customer lifecycle looks like, etc.
• Inactive Customers: you’ll get to the bottom of lifetime value (which can help with paid ad spend planning) and retention. How many purchases did they make total? Why did they stop purchasing from you? What could you have done better?

These are general segments that could apply to any store. You might want to get more specific. For example, isolate customers based on product categories or new customers who purchased from you twice in 6 months.

Just be sure to start with a goal or a list of questions you’d like to answer. Then, work backwards.

So, if you’re looking to understand lifetime value (LTV), you don’t want to be talking to new customers. Likewise, if you’re looking to better understand that initial purchase process, you don’t want to be talking to people who have been buying from you for years.

Make sure the segment(s) you’re targeting are in a position to help you answer the questions you have. They have to be at the right stage of the buying cycle, they need the proper pain/product awareness, etc.

Step 2: Conduct qualitative research

When you know what you want to know and which segments can help you find that out, you’re ready to start diving into qualitative research. Typically, that has four elements when it comes to copywriting: internal interviews, customer interviews, surveys, and testimonial/review mining.

Joel Klettke of Business Casual Copywriting and Case Study Buddy explains why qualitative research is so important:

If there's one thing most companies miss, overlook, or ignore, it's that every single conversion is the result of a conversation your lead is having with your copy.

With qualitative research, you have a chance to look at the answers before you take the test by asking the questions you know your leads are coming into your site asking. You can take their answers, and then turn around and bake them right into your copy, in your customers' own words.

I know of no other factor that makes a bigger difference to the results of your copy than the quality and depth of research you conduct.”

1. Internal Interviews

For the sake of simplicity, I’m going to imagine you’re trying to understand how you can increase initial purchases on your site for the remainder of this article.

Before you talk to your visitors and customers, it helps to chat with internal employees who deal with customers every day. For most stores, that’s sales and support. This can help you understand what’s working and what’s not early on, allowing you to ask customers more insightful questions.

There’s a good chance you are your own sales and support team. If that’s the case, ask yourself the following questions:

• What questions are most commonly asked by visitors?
• What frustrations are vocalized most often by visitors?
• What pains do visitors arrive at my site looking to solve?
• What benefits do visitors arrive at my site looking to gain?
• What objections to buying do visitors have?
• How do I address those questions and objections successfully when I encounter them?

It helps to check support logs during this process. Otherwise, you might end up with bias answers. Go over the logs from the last 3-6 months, depending on volume. Highlight recurring questions, pains, benefits, objections, frustrations, etc. 

If you’re not your own sales and support team (perhaps you outsource support or have a freelance pay per click pro), take the time to talk to those front-line people in person. Ask them the same questions and, again, look for recurring themes.

James E. Turner of SNAP Copy recommends that you:

Get 'em talking for a while, to get past the ‘best business answer’ period and into the ‘but really, this is how it is’ phase.

2. Surveys

There are two types of surveys you can use in the copywriting realm…

1. On-Site Surveys: exit or intent questions that popup automatically at high-value moments.
2. Customer Surveys: questions that go out via email to the segments you’re looking to better understand.

Both can help you better understand that initial purchase decision. 

A great survey really, truly depends on your unique objective/questions. However, there are some best practices to keep in mind when designing your copywriting surveys. 

1. Less is more: the more questions you ask, the fewer responses you’ll get. I recommend aiming for 3-5 questions for customer surveys and 1-2 questions for on-site surveys. As an internal filter, ask yourself what you’ll do with the answer to the question. How will it help with your objective? If it won’t, don’t ask it.

2. Yes/no and multiple choice questions are lower value: yes/no and multiple choice questions are not useless for writing copy. Quantitative data can help you identify problem areas, too. However, you’re looking for the voice of the customer to guide your copy, so open-ended questions will be higher value.

Jen Havice of Make Mention Media offers a word of caution, though:

Getting answers to why your customers are seeking your products out and the information they need to see on your website to take action is critical to writing more effective copy.

They key is to ask questions that relate to their actual behavior versus stated preference. This means asking the questions about what they actually did, not what they want to do.

Why is this important? Because stated preferences are notoriously unreliable. People will say they want one thing but then do the opposite when going to buy.

So it's best to ask questions like: What was happening in your life that caused you to start using our product? or What alternatives did you consider before using our product? That way you'll have a much better idea what's driving your customers to or away from you - making your messaging a whole lot stronger.”

As an exception to the rule, it can be useful to ask a yes/no or multiple choice question initially on on-site surveys, followed quickly by an open-ended explanation.

This is the foot-in-the-door method. You have less goodwill with visitors, so you might see a lower response rate for on-site surveys. To combat this, you ask a simple yes/no question to get your foot in the door and then the visitor feels somewhat obligated to comply with the rest of your request.

3. Be aware of bias that can creep in: you are a very bias person by nature. The more you doubt that, the more true it is (it’s called bias blind spot).

Those biases can easily creep into your survey questions and analysis. For example…

• Experimenter Bias: when your personal experiences make objectivity very difficult. For example, you might unknowingly communicate what you expect the results to be to those you survey, perhaps through a leading question.
• Confirmation Bias: you interpret data in a way that confirms your pre-existing beliefs or your hypothesis while disregarding or downgrading data that does not. For example, when you’re analyzing the results of your survey, you might unknowingly overlook the data that disproves your hypothesis, only highlighting the data that supports it.
• Curse of Knowledge: once you know something well, you find it very difficult to think about a related problem or situation from a lesser-informed person’s point of view. For example, you interact with your store every day, so it can be difficult for you to notice the user experience (UX) problems first-time visitors face.
• Selection Bias: if you don’t choose a representative sample, true randomization can’t be achieved, which leads to bias results. For example, you only survey 15 people. (More on this in a bit.)

The more aware you are of the potential biases that could be affecting your surveys, the better. You can’t eliminate all of your biases, but you can mitigate them through awareness and preventative measures.

One common way bias sneaks into surveys is through leading questions. A leading question suggests the answer to the respondent. So, for example…

Where do you like drinking beer?

...is a leading question. As is…

What do you love about your stylish new tee?

4. Relevancy is king: this is made easier by the fact that you’ve already defined your audience and segments, but it is particularly important for on-site surveys. Your goal is to ask the right question to the right person at the right time.

So, for example, you might ask someone showing exit intent what prevented them from purchasing today via an on-site survey.

5. Simplicity is queen: phrase your questions as simply as possible. The less your respondents have to think about your questions, the better. All they should have to think about is their answers. That means short sentences, eliminating jargon, focusing on clarity, being specific, etc.

Be clear, be brief, be simple.

6. Sample size and representativeness matter: as you now know, selection bias is very real. It can be very damaging as well.

If your sample is too small or inaccurately represents the entire audience/segment, you have inaccurate results.

For example, only collecting 10-15 survey responses is not enough to accurately generalize the results. Similarly, if you survey people who only purchase from one product category, you can’t generalize the results to those who purchase in other product categories.

Typically, you’ll want to collect about 250 survey responses. If you collect less, you won’t be able to accurately notice trends and patterns. If you collect more, you’ll likely just end up unnecessarily investing more time into analysis.

For representativeness, be sure the sample selection is truly randomized and representative of the entire audience/segment.

7. Design your surveys with open rates and response rates in mind: Lianna Patch of Punchline Conversion Copywriting explains it best:

No one gets a survey email and is like ‘OH BOY, LET ME GET RIGHT ON THIS’ (unless they have an axe to grind). So you have to entice users not only to open the email—that’s your subject line’s one job—but to click through to the survey and complete it. These are all individual wins, but none really matter unless you get that survey response.

Your survey email(s) should be energetic and emphasize the end benefit for the reader. It’s not ‘We’re trying to improve our services,’ it’s ‘Tell us what YOU really want so we can give it to you.’ If you can, be explicit about how you’ll use the data: ‘Your response will help create a better version of [Our Product/Service] in [This Particular Way].’

Hopefully you’ve already trimmed your survey itself down to no more than 3-ish questions. In your email, emphasize how short the survey is, how little time it will take, and how grateful you are in advance that your user will help you out. Be careful of offering rewards for survey completion, since you might skew your data by filling it with people who aren’t really your target user; they just love Amazon gift cards.

And obviously, never bait and switch users. I recently got a survey email that asked for 2 minutes of my time. Know how many questions were on that survey? 60. SIXTY! Most of which were open-ended. Smdh.”

3. Customer interviews

Customer interviews are valuable to the copywriting process at any stage of growth, but they’re especially useful when you’re small. If you’re a brand new store owner, it’ll take quite a while to collect those 250 customer survey responses, right?

Again, don’t run out and start interviewing anyone who says “yes”. You want to spend your time wisely, only interviewing those who can offer you the most insight. Stay within the segments you’ve selected and screen for specifics (e.g. purchases, purchase frequency, demographic, etc.)

The questions you ask during an interview are important, so spend time choosing them wisely.

Try to strike a balance between demonstration (“Show me how you would…”), tasks (“Find a pair of skinny jeans for $90 or less.”) and behavioral (“What was happening in your life that caused you to start using this product?”). Treat customer interviews as a cross between user testing and surveys.

It’s cliche, but using the good ol’ who, what, when, where, why and how still works.

Kira Hug, co-founder of The Copywriter Club, shares her interview process:

The most powerful research tool I use to write conversion copy (AKA sales emails and landing pages) is as basic as a 20-minute phone interview. While I find a lot of value in surveys - because you can collect a ton of data from hundreds (or even thousands) of people - I find nothing beats two people chatting on the phone.

Typically, I follow every survey with at least 8-10 customer interviews. This gives me a chance to go deeper into an individual's story, challenges, desires, goals, objections and more. It's amazing what a stranger will share with you in only 20 minutes, when you ask the right questions.”

Whenever possible, conduct these interviews via Skype or Google Hangouts. Phone interviews are useful, but they eliminate some context because you can’t read body language, which is helpful in tests like a contextual inquiry. Record the session so you don’t have to worry about taking notes, which can take away from your interview flow.

Then organize and cluster responses into your spreadsheet:

You’ll get better at customer interviews over time, so don’t worry if you don’t knock it out of the park the first time. As you continue, you’ll learn how best to communicate with different types of people, how to ask smarter questions, etc.

4. Testimonial and review mining

Third-party sites are full of testimonials, reviews, complaints, etc. that you can tap into. They’re usually less bias than if they were solicited. It’s just a lot of voice of customer goodness to make your copy stronger and more persuasive.

A quick Google search can tell you how you’re perceived and specifically what people are saying about your site/products.

If your site is new, this is perhaps less true. If you’re doing dropshipping, though, you can get copy cues from reviews of the products you’re selling. If you’re not, consider where people might be reviewing your products locally (e.g. Facebook, Twitter, online creative/craft groups, etc.)

James offers a word of advice on testimonial/review mining, encouraging you to lean into the negative feedback:

Make sure you take a look at both positive and negative reviews - and if you only have a chance to look at one type, go for the negative. This is where you'll find the angst/pain that drives people to frustration.

Particularly if you're in a crowded, recurring-purchase type of market, if you can point out that, unlike your competitors' faulty versions of your product, yours doesn't wear out after one use, or leave sticky marks on your baby's hair, or explode in your pocket, then you're away to the races, and can lead with that.”

Step 3: Identify and document patterns

Now you’re looking at a spreadsheet with a lot of data from a lot of different sources. It can feel overwhelming, but rest assured your job is only half done. Next, you’ll need to dive into the data and begin identifying patterns.

At this stage, you’re looking for: 

• Words and phrases that stand out to you, that were particularly memorable or often repeated.
• Objections, products, benefits, questions, pain points, points of friction on the site, etc. that were often repeated. 

Of course, you’re also looking to understand the way the segment speaks and the words/phrases they use. This will help you write the way the audience speaks, in words and phrases they identify with.

Be careful not to accidentally insert your assumptions into the research. Simply look at what you learned from your research, what your audience told you and spin that into persuasive copy.

It can be helpful to take data from the spreadsheet and organize it based on the exact page you’re writing copy for. For example, a product page:

(Note that if you uncovered any on-site points of friction during your research, you can go ahead and implement UX fixes at this stage.)

Step 4: Define the messaging hierarchy and wireframe

Whether you’re crafting copy for a product page or a Facebook ad, you now have everything you need to write data-informed, customer-driven copy that converts. Now all you have to do is turn the data into solutions. 

That starts with defining the messaging hierarchy, which helps visualize the importance of each of the messages you’re trying to communicate. The more frequently a pain point or benefit or question came up during your research, the higher it should be on your messaging hierarchy.

Once you have that high-level concept, you can start building a wireframe. (Even if you do your own site design and paid ads, this can be a helpful exercise.) For example:

Joel has given an entire presentation on how to create wireframes using Balsamiq. Be sure to check that out. (He’ll explain why my placeholder copy in the wireframe above is a big no-no.)

All that’s left now is to turn that wireframe into reality and start reaping the rewards of data-informed, customer-driven copy that’s crafted to convert. Remember, you can apply this copywriting process to any medium (email, on-site, paid, etc.) 

FREE TEMPLATE: Grab the spreadsheet we’ve been using throughout the article here.

---

Haven't created your store yet?

No problem. Start your free 14-day trial of Shopify—no credit card required!

Email addressStart free trial