How to Improve Landing Page Conversions with Live Chat Transcripts

Written by Guest Blogger on July 14, 2015

Posted in Form Optimization

If your website landing pages are suffering from low conversion rates, you’re not getting visitors to complete your forms, and you need advice on what really works—so you can start measuring, learning, and improving your landing pages today—you’re in the right place.

Whether you own an ecommerce store, a SaaS product, or a blog, we’ve got you covered.

What is Qualitative Data Collection?

To improve your landing page conversions, you need to follow a process for conversion optimization. The process allows you to learn from your current website and traffic and make changes based on data—not guessing or the highest paid person’s opinion. Data!

The process is as follows:


The first step is to set your business objectives, goals, KPIs, and targets.

Next up is data collection. When you think of data collection, you think digital analytics (like Google Analytics). This is quantitative data—what happened where and when. This is just half the story.

The other side of the story is qualitative data. This is an exploration of user behaviour, which sounds very much like user experience; and it is! Qualitative data is the why: the motivations, the pain points, and the human emotions and behaviours that cause someone to convert on your landing page from a visitor to a customer.

Live chat transcripts are a form of qualitative data. They allow you to learn visitor reservations at the point of sale and what information is missing from your landing pages, and they help to reveal your visitors’ true motivation around spending money with you. In short, live chat can help you learn more about your customer and improve your marketing messaging. This, in turn, increases your conversions and—more importantly—profits.

Enough of the theory, let’s get practical and actionable!

Setting up Live Chat Qualitative Data Collection

I recommend Olark for live chat software. The customer service is second to none, and the product is a breeze to set up and use.


Head over to and simply follow the instructions to set up an account.

Why You Need a Live Chat Script

Aside from answering your visitors’ questions and queries, I advise you to set up a live chat script for your operators to follow. This allows your additional visitor questioning to be focused and makes the data you collect more actionable. The script should correspond with a spreadsheet where you can record your live chat transcripts and analyse them.

Make sure your script focuses on learning the following information from the visitor:

  • Name
  • Sex
  • Product of interest (if you have more than one)
  • Category of interest (e.g., price or feature)

A Framework for Analysing Your Live Chat Data for Visitor Insights

Here is a template spreadsheet I have created to help with your live chat data analysis. Let’s walk through the different columns and learn what they are for and how to fill them out.

Chat ID

This is the ID of the chat transcript in Olark. It is important to record this so you know in the future which chat transcripts have been completed.

Name, Age, Sex

These fields are self explanatory. They require you to read through the transcript once and see if any of this information is revealed by the user.


If you are an ecommerce store or a company with multiple products or services, note here which product or service is being discussed in the transcript.

Category of Interest

This section allows you to categorize and divide the chat transcripts into groups. The categories could include things such as features discussed, order details (e.g., delivery times), or even worries or reservations users have about making the purchase.


This is where you write your impression of the whole transcript and what you feel is the biggest takeaway from the chat. Focus on customer words that express a pain point, an information gap in the landing page, or a desire or goal the customer wants to achieve.

To Do

This is for any actionable takeaways from the chat transcript. For example, note if someone complained about a bug or if his or her feedback needs to be added to your customer feature request list.

Real Quotes

After reading the whole transcript, write down one or two of the most important sentences the user said, in his or her exact words. It is really important to use the customer’s real words as you will collect all the data at the end of the process for another analysis exercise. Again, focus on sentences that express a customer’s pain points, information gaps, or desires or goals.

A Step-by-Step Process for Analysing Live Chat Data

Here is a simple step-by-step process to follow to analyse your chat data.

Step 1:

Log into Olark and jump into the transcripts tab.


Step 2:

Next, open up your spreadsheet from the link above and duplicate it into your Google Drive.

Step 3:

Open the latest chat transcript and begin to note the details as explained for each column above.

Step 4:

Once you have completed all the chat transcripts or a batch of at least twenty, tally up all the categories and find the top three to learn what customers are focusing on the most. The easiest way to do this is to search for a category and see how many results there are.

Step 5:

Select and copy all the entries from the “Real Quote” column, then navigate to this website. Paste all the entries into the text area field at the bottom of the page, then click “Go!” This will create a word cloud from all the customer quotes. Screenshot the word cloud and save it.

Step 6:

Look at the word cloud. What are the three most prominent and insightful words on the word cloud? Make a list of the top words you feel are insightful or point toward a pain, reservation, or desire.

Below is the word cloud I created using the “Real Quote” column from some sample data from an online technology course sales page.


Let’s interpret the word cloud together. You can ignore words like mentor, course, UX, design, and enroll since those are part of the product’s natural language and don’t tell us anything about how the customer feels.

We knew from other studies that prospects were worried about time (specifically, start times). This is echoed and validated with the presence of the words start, started, and days.

We also learned prospects were worried about their capabilities and what support they would get. This, again, is seen in the word cloud with the presence of the words consider, able, and support.

Word clouds can give great insights into the key feelings your customers have, but you must look for patterns not only in single data sources but across your data.

Once you’ve determined the keywords from the cloud, make those words into a list. This list of customer words can then be used as input into your sales and marketing copywriting. You want to ensure you use the same language and exact words your customers use to express their problems and goals when trying to optimize your landing page conversion rates and your profits.

How to Apply Your Learnings to Your Landing Page for Increased Conversions

From your data analysis, you should have three big takeaways:

  1. The most recurring reservations or missing information from your landing pages as mentioned by visitors
  2. A list of “Real Quotes” your visitors typed
  3. A list of visitor vocabulary

Use these three takeaways as an input to your copywriting process. Attempt to remove the reservations visitors have by plugging any information gaps into your landing page copy. Use real visitor quotes and vocabulary to improve your call-to-action button copy and section headers so your messaging is more aligned with how your visitors talk and think.

You Can Improve Your Landing Page Conversions…

…But you have to take action.

You must put into practice the theory and processes you learned here today. Listening to your visitors and learning about your customers is just the start. You have to test and iterate on your landing page designs to increase your conversion rates.

With this in mind, I’ve put together a bonus download for those of you who are serious about increasing your conversion rates. What you get:

  1. Illustrated list of 25 CRO best practices to test in your next redesign
  2. Live chat data analysis pack (spreadsheet template and process PDF)
  3. BONUS: The bonus content from all my blog posts!

Click here to access the free bonus area.

About the Author

Giles Thomas writes about ecommerce marketing at Acquireconvert and He is the founder of Shopify marketing agency Whole Design Studios. Connect with him on twitter @acquireconvert.


Want more tips on creating high-converting landing pages?

Click below to check out our webinar “The Dos and Don’ts of Landing Page Optimization” with Chris Goward and Jeff Blettner.