Many stores under-invest in website usability improvements. Once you have the basics in place, it’s hard to know what usability improvements will move the deal on conversions and sales. In short, you run out of ideas and stop focusing on it.

But with a few small tweaks to your customer service workflow, you can have a never-ending stream of extremely high-impact usability improvements for your team to execute on. The live chat and customer service tickets your team are already working on can be a gold mine of improvement ideas.

At HelpFlow, we provide 24/7 live chat and customer service teams to hundreds of e-commerce stores. In addition to driving conversions with live chat and running your customer service operation, a big value we also provide is helping you improve your business with insights that come from customer service.

In this Partner Thursday post, I will walk you through a simple approach you can use to get ideas to improve sales on your website directly from your customer service communication.

Categorize Pre-sale and Post Sale Conversations

This whole process is going to rely on adding a bit of structure to your customer service conversations so you can analyze them and get insights. The first step you need to do is categorize pre-sale versus post sale conversations.

In addition to having your chat and email, agents quickly categorize with the keyboard shortcut as part of their workflow. You can also automate some of this.

For example, if your chat and help desk automatically sync with Shopify order data for a customer, you should be able to automatically categorize post-sale questions.

Once you have these categorized, you can start to focus deeper analysis on the pre-sales questions. That’s where the big revenue-driving website changes are going to come from.

Tag Conversations Based on Issue Types

This is where you start to add more granularity to the categorization of tickets. Specifically, with pre-sales tickets, the goal here is to tag these conversations as high-level issue types.

We recommend the following to keep the types insightful but not too granular to where your agents will not do the extra work to apply a tag.

Product Recommendation

  • Navigation – this is any question related to somebody trying to find something they are looking for on the website.
  • Policy – this is any question about how you handle certain scenarios for a customer, such as shipping time, return policy, etc.
  • Other – Having your chat and email team do this as part of the workflow will only take about five more seconds per ticket if you have keyboard shortcuts and automations being used. Requiring the team to do this even over a short period of a few weeks gives you a lot of powerful data to analyze in the next step.

Bulk Review Issue Types and Spot Patterns

Once you have tickets tagged, you or someone on your team should spend time reviewing them in bulk. Specifically, dig into individual types and review multiple tickets in big batches.

As you do this, you’ll start to see patterns in those tickets. For example, with one client we worked with that sells video gaming headsets, there was a very clear pattern to a few specific product attributes people asked about to compare the products to each other. This was very clear when reviewing multiple chats at the same time.

During the review, call out a specific type of action that could resolve the issue. In the example above, changing the category page to focus on these attributes might be the action. For policy questions about shipping time, the action might be to better emphasize shipping times on product pages.

Try not to be too specific on the exact action item for each ticket. If you are reviewing chat and ticket conversations in bulk, you’ll start to see a lot of overlap in these action items. From there, you will see a pattern of what specific action items make sense to attack next based on how often they come up.

Create an Improvement Backlog, Prioritize Methodically, and Execute

Once you have built a list of potential actions to take to overcome issues you see in the chat and email ticket conversations, you can organize them into a backlog. This is basically a simple list that keeps things organized, enables you to prioritize like I will share below, and used by your team to pull from for future work.

A simple but effective way to prioritize any backlog is using the ICE framework. Rate each item 1-5 based on:

  • Impact – how big of an impact could this possibly have once implemented.
  • Confidence – how confident are you that this impact will happen.
  • Ease – how easy is this to get implemented re cost, time, etc.

Average out these numbers to assign a single score to each item on the backlog. From there, the highest scored items are the ones you should definitely consider implementing first.

This puts your business in a position where you have a never-ending stream of potential ways to improve the effectiveness of your site, all derived from actual customer conversations. Rather than doing ad hoc brainstorming sessions a few times per year, this can be a systematic approach that enables you to constantly move the needle with improvements.

Like I mentioned, we provide 24/7 live chat and customer service teams to over 100 stores. I’m happy to nerd out in the comments on the specifics of how to integrate this into your workflow with your team.

Jon Tucker


Jon Tucker is CEO of HelpFlow, a provider of 24/7 live chat and customer service teams for 100+ eCommerce stores. Launched in 2015, their team of agents has produced nearly $100M in chat revenue for a wide range of stores using a conversion-focused approach to live chat.


Your Store Needs a 24/7 Live Chat Team


HelpFlow provides 24/7 live chat teams to 100+ eCommerce stores.  See how we can help you save checkout abandons, increase conversion rate, and ultimately drive more sales while providing an awesome customer experience.