From Keywords to Context: Comparing Customer Sentiment Between Top Website Builders

Purpose of Analysis

Put yourself in the shoes of someone who’s just decided to create their first website — maybe for a business, maybe for a blog — doesn’t really matter.

Most people aren’t going to sift through thousands of reviews to compare and contrast different website builders — and you aren’t one of them either. You’ll probably scroll through a few articles that do the comparing for you — all of which tell you different things — or you’ll take a quick look at a random review site and just end up choosing the brand with the highest rating.

Both methods are clearly bad ones.

The same kind of logic applies if you’re the company sitting on the other side. You may be able to observe feedback at a large scale — see how many customers give you 5 stars, 4 stars, 3 stars, or calculate how many keywords (ex. “technical issues”, “customer service”) are mentioned — but accurately breaking down reviews into categories and sentiment is the hard (and way more important) part.

But we did the work so you don’t have to — here’s an example of a quick analysis we did on public reviews for Squarespace, Wix, Hostinger, and GoDaddy, using Cotera’s sentiment analysis program.

The Analysis

The first thing we did was pretty basic - we simply split thousands of reviews for each of these 4 website builders by positive and negative sentiment.

You’ll see that this doesn’t tell us much. There are pretty minimal differences between each website builder in this regard, and it doesn’t tell you anything about what each company does well and what each does poorly.

So we took things a step further by using our program to categorize every review into overlapping topics. From this, we were able to calculate the proportion of positive vs negative reviews across 4 different categories for each brand.

Here, you can see that Squarespace had the lowest amount of negative sentiment in categories like customer service and technical experience, but ranked the worst in terms of user friendliness. Wix and Hostinger appeared to be pretty well rounded in every category — though both companies seemed to be hovering only slightly above 50% positive sentiment in nearly every category. Interestingly enough, GoDaddy users seemed to be the happiest with user experience.

Meaningfulness of Results

Doing an analysis like this isn’t impossible to do on your own, but like we said before, it’s accuracy that’s the real challenge. And the more data you have, the harder it is to do manually.

There’s always the easy route, right? If you want to see how many people are complaining about customer service, the quick solution would be to do something similar to “command F” for keywords like customer service, customer support, support, help, etc. The problem here is that you’ll probably end up with either a) a good chunk of reviews that actually end up being irrelevant, or b) only half of the full picture.

For example, it’d probably be pretty tough to use this method to identify and categorize a review like “Tim answered all our questions and concerns accurately and quickly.”

What’s even more useful than looking at this kind of data at a single point in time is looking at how it changes over time, or even receiving alerts when a new negative review is posted in a certain category (something Cotera’s known for helping companies do). It gives you a chance to resolve issues immediately before things escalate — which will no doubt lower churn rates in the long run.

‍