Best spas in Austin: big data reveals the top 10 most-loved locations

Blog • Data Science

If you’re looking to get pampered in Austin, there’s no shortage of places. A quick search for best Austin spas on Yelp gives you 2,304 choices ranging from massage parlors to beauty salons to acupuncture and yoga studios and every combination in between. Narrowing it down to their “day spa” category helps, but you’re still left sifting through 236 options. Inevitably, you start reading reviews because star ratings and review counts only go so far. But who has time to read and interpret all 6,488 reviews to determine the best day spa?

Well, we know a guy...

Say hello to VADER

(If you’ve already read our post on using VADER to determine the best coffee shops in Austin and/or the people’s choice for best picture, you can skip this section.)

VADER (Valence Aware Dictionary and Sentiment Reasoner) is a machine learning tool we’ve been using recently for a number of clients to conduct sentiment analysis. What’s sentiment analysis? Well, technically, it’s a way of turning qualitative data (reviews) into quantitative data (sentiment scores) to measure the extent to which a reviewer is writing from a positive or negative emotional state, as well as the extent to which the words they use are positive or negative.

In layman's terms: it’s a way to tell how positive or negative a review is, and how passionate that person is.

The exciting thing is that VADER learned to judge sentiment from ALL OF TWITTER as well as a large number of human raters. This helps eliminate bias by relying on “the wisdom of the crowd”: collective opinion is more trustworthy than individual opinion. In fact, in a large study, VADER not only out-performed 11 other models, but it even beat individual human raters. Oh, and VADER can be deployed to analyze massive amounts of data really quickly.

First, let’s take a look at Yelp

Like most people on a quest to find the best spa in Austin, we started out by typing in “best spas” on Yelp. But we’re sorry, Yelp, a yoga studio is not a spa. So, we filtered by their “day spa” category, because these businesses tend to provide a similar suite of relaxation services typically associated with a “day at the spa.” Here’s what Yelp gives for the top 10:

If you’re trying to make a quick judgment by rating or popularity, you’re stuck: seven out of the 10 have the same number of stars, and six have 100+ reviews. Thankfully, we have some handy Python data-mining scripts and VADER to help out.

How we used VADER to find the best spas in Austin

The first step was to scrape all 6,488 reviews and collect them in a single sheet. Then, we removed any location with less than 50 reviews, because it’s not fair to compare a place with 4 reviews against one with 400, and VADER works best with larger sample sizes. That left 48 spas to feed VADER.

Each review was analyzed and assigned sentiment scores. The outputs of this technique are 1) a measure of sentiment intensity of the comment with a range from -100% to +100%; and 2) a percentage breakdown of the actual sentiment based on the words used (positive, neutral, or negative).

For example, “BEST SPA EVER!!” scores high on positive intensity, because VADER takes into account things like capitalization and punctuation. However, only 33% of the words in the review are strictly positive (“best”). On the other hand, “Love Viva—beautiful, amazing” is less intense, but has a higher proportion of positive words (“love,” “beautiful,” and “amazing”).

As you can see, VADER is great at analyzing the data, but we’ve developed a method that helps us assess the data.

Putting it all together

After VADER has taken apart the data, we put it back together in a new way that allows us to see the value of a review. We multiply the positive intensity by the percent positive words, and the negative intensity by the percent negative words, to derive a “Passion Score” (PS). The PS—on a scale of -100 to +100 with 100 being the most positively passionate—gives us a holistic idea of the true sentiment behind a review, giving equal weight to the words a person is using AND how passionate they are about what they’re saying.

Using PS, here are the best, most-loved Austin spas:

Important notes: Steven Todd Salon does not offer massages; all the rest do. Yen Nail N’ Spa actually had the highest passion score of 25.5, but we omitted it because it’s in Georgetown. Similarly, Spa Social (PS score 24.0) and Pfinesse Salon & Spa (23.5) were removed because Round Rock and Pflugerville aren’t Austin, either. Bee Cave, while similarly “out of town,” at least still has an Austin address.

But wait, there’s more!

Averages are helpful, but they rarely tell the full story. And do people really feel exactly the same about Hiatus Spa and Steven Todd Salon? When you visualize all of the passion scores for each individual review, you can learn a lot more. Take a look at this lovely box and whiskers plot: 

What’s a box and whiskers plot? It’s a way to visualize the distribution of a dataset. It shows the average of a dataset (the tall line in the middle), where 50% of the data is concentrated (the box), and the bounds for any outliers (the whiskers to the left and right). You can think of it as a tug of war of passion, with negative people on the left and positive people on the right. The more spread out the data is, the more opposing opinions there are.

For example, Hiatus Spa and Steven Todd Salon may have the same average passion score, but Hiatus has a slightly greater concentration of positive reviewers (that’s why their “box” is more to the right). That said, they also have more negative outliers (that’s why their left “whisker” is more to the left).

And the winner is...

Congratulations, Massage Sway! Based on our analysis and assessment of nearly 7,000 Yelp reviews, you have the most positively passionate fans.

Want to learn more about this project or have an idea about sentiment analysis applied to your work? Contact us

More Like This