Analyzing Analysis: Bloomberg botches butt surgery analysis

This week, Bloomberg published an article (paywall alert) claiming that “data show demand for butt implants soared during the pandemic.” It’s an eye-catching idea but one that we think doesn’t hold up. Basically, their butt analysis is bupkis. And, unfortunately, they missed the opportunity to find meaning in data.

In our Making Better Decisions workshops, we apply a decision-making framework with starts with a series of of questions to understand the situation:

  • What do we want to know?
  • What does this mean?
  • Have we reached minimal viable meaning?
  • If not, what else do we want to know?

We’ll apply this framework to the Bloomberg analysis:

What’s the source of the data?

The American Society of Plastic Surgeons conducts an annual survey of plastic surgeons to estimate the number of procedures completed each year. The survey methodology for 2020 hasn’t been released but in previous years, the survey was sent out to 20,000-25,000 active physicians, 500-600 surveys were returned and 700-800 physicians included in the final sample (it’s unclear how those extra 200 physicians are included). The survey firm calculates this as a 95% confidence level with 4%-5% margin of error. Importantly, the 2019 survey has been removed and been replaced with a note saying that “the 2019 procedure statistics projections have been updated as of April 27, 2021, to reflect a more comprehensive set of board-certified physicians performing these procedures.”

What does this mean?

Surgeons are likely a decent source of information on their own practices. But this is a survey that will include inaccuracies based on biases and bad memory. Small sample sizes are not necessarily bad but they can include data variances, especially when dealing with small numbers. We don’t understand the change in 2019 survey procedures and are left wondering how valid year-over-year comparisons are.

Have we reached minimal viable meaning?

No, we’d like to know how those extra 200 physicians are included and whether the respondents vary in terms of region, practice area, etc. in a way that could affect the overall outcome. We also like to understand the change in methodology in 2019. Unfortunately, this isn’t available but the absence of info raises lots of questions. Overall, we’re left feeling that this survey data should only be used for bigger picture trends.

Next, how does the butt implant data in 2020 compare to previous years?

Bloomberg correctly states that butt implant surgeries reported by the survey were up by 22% in 2020 versus 2019. However, Bloomberg doesn’t disclose any data before 2019.

Let’s look at a 10 year history:

What does this mean?

The increase from 2019 to 2020 may seem to be a significant one year jump but it appears to be insignificant when considered over a longer period of time.

Have we reached minimum viable meaning?

Yes, for 2020, it seems that we know that the increase doesn’t mean much in historical terms. But we’d really like to understand what was driving the increase in butt surgeries in 2015 and 2016. That might require some analysis of social media and reality TV show trends that we don’t have time for today.

Next, how do butt surgeries compare to all cosmetic procedures?

The data for butt surgeries seems not to be correlated to that of all cosmetic procedures. While butt surgeries peaked in 2016, all cosmetic surgeries continued to increase each year other than an unexplained dip in 2017 and the pandemic-related dip in 2020 (on average plastic surgeons were locked down for more than 8 weeks in 2020 or 15% of the year).

How does the butt implant data vary across demographics and could that provide any meaning?

Interestingly, more than half of butt implant surgeries are performed in the South (from the Atlantic to Texas) while more than half of butt lift surgeries are performed in New England and the Midwest.

What does this mean?

We don’t know! But it’s interesting to see the regional differences. For instance, in 2018, 51% of butt implant surgeries were performed in the South while only 35% of total cosmetic surgeries were performed in the South.

Have we reached minimum viable meaning?

No. It would be interesting to dig more into geographical and age differences among the procedures to see if there are any interesting trends.

Finally, is the increase in butt implants due to the Zoom effect?

In the visualization, Bloomberg claims that “social media envy coupled with the ‘Zoom effect’ sent us scurrying to rectify the ‘secretary spread.'” This is presumably based on the quote from a dermatologist in the article who claimed that homebound stagnation and Instagram led people to “give their bottom line a much-needed boost.”

What does this mean?

Not a lot. In the press release for the 2020 survey, the ASPS says that Zoom calls created a “rush on facial procedures.” Logically, that makes sense. We’re all spending a lot more time looking at our faces and having other people look at our faces. We’re not sure why Bloomberg made the leap that being on Zoom would cause people to want to fix parts of bodies that are off camera. We’re also not sure why Bloomberg decided to use a sexist term like ‘secretary spread.’

Have we reached minimum viable meaning?

No. It would be interesting to track surgeries in 2021 to see if there is any connection between a year being on camera and cosmetic procedures. For now, it seems like quite a stretch to claim that the butt implant increase in 2020 has anything to do with Zoom.

One more thing…

Bloomberg’s visualization is remarkably confusing. Here is our list of things that are wrong. Take a look for yourself and let us know what else we might have missed:

  • The Y axis is confusing with labels on the right and a background shade that doesn’t mean anything.
  • The use of circles to represent size isn’t helpful because it isn’t possible to distinguish relative size in any helpful way.
  • The one data point they are focused on, buttock implants, is so small that you can barely see it…showing that the focus of the story may be insignificant.
  • The red and blue distinction isn’t helpful since it isn’t described anywhere nor do we know if or why it matters. The data point that’s the focus of the article, butt implants, is so small that it’s very difficult to tell if it is red or blue.
  • The circular arrows are distracting and don’t make sense. The description text connected to the arrows is in the same style as the labels which doesn’t provide any differentiation. The descriptions themselves aren’t supported in the data or article. For instance, the attribution to “choosy patients” for the decline in hair transplants has no supporting evidence.

Our analysis of the analysis

The core of this analysis took us about 45 minutes of questioning and digging for answers on an iPhone. A deep analysis of the data would require more time but a simple series of questions allow us to call bullshit. Bloomberg cherry-picked two numbers that helped create a clickable title. But there isn’t any validity to the claim that butt implant surgeries in 2020 are anything remarkable, let alone the “soar” that Bloomberg claims.

Bloomberg also missed the opportunity to find meaning in the data. They went for a quick hit story with “butt” in the title rather than trying to understand what plastic surgery truly says about us all in a pandemic. Given that people are staring at themselves more on Zoom calls while also not being seen in person, we wonder whether plastic surgery data could serve as a proxy for what this pandemic effect means.

When you’re next asked to draw a conclusion from data, remember to stay curious as long as possible and keep asking “what else?” until you’ve reached minimum viable meaning. MVM will vary based on the situation but you’ll know it when you reach it. Humans may be subject to cognitive biases and may be easily fooled but we’re also very good at figuring out whether something makes sense as long as we stay curious and keep questioning until we’ve found meaning.

Share on email
Share on facebook
Share on linkedin
Share on twitter