Well, geez. Who would have guessed? Apparently, that would be Spencer Chen, head of Marketing and Growth for Frontback and former marketing development lead for Appcellerator. He'd noticed a trend at tech conferences and had an opportunity for A/B testing, so he scienced! (Explanation of A/B testing and conversion rates with examples from Kinja, Cracked.com, and Adobe.com at the footnote*.)

My theory from years of being a part of trade show staffs is that the booth babes we hired were actually a drag on lead-gen. Up to that time, it was all empirical evidence based on being at shows where we had money to hire booth babes and events where we didn't. I noticed that we had always done better without the booth babes but it was just silly to suggest that we did better because we didn't have hot babes at the booth. I mean, I had a better chance of convincing my co-workers that the sky was purple.

His company had rented booth space at a tech conference and the organizers circled back to offer the company a second free slot; he accepted and decided to put his hypothesis to the test. He hired professional "booth babe" talent as well as "informed" local ladies with conversation skills and a personal knowledge of the conference area to see if the professional booth babe was better or worse than the local woman. Now, there are some issues with his experiment design, which I'll get to later, since these issues are good areas for further testing.

The booth that was staffed with the booth babes generated a third of the foot traffic (as measured by conversations or demos with our reps) and less than half the leads (as measured by a badge swipe or a completed contact form) while the other team had a consistently packed booth that ultimately generated over 550 leads, over triple from the previous year.


The short version of his observations:

  1. Booth babes intimidate the awkward nerds they are supposed to appeal to.
  2. Booth babes have a job of looking pretty and thus aren't motivated to learn the product that they are shilling for.
  3. Executives look for professionals to talk to, not booth babes.
  4. Booth babes generated less leads but also lower quality leads.

Read the full article here for more detailed findings. Now to the problems with the experiment design and other items that should be tested.


Location: The second booth was a free booth offered to the vendor. To me, this implies that the vendor booths didn't get filled and the organizer wanted to ensure that all booths were filled. This also implies that the free booth may have been in a poor location. It is important for weighing the validity of the study to ask which group was at the free location. If the free location had the professional booth babe, the location could have a negative effect on sales, skewing the booth babe results down; if the "grandmothers" were at the free location, skewing the non-booth-babe results down. I would love a clarification on which group was at the free booth.


Local knowledge vs. professional knowledge: the women who were hired were hired in part to converse with potential leads and give them local dining and entertainment information. In the past, women who work as programmers have gone to conferences and been insulted as clearly uninformed, and since they were uninformed anyway, the role should have been filled by a booth babe. [seeking citation, see footnote**] If the booth babe equivalent were actually a sales rep or product manager or other professionally informed role, we would expect the results to be different. It would be interesting to determine if the local color aspect is more or less valuable than being informed about the product.


Professional appearance vs. scantily clad: similarly, it would be interesting to test whether a professional appearance on a professional booth babe is seen as more serious, and thus higher quality, generating more or better leads. If executives prefer a more serious look on the female reps, this could discredit the idea of booth babes completely, but that isn't precisely what happened here. A good control would be to have uninformed booth babes dress as professionals on alternating days or to engage twin/ equivalently hot booth babes.

At any rate, the results are definitely interesting and hopefully this will lead away from the assumption that uninformed eye candy is a requirement of doing business at conferences. At the very least, booth babes are the people pointed to as Platonic Ideal "fake geek girls" at conferences even though they aren't faking being geeks and are known to everyone to be professional eye candy; eliminating booth babes could help the way that women in tech and geek women are treated. If you have ideas for other controls and variables, please expand in the comments.


*The idea of A/B testing is that you have two near equivalents, you offer both and see which one works better. For those familiar with Kinja (our feisty blogging platform), you can see A/B testing in features like the different commenting system at Valleywag. Valleywag has one commenting display and Jezebel has a different commenting display; by comparing statistics between Valleywag and Jezebel as well as between Valleywag current and Valleywag old school, the Kinja developers can decide whether to move Jezebel to the Valleywag display, tweak the Valleywag display, or change Valleywag back to the Jezebel display. In the recent Cracked.com redesign, some users were presented with the new look and given a survey for providing feedback; after tweaks, the site was changed to the new display.


Adobe.com and many other websites do A/B testing by assigning a group of users to an experimental display and seeing if that change affects free trials and sales, with sales considered a "conversion." (Some sites would consider starting a free trial to be a conversion; I don't recall if Adobe.com considers this a conversion.) By measuring the conversion rate, the business can decide on a course of action that has the best chance of business success.

**IIRC, there was an incident somewhere between 2004-2009 in which a female programmer went to a gaming conference to demo the cutting-edge battlefield AI that her company was developing for a game that I believe involved centurions, after which she related that she was repeatedly interrupted and dismissed on the grounds that she was a woman, asked where the booth babes were, and denied the ability to give demos. The game engine may have been initially developed by perpetual, but I'm not certain. If someone recalls the incident in question and can send me a citation, that would be greatly appreciated.