February 12th, 2010

The edges of search firm automation

Over the past few years I’ve pushed the boundary around Search Firm Automation.  Specifically, just how far can one automate the process before engagement rates fall significantly.   Once such project involved a partnership with a company in Canada called Boxpilot.  (www.boxpilot.com)

Here’s what happened and the results we achieved.

The open position

We were hired to find a senior business development leader for a 1.4 Billion IT services firm.  The ideal candidate would have at least 12 years of outside sales experience, a proven track record, and currently involved in the selling of a specific consulting service.  The ideal candidate had to be able to “touch” the regional office at least 3 days per week as well and therefore needed to live within a specific geographical area.

What we did

Because we wanted to examine the use of Boxpilot “in the wild” we created a list of passive candidates we’d never spoken with before….pure strangers who had the background, experience, and location that fit the open position.   Our list included 50 targeted individuals we wanted to engage on this project.

We then found the phone numbers of these targeted individuals, confirmed the phone numbers, and sent the list of 50 to Boxpilot.

We recorded one general voicemail that we wanted Boxpilot to deliver to each of the 50 targeted passive candidates.  Because we were going to use the same voicemail for each of the 50 we did not use any specific names…just a general voicemail stating that we’d like to speak with the targeted candidate about roles in their industry and geography and a bit about who we are and what types of roles we recruit for in their area.

What Boxpilot did

Boxpilot employees personally dialed each number on the list and navigated the voicemail system to arrive at the target’s voicemail.  Once they confirmed they were in the right voicemail box they delivered OUR voicemail to that box.  Thew were able to deliver our pre-recorded voicemail to approximately 75% of the people on our list.   They only left the voicemail once per person.

The Results

Eighteen percent (18%) of the targeted passive candidates who received our generic pre-recorded voicemail called us !  We never used their name and we didn’t even place the call.  This was slighly higher than our null hypothesis of a 10% call back rate.   Of the candidates that called us back none ever mentioned that the voicemail was left after-hours and that we hadn’t used their name.


Given an 18% call back rate when reaching out to complete strangers the call back rate would be significantly higher if the list was warm.  Also, due to the low volume of targets the cost per candidate call back was high.  Our next plan is to use traditional phone calling for the initial call and then consider using Boxpilot for that very important second call to those that didn’t respond initially.

I see additional uses for this methodology in the recruiting industry and I will continue to post efforts and results.

You must be logged in to post a comment.