Five Scenarios for the Future of Recruiting 6: Invasion of the Shallybots

Spring 2010 conference-logoI spent an hour on the phone yesterday with yet another entrepreneur who imagined that the future revolved around the “eHarmony for Jobs.”(The idea was tired a couple of years ago.) I regaled him with stories of Intellimatch, itzBig, JobFox, and 40 other matching services. They all planned to use structured profiles and assessment tests to ensure a fit. The primary problem with these schemes is that they always require too much investment of time (from candidates and employers alike) to actually work. The secondary problem is that the investment required to make the technology make sense is measured in billion$, not million$.

Lots of forecasts for the future of recruiting and HR focus on phenomenal breakthroughs in technology’s ability to personalize and match environments. That’s probably not really going to happen in the foreseeable future. The triple disciplines of sourcing, attraction, and selection will continue to require human intervention at the decision making point.

Some commentators are beginning to notice that the world is already heavily served by a bot-mediated culture.

Forget about HAL-like robots enslaving humankind a few decades from now; the takeover is already underway. The agents of this unwelcome revolution aren’t strong AIs, but “bots” — autonomous programs that have insinuated themselves into the Internet and thus into every corner of our lives. Apply for a mortgage lately? A bot determined your FICA score and thus whether you got the loan. Call 411? A bot gave you the number and connected the call.

Highway-bots collect your tolls, read your license plate, and report you if you have an outstanding violation. Bots are proliferating because they are so very useful. Businesses rely on them to automate essential processes, and of course bots running on zombie computers are responsible for the tsunami of spam and malware plaguing Internet users worldwide. At current growth rates, bots will be the majority users of the Net by 2010.

We are visible to bots even when we are not at our computers. Next time you are on a downtown street, contemplate the bot-controlled video cameras watching you, or the bots tracking your cellphone and sniffing at your Bluetooth-enabled gizmos. We walk through a gauntlet of bot-controlled sensors every time we step into a public space and the sensors are proliferating. — Paul Saffo (See here (long video), here (book) and here for a crisper understanding of the bot-mediated future.)

In other words, rather than a huge, monolithic big brother, it’s more reasonable to expect something like a swarm of little tools: Shallybots.

Contemporary software development processes emphasize incremental progress rather than grand vision. While the “big idea” is certainly an important force, tools like the scrum methodology focus on the delivery of high priority results in an iterative environment. (Translation: keep your eye on what’s important.) Older approaches to technology projects echoed their industrial roots. The big project/big picture approach with a cascading series of “waterfalls” worked to get a man on the moon. It doesn’t work as well when you want to make daily forward progress.

So, we will increasingly inhabit a world that is riddled with bots. These one-task-at-a-time bits and pieces of automation will increasingly hold the responsibility for Internet filtration. To the extent that a job req is designed to fill an empty slot (find a replacement worker), bots can be developed on a case-by-case basis. The technique doesn’t work as well when the job is brand new or the organization is small.

That suggests a world with lots of job boards, sourcing bots, and highly targeted advertising networks.

There’s an incredible temptation to think of the future as more of the past with a little something extra. Even asking a question like “What is the future of recruiting” assumes that things will continue to behave similarly. More likely, lots of little things will get automated and we’ll develop architectures and nomenclatures for the new structures.

Hadoop is the network architecture that underlies the calculation speed of Google and Yahoo. By organizing around single-instance problems spread across many, many servers, the framework produces quick results to one-off problems. It’s another aspect of the move to solving micro-problems rather than their imponderable macro cousins.

Article Continues Below

There are only a few instances of scrum methods and Hadoop implementations in the recruiting and HR space. Rest assured, they are coming and will be the foundation of the next waves of change. The essence of the approach is to take a lot of bite-sized moves to produce change.

In the Shallybot scenario, tens of millions of little alerts and triggers are constantly going off as you move through time and space. Potential employers know you by characteristics and have set thresholds for paying for your attention when they need it. At the same time, sourcers, working at complex dashboards, monitor availability and requirements while they continue to try to discover novel answers to oft-repeated queries. Selection bots use behavioral indices that resemble credit scored background checks to winnow the funnel.

As the tools create ever-refined personalization, recruiting becomes less and less standardized. Nuances for regionalization, industry specifics, cultural attributes, and other factors. In the Shallybot future, recruiting produces better matches as a result of 10,000 little things rather than one big one. It’s the opposite of the eHarmony for jobs.

This research is sponsored by Pinstripe Talent.

To read the rest of the series:


13 Comments on “Five Scenarios for the Future of Recruiting 6: Invasion of the Shallybots

  1. Sorry if the Shallybot nomenclature was a little obscure. When I think about the way that bots work, I imagine highly refined single purpose automatons that relentlessly do the same single thing over and over again.

    In that sort of automation (which is the logical extension of the newer, agile coding techniques, you end up with software that works more like a swarm thatn the relatively single purpose stuff we’re used to.

    I coined the term ‘Shallybot’ as an homage to Shally Steckerl. My sense is that the essence of the search techniques he teaches is this single minded pursuit of a very specific objective. While the rest of the market is focused on one-size-fits-all toolsets and approaches, Shally seems to be demonstrating the foundations of future sourcing operations.

  2. John. I think I get it.. but I have to we all talk into the echo chamber constantly..what specific examples of “search techniques” can you point to that Shally (or anyone) has put into practice that connect to your thinking.

    Thanks John

  3. Thank you, John, for eloquently stating what many in the job board industry have felt for years: that the concept of matching is simply not feasible because it requires the complete buy-in of both the employer and candidate. Ironically, those who seem to be the most excited about matching technology are often the worst fits for it because their excitement is largely based around the promise of time savings. They seem to feel that the matching technology will allow them to apply for the right jobs or find the right candidates in less time than it takes them to do that now on more traditional job boards, yet the technology won’t work unless they spend more time than they are now.

    My conversations with job board industry leaders and people at some of the matching sites reveals that the clients who have tried some of these sites inevitably walk away disgusted don’t seem to understand that it isn’t the technology that failed, it was their efforts which weren’t adequate. Maybe they were over sold by being promised that the technology would do all of the work for them, but savvy buyers should have understood that the software only knows what you tell it and if you’re not willing or able to tell it what it needs to know, then the results you’re going to get will be crap.

  4. IMHO, Shally-bots (or as I call them: “Stasi-bots”, after the East German Secret Police) will make much sourcing obsolete. Why? When you know where someone is and what they’re doing, you don’t need to look for them. Consider this ~2012-2015 scenario: a bright 16 year old girl wins a prize in a regional computer-science competition for high-schoolers. This sort of thing draws the attention of one of “The Corporation’s” many Stasi-bots, who are then alerted to use their data-mining and aggregation subroutines to assemble a digital dossier on young Ms. “Sinha”. By analyzing her DD, it’s determined that among other things, she is quite a fan of the latest trend in Young Adult Fiction: YA Zombie Novels, so she receives a YA Z-themed text-card (eye-balled for appropriateness by a $2/hr Virtual Assistant in Ho Chi Minh City) inviting her and her parents to visit “The Corporation’s” campus over spring break at “The Corporation’s” expense. It all starts from there….

    Your thoughts.

    Keith “Reads Too Much Old Cyber-Punk” Halperin

  5. Keith — Your vision seems realistic from a technology standpoint yet unrealistic from a practical, corporate recruiting standpoint. I can see a third party recruiter that stands to make tens of thousands of dollars from a placement investing the resources necessary to make something like this happen, but how about an employer that hires hundreds or even thousands of entry-level candidates a year? One of our clients receives 200,000 resumes a year and hires 8,000 of them. There’s just no way that a scenario like this will work for that kind of mass recruiting.

  6. Thanks, Steven. IMHO, there’s very little in recruiting (except the most valuable things like closing and advising) that can’t be accomplished with increasingly cheap massive computational power and increasingly skilled inexpensive labor, along with the will to use both. There’s a guy who has proposed cross-referencing every email (one outfit I spoke to claims to have 800M emails), phone number, physical address, and name in the US and then applying datamining techniques to create those “digital dossiers” I mentioned based on purely public information. Imagine if we all could “know everything about everyone”?

    Be Seeing You,

    Keith “We want information… Information…INFORMATION!” Halperin

  7. I assume your reference to 800M means 800,000 and not 800MM which would be 800 million. A list of 800,000 is actually not all that large. We have a double opt-in list of 10 million college students and recent graduates with up to 700 fields of data per person including valid emails for all of them. We also have permission to text about 60 percent of them so cell phone numbers for about 6 million of them.

    The reality is that data is cheap and anyone who is saying that a huge database is the next best thing should advance their calendar forward from 1996 to 2010. It is how you collect and use the data that matter now.

  8. No Steven, I meant 800,000,000. Of course you need to know what to do with the data, and that requires intelligence to organize and interpret the data and not just storage or computing power. An example:
    Here’s my google which I have on my resume.
    Tell me about me- not the facts, the MEANING of the facts.
    That’s the REAL ticket…..

    Have a Great Weekend, Everybody!

  9. It’s great to watch this conversation evolve.

    One of the most interesting possibilities is that the data that’s in the rest of the organization is going to migrate towards recruiting rather than vice versa. There are really interesting experiments and projects already underway to use the full spectrum of data in the social graph for recruiting. ( In not so many years, that will be 6Billion email addresses)

    The game is moving beyond the previously understood limits, as Steven rightly notices. Keith’s vision just hints at the levels of complexity that may be discovered in the next generation of software evolution.

    My view remains that lots of little bots that do one thing really well will be harnessed to do a lot of the stuff that we’re just beginning to imagine. I’m not so paranoid as to think of them as nano-secret police (although I do understand the sentiment). Benign management of little tim=ny inconvenient questions and processes, executed simultaneously produces something like we’ve not seen before.

Leave a Comment

Your email address will not be published. Required fields are marked *