November 7, 2005
Recruiting Top Candidates- What Does “A” Caliber Mean?
I received a question on a recent posting about what constitutes an “A” caliber person from a recruiting point of view. In my view, this is one of the toughest questions faced by management teams. I had been thinking about the issue when I read Friday’s New York Times article on Tiger Woods, “Tinkering With Success.” The basic theme of the article is that Tiger has been an excellent golfer for a very long period of time (winning eight major championships through 2002), but has spent the last 20 months changing his swing (and his coach). The results have been tremendous this year…he has won two major championships and six tournaments. Tiger is an “A” caliber person!
The problems with this observation in the context of recruiting into your company (besides the fact that it is obvious) are:
- It does not (by itself) help predict if other people will be “A” caliber at golf as it is describes the result rather than points out characteristics that help predict the result.
- Even if it did point out predictive characteristics, the predictive “algorithm” would only work for golf (i.e., While he is “A” caliber at golf, Tiger probably is not an “A” caliber engineer!).
My involvement on the issue of finding “A” caliber people over the years has led me to the following “best practice” approach:
- Focus your attention on only one or two positions at a time. This is extremely important, as the effort required to do the job right is considerable. If you do get it right, you will significantly improve your company (I continue to be amazed at how the addition of one key person to an emerging growth company has had such a significant effect on the companies that I am involved in). Given that you want to get the “biggest bang for the buck,” I would suggest that the possible focal points are:
- a position that is currently a significant hole in the team,
- senior level positions (C-level/VP level), or
- a function that you intend to hire a large number of people into over the next year or two (probably a specific sales or service function if you are at the expansion stage).
- Output has an “efficiency” angle and an “effectiveness” angle. Efficiency is all about pure input/output productivity (how many sales calls per week does a salesperson make, how many lines of code does a developer make), while effectiveness is all about quality (how many sales per sales call does a salesperson make, how many errors per line of code does a developer make. I know, the examples that I give are really simple and not the only measure of output for the functions mentioned, but hopefully they serve to get the point across. Also, don’t confuse these simplistic examples with Taylorism. The intent is to get you thinking on the topic, not pulling out a stopwatch!)
- It is relatively easy (at least conceptually) in an individual sport like golf to determine “A” level output. From an efficiency point of view, you can look at the number of tournaments/top tournaments that a player enters for a given season. From an effectiveness point of view, you can look at tournament wins or wins against top competitors as a couple of high-level metrics. You could also disaggregate the measures of the game into some components like length and accuracy of drive, pitch accuracy, putting accuracy, ability to recover from tough situations. You could also disaggregate each of these components further. For example, for the drive swing, you could look at club speed at impact, body speed, arm speed, form etc.). I have not read (or looked for) a book on disaggregating golf, a great read on the use of statistics in baseball is “MoneyBall: The Art of Winning an Unfair Game” by Michael Lewis.
- When you move to a team sport, like basketball, the issue of determining “A” level output gets more complex for a few reasons. First, basketball involves specialist roles (e.g., Point Guard, Center, Forward), and you need to make sure that you have defined the “A” level output for each of the roles separately. Second, each basketball program has a somewhat different strategy (fast-break running game, outside shooting game, inside shooting game, pressure defense, etc.), and you need to make sure that you have defined the “A” level output in the context of this strategy (or planned new strategy!). Third, teamwork matters. The capabilities and style of the other players already on the team will help determine what “A” level output of the player/position in question. If the new player does not “fit in” with the rest of the team, the output of the entire team could suffer.
- When you move the issue to an emerging growth business, determining “A” level output gets even more complex. You not only have the same role/position specification, strategy, and teamwork issues as the team sport example, but also have additional burdens of constantly changing need to move from generalists to specialists as the organization grows, as well as very dynamic changes in markets, competition, and regulatory issues as markets and companies evolve (think about changing the rules of basketball and adding new players on the other team WHILE the game is in progress). Finally, you do not have the benefit of transparency, where you can gather up all of the output data from all of the other companies that have similar functions…you only have your own data to work with!
- So how do you determine the right “A” level output for a given position, given the dynamic nature of the business? This is part art, part science, but to the extent that you have people already in the position/role, you need to “tease out” the output from the best relative to the rest (assuming that you have one or more people performing at the “A” level). You should also reach out to your network and advisors to get as much information and as many viewpoints as you can from them, especially if this is a new position (if you have a good VC, this is easier, as you will generally have access to other companies in the VC’s portfolio as well as the experience of the VC. Also, really good professional recruiters can be helpful in all stages of the process, including this stage.). Take all of the data and ideas that you can, and distill it down to what output that you think constitutes the “A” for the role/position in question (and what is the priority for the output that you are looking for…conceptually similar to a “hierarchy of needs” for the position in question). This is important, as you may decide at some point to “back off” your original criteria (for now). If you do, you will want to know what you “must have” vs. what is “nice to have” for your current stage of development.
- Try to make sure that the characteristics are “causal” in nature. This is the toughest step, in my view, as you need to separate correlation (things that move in similar ways) from causality (things moving that cause other things to move). An example of this (from my college statistics course) is that someone once figured out that as the number of Bishops in England went up, the number of fires in England went up. Cause for alarm if the Bishops CAUSED the fires! But it is only correlation (the two things moved in similar ways). The cause of both the Bishops increasing and the Fires increasing was that the population of England was increasing (the population increase led to an increasing need for Bishops and increasing accidental fires). More relevant to golf, (from Tiger Wood’s fun facts), the fact that Tiger wears red on the last day of a tournament is less causal in nature (at least in my view) than the fact that he first started swinging a golf club at 11 months (note: neither are probably the best criteria to use for predicting golf success, but the latter is probably more predictive!)
- You need to determine the time scale for the predictive characteristics. For example, do you need someone to “hit the ground running” (in which case, you need characteristics that predict the person’s output today) or are you hiring less experienced people that you want to grow over time (in which case, you need characteristics that predict that a person will develop into having the “A” caliber output over time). For example, getting a “plug and play” salesperson to sell advertising may suggest characteristics such as “currently selling advertising to the same advertisers for the same type of media,” “sells a lot,” and “works well with others.” Getting a less experienced person right out of college for the same type of role might be characteristics such as “leadership,” “aggressive,” “outgoing/extroverted,” “team oriented,” and “passionate for this job.”
- Make the characteristics comprehensive. You need to make sure that you encompass the role.
- Keep it as simple as possible. No need to get overly complex here. Boil the characteristics down to the minimum necessary. (This point is not in conflict with the prior point.)
- Turn the characteristics into useful screening criteria. For example, you can measure “sells a lot” by the salespersons W2 (that is, what (s)he got paid last year) and the written compensation plan for the position, “works well with others” through personal reference checks, the college graduate’s “leadership” through extra curricular activity leadership in college, etc. The key is that you can get the facts that help determine if the person has the characteristics that you want. (Btw, you need to do real work here when you are doing diligence!)
- Once you start interviewing, revisit and refine these criteria. I find that the process of interviewing helps to both identify the “land of the possible” and is also quite useful at improving your thinking on ingoing criteria.
- Note1: You can make these activities more comprehensive, and possibly more accurate, with activities like psychological and behavioral testing, but I have not seen it “up close and personal.” My sense is that it is probably better suited for large scale hiring efforts that can not put the time into a deep understanding of the candidates and may be better suited for finding “minimum competency” rather than “maximum competency” of hiring “A”s as I discuss here; therefore, I believe the activity is better suited for the large companies.
- Note2: I have used an organizational psychologist to help with the defining characteristics and interviewing periodically. His high emotional IQ and different angle on the issues has been quite helpful.
- Note3: When companies have the luxury of lots of data and resources to analyze the data, this approach can (an has) been taken to the extreme through regression modeling. I have not, however, been in a position with any of my emerging growth companies where we have done (or even considered) this. It is really more of a large company activity.
- You will want to have a broad enough sample of different input characteristics so that over time you can tune the characteristics that make an “A” caliber person in the function. Since there are a lot of factors other than “population” that determine the “number of fires” in England, there are most likely a number of more idiosyncratic factors that define your optimal job candidates for a particular position.
- Diversity will help you over time as your business and its needs change (think about Charles Darwin’s ideas!).
- Diverse people also have diverse interests and motivations. With diversity, you essentially are allowing the business to run along its efficient frontier, which essentially means that it allows you to maximize your overall output (through increased “intelligence” and “network” of your company) while it averages out the hills and valleys in individual motivation and performance.
That is it in a nutshell. This process should help you determine what constitutes an “A” caliber person for a given role in your company. One last thought on setting the “A” criteria. Tom Watson says, “Tiger has raised the bar to a level that only he can jump.” There is not another Tiger Woods, so if you set the bar at that level, you will not be successful. Set a bar that is realistic for your situation and gets you the top candidates that you can find. The key is to “raise the bar to a level that only a few can jump!”
Also, once you have the criteria for the ideal candidate, make sure that you are doing the work necessary to turn the paper candidate into the productive employee, including:
- Put a Process in place that “screens” for the right characteristics. See my post on “Recruiting Best Practices” for some thoughts on the screening process.
- Set up the “environment” so that the individual can thrive. (See my post on “attracting, motivating and retaining” for some thoughts on setting up the right environment.) If you do this right, you will get all your staff, not just the “A” caliber staff, working above their natural level.
This process works well, and it should work well for you.