How should we measure gender-based equality for opportunities at work? originally appeared on Quora: the place to gain and share knowledge, empowering people to learn from others and better understand the world.

Answer by Jessica Margolin, Consultant: Organizational Design & Impact, on Quora:

Financial folks know that diversification lowers risk -- it's a simple mathematical proof -- so it has always been interesting to me that we humans, social animals that we are, feel safer in groups that have a shared history or context. I guess it's yet another cognitive bias, like "sunk cost."

Now that the issues surrounding gender-based harassment have become front-and-center, many companies are reviewing their policies for gender and other kinds of diversification.

We all know that it's easier to communicate progress when we're quantitative, so it's a natural urge to wonder how to best measure gender-based equality for opportunities at work. It's a fine question about metrics design, but even given that's what I do -- it's my hammer for which everything is a nail! -- I believe that's a bit premature.

The trivial answer is that only one success metric matters: monitor the ratio of males to females at each level, and when that's 50:50, then you're finally balanced. After all, as a woman with MS and MBA degrees, I always wonder when a company's employment ratio across each level (especially when it comes to the C-suite) isn't 50:50, why that might be. It's a situation that invites speculation.

But even so, we have many moving parts to get to the point where the hiring funnel is smoothly egalitarian for both men and women, which is why I believe measurement is premature.

Focus on process, and process indicators.

The fundamental component that must be addressed is our uniquely human desire to find people that we can relate to, with whom we have shared experience and context. In corporations, this is usually called "fit."

When companies include "fit" in their hiring effort, they're very likely to skew towards inequality.

But you can adjust the way you evaluate fit and, at the same time, smooth out the paths that will lead a diverse workforce to your door.

(1) Deconstruct which aspects of fit are necessary and which aspects of fit aren't.

Everyone in a scientific group is rigorous in how they conduct research, for example, and that professional competence may be apparent in how they speak or think about a wide variety of things. However, if they like dogs/drink beer/play ping pong/enjoy horror movies, that's not actually relevant to their ability to do their work with joy. Be sure you understand the causality.

These social fit components may seem innocuous, but by the time you pull together a persona that hits several of the same areas of interest for each of the key interviewers, you're going to create a homogeneous culture.

Check your own company description to ensure that "what you like to do" isn't actually "how you like to live your life." Separate the relevant components of core competencies from the social components of "fit."

(2) Explicitly manage the risks associated with homogeneity.

When you're interviewing, specifically avoid connecting only over hobbies or stage-of-life. This keeps the interview neutral. The alternative is to run the risk of having women self-segregate: if they're "cat/vodka/swimming/art-house" they are likely to feel awkward around all those "dog/beer/ping pong/horror flick" people. When people feel awkward, they often act awkwardly, and that may mean you're not getting a solid interview from them in the first place.

Similarly, if someone seems great for the job except you don't immediately want to socialize with them, be self-aware enough to give them 100% thumbs up on hiring. If you absolutely must express your disinterest in being stuck in an airport with them, then you can say, "I was unable to find a common area of interest." If it comes down to two great candidates in the end, and one is a man you'd want to hang out with and one is a woman who seems perfectly fine, but you don't really "get", you need to hire the woman.

It may feel riskier, but it's not. Or rather, the risks are those associated with homogeneity vs. the productivity risk that you'll not have as much fun at work because she's "weird" in some undefinable way. You're trying to fix the homogeneity risk. So fix it.

(3) Be very aware of sexual dynamics, by which I mean completely block all thoughts about sexual dynamics.

On this, you're going to need to assist each other in changing your habitual thoughts. It doesn't matter if she's pretty, wears makeup, has a husband, is lesbian or straight, or poly or mono. Do not even begin to think about whether she wants babies, has a family, or "has hot friends." She doesn't have to smile; forget leaning-in or leaning-out, and who cares whether she's a feminist? And for god's sake leave her religiosity--or lack thereof--out of it. Just evaluate her for the job.

It's very hard to avoid the urge to be with people just like us. This urge is the cognitive bias we're working against when we evaluate people: we're social animals, and people who are like us feel safer. But if you want to minimize homogeneity risks, you've got to avoid seeing a woman as an exception, subject to all this additional scrutiny. Everyone enjoys shared context--me too!--and we all have to not bring this preference into our workplace. The social benefit, of course, is learning about different contexts.

How do you measure progress when so much is changing?

What I've described above is no different than any situation where you're setting up and measuring behavioral change goals. These are general approaches that will work when pursuing any type of diversification strategy.

Like any cultural transformation, the process of change is very difficult, but self-sustaining: once you establish an effective cultural process, it will reinforce itself.

(1) Set up your culture.

Understand who in your organization is good at interviewing candidates in an unbiased way, and who is not. Coach the latter, but more importantly make sure everyone knows who the "heroes" are, and why they are heroes.

Anoint your "clergy," the people who explain the meta-information around the changes in how people interact and how to adapt to a large percentage of employees from different contexts. The employees who make good "clergy" are not typically in a position of organizational hierarchical authority. Their authority is that of any natural teacher, who repeatedly simplifies difficult concepts and keeps people engaged when they're frustrated. Cultural "clergy" can coach people one-on-one, allowing employees to grow into the new cultural norms. Be sure you clearly communicate expectations, and provide positive incentives for achievement as well as consequences when they're not met.

The measurement goals here are then to:

  • Identify and provide resources to your heroes and clergy.
  • Develop systems that ensure the clergy are visible enough to be discovered by those who need to find them.
  • Understand how well the norm is settling in, and how often there are problems with it.

You need to know and clearly identify what the norms are, and then determine how to best measure their acceptance.

(2) Determine the existing state of your cultural systems--onboarding, training, reviews, and promotions. Set goals for growth, and try different programs to see what works as a coaching path in your organization.

How skilled are your interviewers? How might they improve? Do you have a program training people how to interview? A webcast? A handbook? A list of questions interviewers must ask every candidate?

In a large organization, goals are set up as individual objectives. In a small organization, the goals might be simply to get these things delivered.

You're also setting up and evaluating concrete goals deep in your organizational processes because your new hires today will be your interviewers tomorrow. As your culture migrates, a more diverse workforce is going to impact processes and programs across the board. That, after all, is the point: you want the new hires to manage risk, and that won't happen if they aren't heard and their ideas are ignored when they do point out risks and propose solutions.

(3) Look at your hiring funnel, and see where you're losing the women.

  • Are you sourcing enough women?
  • Are you getting them in to see you -- do they make it past the phone screens?
  • Are they doing OK in the in-house screening process (or shooting themselves in the foot somehow)?
  • Are you finding that women don't take your offers, or are unhappy once they're employed?

Each situation requires different interventions, and if you're serious about increasing diversity in your company, you must focus on learning how to adjust your recruitment and retention.

These goals, the ones associated with your hiring funnel, can get very numerical and "hard," and reported up and out to investors. But beware: because this risk mitigation will require cultural change goals, metrics about gender balance won't be terribly meaningful until you understand what's going on more thoroughly. If you can resist the temptation to make a dashboard of your funnel before you have some idea what you're going to do to move things forward, you can plan out a detailed and well-designed solution rather than creating a "bug fix" that introduces more bugs than it solves.

While the path to lowering homogeneity risk is difficult, it's well worth the trouble, both because of the risks it mitigates and the new opportunities it provides. When you focus on lowering a specific type of risk and put your full efforts towards that goal, you'll clearly see which metrics optimize the information you need in order to lead your organization through cultural transformation.

This question originally appeared on Quora - the place to gain and share knowledge, empowering people to learn from others and better understand the world. You can follow Quora on Twitter, Facebook, and Google+. More questions: