The Chenope Blog

We bring more things to light

January 23, 2015
by chenope

Is There A Diversity Problem in Silicon Valley?

Recently, I’ve been reading a lot about alleged gender discrimination against women in the valley. While I am a sharp critic of much of what goes in in the valley, and have surely seen many incidents of atrocious behavior of different kinds, I have to admit that I find these claims of gender discrimination rather mystifying.  This is because I’ve never seen any evidence of it. Now of course that doesn’t mean it doesn’t happen. But on a widespread basis, I doubt it.

By way of background, I am a woman who has worked in various technical positions in the valley before becoming a CEO about 15 years ago.  Over the course of time, I have probably hired more than 100 developers.

Almost of all of them were men.  In fact, I believe only once did I hire a woman, and she was really more of a mathematician who subsequently learned to code. In a few cases, I inherited developers who were women, but even these I could count on one hand.

Now why is that?  It isn’t that I needed to do more to recruit women, which is something that I hear said a lot. As most people know, most software developers are men. This becomes even more true the deeper you get in the stack; the further away you get from UI stuff, the male/female ratios get even worse. I don’t claim to know why this is. I would suspect however that there are multiple reasons for it.  Whatever the reasons, I do not see it as particularly calamitous.  After all, what if most women simply don’t find programming appealing?   Is that really so horrible?  Many professions skew significantly by gender – for whatever the reasons.

Discrimination implies that the candidates are there but, whether by omission or commission, are essentially being ignored in hiring and promotion. I would argue that this happens far less in the valley than it does most places.  This is because the incentives to hire and retain women are so particularly compelling in the valley ecosystem.  While some may find the following analysis less than high-minded or cynical, it is extremely practical and is as follows:

  • Larger companies are highly incented to hire and retain women developers.  Any large, persistent gender imbalance makes the company vulnerable to various types of sex discrimination suits, which amongst other things are bad PR. However, having good ratios is good PR – and even better for recruiting, in both genders.
  • Startups may fear lawsuits somewhat less, and are largely immune from statistical standards. But they are at least equally motivated to hire female developers for two reasons.   The first is that in startups, recruiting really is everything.  Having women around, especially in technical roles where they are more accessible, is great for recruiting. For one thing, women are likelier to feel more comfortable joining a group in which they are not the first woman. In high octane startups in which the hours are typically long, dating among employees is extremely common.   And even when that is not the  case, younger male programmers who are often shy around women usually value the opportunity to really interact with women.  (Keep in mind that most people don’t write code for a living after the age of 40, so this is a real issue.) A second reason is the increasing prevalence of so-called “acqui-hiring” in the valley, in which large companies buy small ones based solely on a bounty per engineering head, often $1M per head. Large companies may crave those female developers, so having them is a plus.
  • In addition, tech companies who are often short on basic business fundamentals, place far more value than companies in other sectors on bring trendy and progressive.  Having floors full of men, with the occasional secretary sprinkled in doesn’t cut it.

The reality is that the CEO’s of IBM and HP are both women, though you rarely hear either mentioned as “women CEO’s”.  (For that matter, Mary Barra, the CEO of GM, studied electrical engineering, as did Ginni Rommety, IBM’s CEO.)  I suspect the reason for this is that in both cases, they are “real CEO’s” in situations that are to say the least very challenging and not at all for the faint of heart.   Difficulty in turnarounds is directly proportional to the size of the company; the larger the ship, the greater its moment of inertia – and these are really large, not to mention old, ships.

However both cut far less glamorous  figures than the younger, more attractive Marissa Mayer at comparatively small Yahoo, with its proportionally quite large Alibaba warchest.  In contrast to Rometty and Whitman, Mayer’s gender is frequently cited, but usually in the context of deflecting criticism or concerns about her managerial competence. For example, what seemed to draw the most public outrage on Twitter in the aftermath of an excerpt of Nick Carlson’s book in the NYT was Mayer refusing to hire Gwyneth Paltrow as some kind of guest editor because she didn’t have a college degree. That kind of thinking is problematic, but isn’t in any way gender-related.

Gender diversity though is only one kind of diversity.  My friend the late Frank Schirrmacher, one of the publishers of the Frankfurter Allgemeine Zeitung (essentially the NYT-equivalent in Germany) often lamented the lack of diversity in the valley.  However, his concern was diversity of thought, noting the extreme similarity of the viewpoints expressed by public figures in the valley.  This is a very legitimate concern.  Problems brewing for Google, FB, etc in Europe stem directly from this, but I believe are generally poorly understood in the valley.

Here’s another concern: Increasingly one sees companies getting funded, or given a bye, largely based on a very particular pedigree of school and former employer.   While this narrow-mindedness may not make the gender diversity worse, it surely cuts out those who go to working class schools such as Drexel.  Or veterans who got their degrees while in the military through whatever school happened to be accessible.  Or, as in the above anecdote, don’t have degrees at all, but could make potentially stellar contributions like Ms. Paltrow. Nor will you hear many people with accents that suggest that they hale from south of the Mason-Dixon line.  Nor do you see really overweight people on the Google campus.  I could go on, but you get the point. When viewed in this fashion, Ms. Mayer is getting in trouble not for being different from those around her (e.g. a woman) but ironically for being the same in a context in which it was inappropriate.

Not everyone would consider such things a question of diversity.  Certainly the law is specific about what diversity is actionable legally.  But in the end, it is never good for anyone to only associate with people who are very similar to themselves – and it is even worse for the leadership of consumer product companies.    I think this is a real problem – one of many in the valley that warrant some outside scrutiny.  Attention on non-problems is only de-focussing.



January 17, 2015
by chenope

More Tech, More Privacy (Expectations)

Since an old Merc interview of mine is now finally showing up in print, the topic for this week’s blog will be the interplay between technology innovation and privacy entitlements.

There are two constants when discussing privacy and technology:

  • Each new wave of technology adoption brings with it expanded privacy entitlements.
  • Privacy is always a trade-off with something else. And it always will be. What does change is what the ‘something else’ is.

Back in the day, privacy was most often traded-off against simple convenience. Sure, you could take the few minutes to go outside with your cell phone – or in the dark ages, walk to a pay phone – to avoid those in neighboring cubicles overhearing your conversation. Or you could just break down and risk your co-workers overhearing bits and pieces of your conversation. With old technology or new, you still had to go out of your way for privacy. Cell phones certainly brought more privacy – for one thing, you could take them anywhere – but also vastly greater expectations of privacy. Think about it: most people these days would rather have a friend or stranger rifle through their wallet than browse their cell phone contents.

In more recent years, the privacy trade-off has increasingly been with the desire for unfettered online self-expression in forums such as Twitter, FB, etc. For example, a woman delighted with the cool new pair of shoes she just bought may want all of her friends to see how fashion-forward she is, and so will likely post a photo of said shoes. There’s no problem at all with that, until/unless the shoe maven in question feels like she is being stalked if all she sees for weeks afterwards are ads from the brand of the shoes and competitors of same – as she very well may. But it can be far worse still: what if it turns out to be the case that buying a certain type of shoes (say stilettos) at a certain time of week (say Friday afternoon) is strongly correlated to the purchase of particular types of lingerie or condoms within the same few hour period? The relevant ads from nearby stores are all but certain to find their way to her cell phone in a timely manner, potentially causing giant embarrassment.

But is that really a violation of privacy – given that shoes were purchased at a popular store on a well-trafficked street where there was no privacy – or just brands playing the real world probabilities? And even if it is the former, at what point does the average person come to understand that there is a certain amount of randomness in all of this? That even if the ads aren’t random in a purely statistical sense, the correlations are often weak ones – if still profitable. Our shoe maven may have simply fallen in love with the shoes, and just happened to be passing the shoe store on the way to visit her grandmother. In other words, you can’t judge a person by the ads which target them. When viewed in this light, unfortunately-timed ads are not so much an invasion of privacy as just plain unlucky. If you do think that this example is a violation of privacy, you ought to consider the following question. What if a shrewd lingerie store a block away from the shoe store strategically stationed someone outside the shoe store to hand coupons to anyone who comes out with a shopping bag? Would that also be a violation of privacy?

The right to online complaining – and just outright venting – is just as much of an entitlement as is bragging. If, for example, you complain online that United’s baggage-handling robots literally pulled your brand new luggage apart, you’ll likely soon get ads offering deals on a new suitcase. Hard to complain objectively about that: for one thing, maybe the ads are actually useful given the circumstances, and for another it is hard to argue that privacy should attach to a publicly made complaint. Still, even in this scenario, one can understand why the ‘right to be forgotten’ – forgive the twisting of the European catch phrase – could be desirable; for example, if each time you see one of the suitcase ads, you become incredibly aggravated because it reminds you how shabbily you were treated by the airline.

Nonetheless, at the end of the day, few people are willing to give up their rights to self-expression online. So while there may be much hand wringing and pontification about privacy, most will not change their behavior. Effectively, at least in the U.S., the tradeoff has been made.

The question however remains on how many things in the so-called Internet of Things will reach the level of addictiveness or need such that privacy concerns are all but neutered. Even leaving aside concerns about hacking, there are legitimate concerns about what kinds of data could be collected, and what correlations could be drawn from them. In today’s WSJ, there was an article about Google suspending the sale of Google glass. It cited privacy concerns and lack of real world usefulness as the two main reasons. No surprise on either count. Few people can honestly say that their glance hasn’t lingered over long on something that it really should not have – call it the rubbernecking effect. Right there, there is a lot of potential bad juju to have to outweigh with irresistible value that has yet to be identified. Likewise, I suspect most people don’t want their cars telling them – or worse still, others – how well they are driving when they are having a hellaciously bad day. And while there may be something to be said for major appliances who send email about needed maintenance, or upgrade offers based on actual usage patterns, this could very quickly become irritating if more than very occasional. At some point it isn’t even a question of privacy so much as of common sense. Of course, many ‘things’ have limited privacy impact, if probably also limited value. My internet-connected toaster or fridge might not provide me all that much value, but is very unlikely to do me any harm either.