Leadership has changed — and in a big way — since the Web started upending the status quo two decades ago. That’s why we’re launching our new Cultivate event; we realized that businesses need new types of leaders, and that O’Reilly is uniquely positioned to help engineers step up to the job.
At the start of the 21st century, Google was in its infancy; Facebook didn’t exist; and Barnes & Noble, not Amazon, was the dominant force in the book industry. As we’ve watched these companies grow, we’ve realized that every business is a software business, and that the factors that made Google, Facebook, and Amazon successful can be applied outside the Web. Every business, from your dentist’s office to Walmart, is critically dependent on software. As Marc Andreessen put it, software is eating the world.
As companies evolve into software businesses, they become more dependent on engineers for leadership. But an engineer’s training rarely includes leadership and management skills. How do you make the transition from technical problems to management problems, which are rarely technical? How do you become an agent for growth and change within your company? And what sorts of growth and change are necessary?
The slogan “every business is a software business” doesn’t explain much, until we think about how software businesses are different. Software can be updated easily. It took software developers the better part of 50 years to realize that, but they have. That kind of rapid iteration is now moving into other products.
Software breeds agility, but agility doesn’t exist in a vacuum. Software businesses discovered the need for deep cultural changes — new communications patterns, new models for trust, authority, and accountability, new collaborations between different disciplines. There’s an important lesson here: if you want to become a 21st century business, you can’t just install some tools and imitate some practices. To transform your business, you must change your business culture. Success is as much a cultural problem as an engineering problem. If you want to prosper in the 21st century, those cultural transformations are critical.
At Cultivate, we will be looking at the challenges facing modern management: adopting rapid iteration, design-first thinking, becoming data-driven in reality, and providing a working environment in which all staff can contribute. We want to train a new generation of business leaders who understand the relationship between corporate culture and corporate prosperity. We are looking for engineers and other product developers who are leading product teams or moving into management, and managers who want to sharpen their skills and understand the challenges their organizations are now facing.
Here are some of the issues we’re discussing:
Although the Web was in full swing by 2001, for the most part, we still treated Web applications like 1980s software. Fifteen years ago, “Release early, release often” was a mantra for Open Source, but Web developers were only starting to realize just how radical that could be. Now, Web-native companies push new releases to their sites many times a day. Some, like Facebook, even have new employees make a change and push it to the site on the first day.
Other companies, including larger and more traditional enterprises, have taken notice. They want the same fluidity and agility in their own operations. They want to be able to release early and often, even with physical products. They have noticed that it’s not just companies like Etsy, Facebook, and Google: Tesla has avoided physical product recalls by shipping software updates to all the cars in the field. They have also noticed that Tesla is planning to deliver partially autonomous vehicles, not by releasing a new product, but by updating the software on cars already in the field. And other companies, including giant industrials like GE and Caterpillar, are poised to join in the game. Indeed, they have realized that they have to join in the game, or they may not survive.
But agile product development isn’t simply about installing software tools that support continuous delivery and testing. The best tools in the world will do you no good if changes have to bubble up to the top of a corporate hierarchy before they can be approved. Agility is really about communications patterns. Likewise, agility is about trusting the people closest to the product to make good decisions. Communication patterns aren’t technical issues; they’re deeply embedded in an organization’s culture.
Using data effectively
We started the Strata conference just when “big data” entered people’s consciousness. Wherever “big data” might be on the hype cycle, data has always been valuable and will remain valuable. Organizations that use data wisely and appropriately will succeed. Organizations that fail to use data appropriately will gripe that “big data isn’t all it’s cracked up to be.”
But what does “using data appropriately” mean? In Data Driven: Creating a Data Culture, DJ Patil and Hilary Mason wrote about having intelligent discussions about data. Being data-driven isn’t about coming up with an analysis that justifies the HIPPO’s (Highest Paid Person in the Organization) preconceived notions. Nor is data about coming to a meeting armed with tables and graphs so you can “win.” DJ and Hilary said that, “Rather than jumping straight into decision-making, we start with a conversation about the data.” For them, using data intelligently “disarms data as a political weapon” so you can have a discussion in which everyone becomes smarter, and can make well-informed decisions. They don’t reject the idea of intuition or going with your gut: data isn’t everything, and sometimes going with your gut knows what the data isn’t telling you. But until you’ve discussed the data, you’re blind.
Again, we’re not looking at a minor change. Building a data culture isn’t about “putting a Hadoop in your machine room.” We’re looking at a significant change in how organizations work. Data analysts have to take part in high-level discussions; participants have to listen carefully, try out hypotheses, and admit they’re wrong without fear of repercussions. Management needs to be flexible and willing to change course if the discussion shows that there are better options. Using data effectively requires cultural changes, not just new tooling.
Design from the start
Apple’s success is one of the biggest stories of the 21st century. In the late 90s, Apple looked like a company doomed to follow the tracks of Commodore and Atari. That’s obviously not what happened. What enabled Apple to succeed, and distinguishes it to this day, is its insistence upon design excellence.
As John Maeda wrote in #DesignInTech, companies need to “start with design, rather than end with it.” It used to be commonplace for engineers to make something, and then throw it over the wall to designers to make it pretty. That doesn’t cut it. That gets you a PC, not a MacBook, a Nokia, not an iPhone. But it isn’t just a matter of spending more money on designers. Maeda noted, “To achieve great design, you need great business thinking/doing to effectively invest in design — and you need great engineering to achieve unflagging performance.” You need to integrate design thinking into engineering teams from the beginning.
Again, this isn’t a simple change. Engineers and designers frequently don’t think alike. Engineers often have little respect for non-technical opinions, and designers are often asking questions about usability and user experience that don’t make sense to engineers. How do you integrate designers into product teams? How do you change developer culture so engineers can hear what designers have to say? Phil Gilbert, IBM’s GM of Design, trains IBM’s product teams and executives to incorporate designers into their teams effectively. “Design first” isn’t a technical problem that can be handled by exposing engineers to design tools, or by simply adding designers to engineering teams; there are cultural problems, communications problems, that need to be solved before engineers and designers can work well together.
Soft skills matter
Engineers are legendary for not having people skills. While that legend is certainly an exaggerated stereotype, it is absolutely true that effective management is about human problems. As Brian Fitzpatrick and Ben Collins-Sussman said in Team Geek, “writing software is a team sport.” Does your staff trust you? Do you trust your staff? Does your group learn from failure, or does it descend into a cycle of blame? Can you deal with the tensions that inevitably arise in a working environment?
Managers are frequently promoted because they have technical vision, or have made important technical contributions — or sometimes, just because it’s the next step. But “soft skills” haven’t been part of their education and career up to that point. And soft skills are all about culture, about rules that prescribe both how we act and how our actions are perceived by others. If managers are going to lead effectively, they have to understand how their behaviors affect others. And they need to understand that their actions define acceptable behavior for the rest of the group.
It all comes down to this: more diverse teams are more successful teams. Yes, we’re talking about race and gender. Talent is rare, and nobody can afford to waste it. We certainly can’t afford to drive 45% of women (and who knows how many others) out of our teams because the work environments are hostile. But diversity has many dimensions. We’re also talking about job skills: engineering, design, data, and more. I’ve talked to people on teams where electrical engineers clearly didn’t respect software developers, and vice versa. Our Velocity conference sprung from the need to break down barriers between software developers and operations staff. Perhaps the most important thing group managers can do is establish a culture in which everyone’s contributions are valued: regardless of gender, regardless of race, regardless of professional specialty or job title. That’s how companies become learning organizations that can react quickly and with agility to the challenges they’re facing.
If there’s one lesson to be learned from the past decade, it’s that breaking down silos is critical to success. If design needs to be part of product thinking from the start, then designers need a seat at the table. If companies want to use data intelligently, data scientists need a seat at the table. If companies want to release and update products rapidly, operations staff needs to be involved. And if corporations are going to be productive, they need to create teams that value the contributions of all the members, not just some.
Management skills aren’t tied to any specialty. You don’t need to be a data analyst to value data, any more than you need to be a designer to value good design. That’s why Cultivate isn’t specific to a single audience, but is relevant to all the audiences O’Reilly serves. We’re presenting Cultivate first at OSCON, but this isn’t a one-time thing. Cultivate will be a part of many of our conferences, including Strata and Velocity. We are committed to training the people who will lead successful teams, now and in the future.