Book Notes: The Hard Things About Hard Things

Notes from Ben Horowitz’s The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  • Pg. 52: “It’s a good idea to ask, ‘What am I not doing?'”
  • Pg. 59: What’s the secret to being a successful CEO? There’s no secret, but one skill stands out- ability to focus and make the best move when there are no good moves.
  • Pg. 65: Don’t be too positive. Tell the truth. Give problems to those who can solve them and are motivated to do so. Transparency engenders trust, gets more brains to work on hard problems, and encourages problems to be surfaced rather than hidden.
  • Pg. 71: Layoffs
    1. Focus on the future, not the past
    2. Don’t delay- once the decision is made, do it ASAP
    3. Be clear that the company’s performance failed, not individuals. Some excellent people will be lost in order for the company to continue
    4. Train managers- they need to explain what happened, explain that the employee is impacted and the decision is non-negotiable, and they should know and present all the benefits and support the company plans to provide
    5. The CEO needs to address the entire company which will give some air cover for a manager to do the individual firing. The message the CEO gives is for the people who are staying, not those that are going.
    6. Be present and visible after the firing. Show you care by being available.
  • Pg. 88: There are typically no silver bullets, only lead bullets. Sometimes, you have to admit your product is inferior. Improving it is the first and only thing to do. Everything else is tangential.
  • Pg. 91: No one cares about your problems. Just do your job.
  • Pg. 106: Train your people. Why?:
    • It increases productivity
    • It creates a way to manage performance by setting expectations
    • It keeps product quality high
    • It helps retain employees through good manager/employee relationships and keeping employees learning and challenged
  • Pg. 134: Management debt, like technical debt, makes things easier in the short run, but kills you in the long run. Don’t fall prey. Common management debt scenarios:
    • Keeping two good people who compliment each other when one person should suffice. It creates reporting issues, direction and decision confusion, the two people may begin to work cross purposes, etc.
    • Overcompensating an employee only sets you up for skewed salary ranges and the wrong incentive structure.
    • Not formalizing feedback will slowly chip away at clarity and focus for employees.
  • Pg. 174: Bill Campbell’s methodology for measuring executives:
    1. Results against objectives
    2. Management: are they building a strong & loyal team?
    3. Innovation: short term is easy, but imperils long term effectiveness
    4. Working with peers: effective at communicating, supporting & getting what they need from other execs
  • Pg. 179: Culture does not make a company because it takes two things to make a company and they matter more than culture:
    1. Build a product that’s 10x better than the competition. 2-3x isn’t enough to get people to switch
    2. Take the market: competitors won’t sit by and watch you take market share. You need to get it and hold it ASAP.
    3. Culture is important after the above two things are accomplished though.
  • Pg. 181: Focus on a small number of cultural design points and they will influence behavior over the long term. 
  • Pg. 186: Three things become difficult as you grow:
    • Communication
    • Common knowledge
    • Decision making

    The idea is to grow but degrade as slowly as possible along these dimensions. Three things can help a company cope:

    • Specialization of skills among staff: if you find that getting some one up to speed along all dimensions needed takes longer than just doing the work yourself, you need to invest in specialized talent.
    • Organizational design: all org designs are bad. You need to choose the least of all evils.
    • Process: create process as needed as the company scales. process = communication
  • Pg. 188: Choose an org design that best facilitates communication internally as well as externally with customers. Steps for org design:
    • Figure out what needs to be communicated: list the most important knowledge and who needs it
    • Figure out what need to be decided: what types of decisions need to be made most frequently and how can they be maximally grouped under a single person?
    • Prioritize the most important communication and decision paths: for example, I s it more important for product managers to understand product architecture or the market? Optimize for today’s needs- you can reorg in the future, if needed
    • Decide who’s going to run each group: make this decision subservient to the communication and decision needs of employees
    • Identify the paths that you did not optimize: don’t ignore these entirely
    • Build a plan for mitigating the issues identified in the previous step: patch the problems you’re knowingly taking on in the last step
  • Pg. 190: Process = communication
    Helpful things to keep in mind as you build processes:

    • Focus on the output first: what should the process produce?
    • Figure out how you’ll know if you are getting what you want at each step in the process
    • Engineer accountability into the system: which person and which group is responsible for each step? How van you increase the visibility of their performance?
    • Don’t add process too quickly or else the company will seem heavy and slow. Anticipate growth and design process for it proactively, but don’t over anticipate growth.
  • Pg. 214: Ones are strategic decision makers. Twos are execution focused. Some people are ones in their functional role and twos at the executive level. Ones who have other ones reporting to them can be counterproductive since each wants to set their own direction.
  • Pg. 219: Leadership has three traits:
    • An ability to articulate a vision: is it interesting, dynamic & compelling? Compelling enough for employees to stay even when it doesn’t make sense?
    • The right kind of ambition
    • The ability to achieve that vision

Why Invest in Research?

Research is a powerful business tool. Professionals throughout all organizational levels know this to be true, yet sometimes skip it. Common reasons include time, cost and a belief that the factors for success are already known. While I wouldn’t dismiss any of these reasons outright, my experience tells me they’re often myths. Research needn’t be a speed bump, budget buster, distraction or inconvenience. Rather, it’s a practice and mindset that comes with the following positive benefits.

Make better decisions
Whatever your skill set and role, researchers collectively bring strategic thinking, tactical skills, best practices and intuition to the table. Research brings another voice to the conversation — your target audience’s — that is objective, free of organizational bias and able to surface ideas that may otherwise not be seen or considered. These outside perspectives can profoundly change how the world is understood and, therefore, how it needs to be approached.

Reduce risk
Research allows project teams to better pinpoint where attention, time and money are best spent. A team that has a comprehensive understanding of problem(s) will also be a team of effective problem solvers. And while research does come with financial and timeline implications, it reduces or completely eliminates the greater costs incurred building problematic or failed products and services (not to mention the costs needed to fix them).

Fill knowledge gaps
Intuition, experience and personal observations get teams far. Research, in combination, takes them farther. It brings more voices to bear, more subtlety and nuance to what may otherwise be broad generalizations, and pointers when the path ahead is unclear. It also helps surface biases and cultural norms that may negatively impact the project’s success.

Improve organizational dynamics
Research can provide a framework to align people and departments through collaboration and a shared understanding of customers, their problems, and how well the organization is positioned to serve customers and address their problems. Research can rally a team by injecting it with purpose and inspiration.

Speed development
Some perceive research as slowing down a team’s progress. If speed to market is the goal, then this perception may be correct. But if the goal is to build the right product or service for the right people and for the right reasons, then research actually speeds the development process. Teams will ultimately reach success quicker with research than without it and will do so with fewer public stumbles.

If space for research is included in the design and development processes, teams gain an ability to effectively manage and apply their time, money and attention. The results are more likely to be exceptional products and services that customers are willing and happy to buy and use.

Personas and Jobs To Be Done

I read more articles over time decrying personas in light of the jobs to be done (JTBD) framework. I generally favor JTBD myself, but I, like others, are not ready to proclaim personas dead yet. Every tool has it’s time and place.

The issue that comes up in articles seems to center on the personas that promote a shallow characterization of customers. This is indeed an issue for the UX and product management world. However, personas come in many flavors and each has strengths and weaknesses. The question is not whether personas are good or bad generally, but whether or not they convey meaning for a specific purpose. The three persona types below each lie along various continuums of time and effort needed to create, depth of insight given, etc.

Proto persona
Proto personas are making inroads given the trend toward lean principles. They’re initially created from best guesses until customer contact provides information on which to revise them. This is acceptable as long as there’s follow through to ensure greater accuracy.

Demographic/marketing persona
These personas are the ones that rile people. They go something like this: Steve the Stereotype has a wife, two young kids, likes to golf, is 35-45, etc. These are acceptable for some marketing and advertising purposes when you’re looking to reach a block of people who share some similarities and, based on that similarity, can be reached in a consistent manner. ESPN reaches sports enthusiasts, for example. I wouldn’t use these for UX or product decisions. These are the types of personas I rail against too.

Design/user persona
Kim Goodwin has a robust, thorough methodology to create design/user personas (that fits within an equally robust and thorough product creation methodology). For me, it’s the gold standard of personas because they’re built upon solid research of real people. I reserve the right to question the addition of personality details, but those can help create empathy which isn’t a bad thing.

Empathy persona
If your intention is to create empathy among staff who don’t or can’t get regular customer contact, an empathy persona can offer a decent alternative. They’re not perfect, but they do provide a reminder that humans are the beneficiaries (or victims) of our product choices.

My preference is to create proto personas, though I would rarely share these with my client. They’re lightweight, quick and nimble which works for the equally lightweight, quick and nimble way our company works. They’re a means to an end in my workflow and needn’t be a milestone deliverable. Create ’em if you need ’em.

In reality, I’m more apt to filter the best from the design/user persona approach using jobs to be done thinking. I believe well crafted personas have a lot in common with JTBD analysis, it’s mainly a difference in focus: “According to Clayton Christensen, the customer is the wrong unit of analysis for innovators to focus on. Instead, focus on the job that customers are trying to get done when they use your product or service.” In effect, JTBD keeps the goals, needs and behavioral aspects of the best personas—the “what” and “why” of understanding customers—and drops most of the “who” information found in demographic data and filler background stories found in demographic/marketing personas. You end up wringing out the personal notes in order to distill and clarify jobs, situational awareness, desired outcome states and the causality, motivation and triggers that tie all these concepts together. In a sense, JTBD cuts to the chase which is useful in an agile workflow. It prioritizes information, focuses effort and clarifies the root issues at hand.

Of course, your mileage may vary.

Explore Pixar‘s Relationship with Their Customers

Ed Catmull, CEO of Pixar, was asked to comment on kids as consumers. Specifically, the question asked about Pixar’s views on the trends they see amongst their five-year-old demographic (skip to the 24:10 mark in the video). His answer:

Five year olds actually haven’t changed as much. Clearly, the teenage world has changed a lot more because of the way media is spread. So that’s actually the bigger change. For the children, we haven’t seen much. In terms of the way we think about our stories though, we don’t segment them in that sense. That is, we do make movies that children can enjoy but we also make films that we [adults] can enjoy. And we believe very strongly that children live in an adult world. So we want things in the films which they don’t understand. Not that we’re trying to do it. But what makes things interesting for children is they’re figuring out the world. And if you try to make something in a movie so that it’s all easy for them to figure out, you’re actually having a distortion of the world they operate in and that they like to be in.

Does his response indicate that Pixar takes a jobs-to-be-done (JTBD) approach to their films? It’s certainly a stretch to definitively say they do based on one question and answer exchange, but let’s dive in a bit as a thought experiment.

A tenet of JTBD is that segmenting markets by people (i.e. demographics) is the wrong unit of analysis when building products. Instead, the jobs and the  situational context within which they exist provide a clearer link between people and their actions. It would appear from the exchange that Pixar does not ask “what movie will five-year-olds like and want to see?” but rather “what job do people hire films to get done?” Are they hired to entertain? To act as cheap, temporary babysitters? To momentarily escape from normal, daily life?

We can analyze and debate films’ JTBD and the situations in which people hire them, but instead of going down that path let’s consider how Mr. Catmull thinks about it based on his Q&A session.

For Catmull, Pixar isn’t making movies for five-year-olds, he’s making movies that transcend the simplicity and one-size-fits-all mentality that comes with demographic segmentation. He acknowledges that children are customers, but also acknowledges that the end result doesn’t revolve around them to the exclusion of others. He’s more motivated to tell a story about the world we live in—a world in which kids and adults are participants. Therefore, Pixar is less interested in telling a story that only addresses the cognitive abilities of children and more interested in telling a universally accessible story about the world we all experience, young and old. This approach cuts across demographic segments making them an ineffective and counterproductive way to consider customers.

We could argue, upon deeper reflection, that a Pixar film’s JTBD is to explore and examine the human condition through the perspective of fictional characters that resemble the world, but are not of this world in order to reconnect with our own humanity. In essence, Pixar reminds us we’re human in a world that so often doesn’t treat us as such. If you buy that (and it’s not important that you do—this is an mental exercise after all), then it’s a small step to realize how such a job wouldn’t fit neatly into carefully considered demographic segments.

Catmull provides additional information for our exercise at various other points in the interview. At one point, he explains Pixar’s belief that personal investment on the part of a director is key to making a film special. He citesUp’s achievement in this regard. It went “…above and beyond the standard vocabulary that’s used in storytelling.” In essence, Pixar entrusts new product development (and the massive investment that comes with it) into a single person’s vision. And it’s expected that the director’s vision will connect with moviegoers on a deep level. We could translate that to mean this: Pixar expects their films to address the unmet or underserved needs amongst their customers to feel human instead of creating films to fit perceived opportunities in a demographically sliced and diced market.

Catmull also suggests that Pixar doesn’t focus group their films. Instead, directors periodically and regularly present their films to a brain trust of internal staff who give feedback and offer commentary. What this tells us is that, again, the vision that drives the film is not limited by superficial market segments. It succeeds because it addresses a latent need that exists in the world. You don’t have to be five to enjoy Up nor do you have to be 35. You don’t have to be male or female, wealthy or poor. None of that matters because the film taps a deeper link with people along the jobs-to-be-done spectrum that would otherwise be missed if demographic segmentation were the only yardstick.

What’s most interesting about the lead quote from Catmull’s above is that Pixar avoids the trap of considering itself a maker of films for kids which ultimately drives their success with them.

Book Review: Remote: Office Not Required

I read Remote: Office Not Required over the weekend (it’s a fast read) to see what I could learn about working with remotely located coworkers. I came away with a simple change of perception that will likely change the way I operate in the future.

Our company has two offices which, in a sense, makes us a remotely located group. I say ‘in a sense’ because those offices are, at most, 30 miles from one another. Before reading the book, the idea that this small distance made us a remotely located organization seemed far fetched. If I needed to talk to someone face-to-face, we’d simply go to the same office. Or, more often than not, the people who I interacted with most already worked from the same office I did- nothing remote about that, right? Well, now I think differently.

The book explains the simple concept that even if you’re across the street from your office working at a cafe, you’re effectively remote. People at the office can no longer physically walk up to you. Yes, they can walk across the street, but otherwise need to make contact via some other method, many of which are asynchronous and therefore don’t require immediate attention like a physical interruption does. That’s working remotely. It’s so simple that, in hindsight, I’m surprised I didn’t make the connection. Just because many of us are in the same office for lengthy periods of time doesn’t mean, as a whole, that we’re not a remotely oriented team. Just the opposite: many colleagues are not in the same office as me and therefore, each and every day, half or more of our staff are effectively remote. And then there are our clients who are hardly ever in our office.

The simple change in perception the book gave me is worth the price of admission. From there though, I didn’t get much tactical value. I found significant portions of the book covered workflow adaptations we’ve implemented and tools we already use. It was great to compare notes against what 37signals does and the other companies they highlight. I suppose some of that does have tactical value, but it seems minimal compared to the larger shift in mindset I took away.

I don’t have much more to say at this point other than a quick critique of the writing and format. Once again, 37signals has published a book that strikes me more like a collection of blog posts than a streamlined, cohesive narrative. This might leave a bad taste in your mouth, but it certainly makes for a quicker, more consolidated package. And who won’t benefit from a quick, weekend read with takeaways you can put to use the same day?