Insights from Paul Graham’s Essays
When I was in college, my economics advisor suggested that I write summaries of all of the nonfiction books I was reading in order to help the information sink in. I was doing an independent reading project with him and the summaries I wrote became the first posts on this blog.
Now that I'm a product manager, I'm doing a lot of independent reading again. It's not that I don't read regularly — I do. Reading is one of my favorite hobbies and I tend to read 1-2 books a week. But most of the books I'm reading now are nonfiction books I'm trying to learn from in order to be better at my job. I don't want to forget the information I've read, and so I figured I'd revive the practice of summarizing my independent reading here.
First up: Paul Graham's essays.
Paul Graham runs the start-up accelerator Y Combinator, which counts Dropbox, Airbnb and Justin.tv among its early-stage investments. He's also the former founder of ecommerce start-up Viaweb, which was acquired by Yahoo! in 1998 and may have been the first web application.
Graham's essays run the gamut from start-ups to public policy to programming languages to Silicon Valley culture to fundraising. Most of his essays are from the early-to-mid 2000s, so some of the specific technology trends he writes about can feel a bit dated now. But by and large his writing is evergreen, and his essays contain a surprising amount of career and life advice (and, let's face it: if you're working at a start-up, your career kind of is your life).
Here were my biggest takeaways:
“If you’re not embarrassed by the first version of your product, you’ve launched too late.” Actually, Reid Hoffman, not Paul Graham, said that. But while Hoffman coined the quote, Graham gives similar advice over and over again.
If you release early, then you can start learning from your users. On the other hand, if you launch only after you’ve executed according to a grand business plan, then you'll find out too late that many of your original ideas were flawed.Start-ups allow you to compress your working life into a few years. In a start-up, you work insane hours. But if you exit successfully after a few years, then you're set for life.
Of course, to try this, you need a fairly high tolerance for risk. Typically, only 1 in 10 start-ups succeed. But there's no better time to try than when you're in your 20s and don't have a family to support (or other life commitments).
Graham puts the turning point for most start-ups at "ramen profitability" — when the company is making enough money to pay the founders' meager living costs.Maker’s schedule, manager’s schedule. Makers need long, uninterrupted blocks of time to do creative work. A meeting in the middle of the afternoon can make the afternoon unproductive even if the meeting is just thirty minutes long.
Webapps allow you to release quickly, i.e., multiple times a day instead of every few months. Building Viacom as a webapp allowed the team to stay ahead of their competitors. Customers also loved calling about an issue and then seeing it fixed later that day.
(Side note on speed: Viacom was also built using LISP, a high-level and relatively obscure language. Graham writes how uncommon LISP was at the time — most software was written in C++ or Java. LISP's advantages as a programming language also helped Viacom outmaneuver their competitors.)Put developers in touch with users. At Viacom, Graham and other developers would take support calls from customers. Most companies — even small, 10-15 person companies — don't do this and instead protect their developers with a dedicated support team. But I like the idea: developers in touch with customers have better insight into their customers' needs.
Don’t die. The best way to succeed as a start-up is to simply not die. Stay scrappy.
Graham also has lots of advice for fundraising, if your start-up is at that stage.
Software Is Eating the World
In short, software is eating the world.
—Mark Andreessen
Oldie but goodie. In 2011, Andreessen was talking about how software development showed more potential than hardware development. In 2014, post-Web 2.0, his comment is just as relevant but could describe how every industry from education to taxi services is being disrupted — and in some cases gradually consumed — by software innovations.
15 Ways to Screw Up an IT Project
Jennifer Lonoff Schiff has a great round up of project management patterns to avoid over on CIO. In brief, the list is:
Having a poor or no statement of work.
Not setting expectations up front.
Not securing management buy-in.
Using the same methodology for all size projects.
Overloading team members.
Waiting or not wanting to share information.
Not having a clearly defined decision-making process.
Not using a project management software system.
Allowing scope creep (or excessive scope creep).
Being afraid to say "no."
Not being a team player.
Poor communication.
Too many, too long status meetings.
Not caring about quality--the "good enough" syndrome.
Not learning from past project management mistakes
But make sure to read the full post for comments and explanations from project managers in the trenches.
Merlin Mann on Credibility
Building credibility can start with something as simple as this: I say what's gonna happen, and then it happens. And then I say something else is gonna happen, and then it happens. And I keep doing that over and over and over . . . .
If you say, "I'm going to estimate how much it's going to cost to do this project, and when we can have it done by to the level of quality you want," and then you do it — I mean, I don't care if that's one day's worth of work. If you do that over and over, dude, you are on another level.
Merlin Mann touches on a core description of project management in Back to Work Episode #127. If you're not already listening to Back to Work regularly, change that because you should be.
Pagination vs. Scrollability
Pagination is a necessary evil when you have too many items to easily show them all on one screen. Linear content flows—such as articles like this—should almost never be broken up into multiple screens. It’s better to show the full article on one long screen than to inflict the pain of additional steps on users when all they want to do is read an article, and thus stay within that one item.
Where pagination comes in handy is for listings, such as e-commerce category pages, search engine results pages (SERP), article archives, and photo galleries. Here, a user’s goal is not to peruse the full list, but rather to find a specific item and click through to that destination page.
Assuming that you can prioritize the list items, users are likely to find what they want close to the top. To focus users’ attention and improve response time, you can start by showing a fairly short list, and then offer pagination options for progressing further down the list if needed.
Users' Pagination Preferences and 'View All', Jakob Nielsen’s Alertbox, April 28, 2013
This one of the reasons I prefer to read ebooks in apps which offer scroll functionality. An infinite scroll is far more immersive. Pagination can be nice when reading on an E Ink screen (which would have a slower refresh rate than an LCD), but every page flip is a necessary evil and a slight interruption.
FastCo interview with Marcus Buckingham, author of First, Break All the Rules
Every manager is different, but what they do differently you could sum up with that mantra I put in chapter two, "They don't try to put in what God left out; they try to draw out what God left in. That's hard enough."
Great interview over on Fast Company with author Marcus Buckingham, of First, Break All the Rules fame. I love his data-driven approach to analyzing business leadership.
4 Ways the Web is Changing Communications
The web is changing how we organize information in four ways:
Taxonomies (think the Dewey Decimal System or biological classification) used to be necessary to organize information, but the web recognizes that categories are often fluid. Would my hypertext novella be found in the nonfiction, new media, prose or poetry section of a bookstore? The answer is neither and all four—it depends on who is looking for the book. With hyperlinks and tags, the web's architecture is primarily relational.
We rarely enter websites via their splash pages and instead access the page we want via Google, Bing, or another search engine. Even though technology allows news to break faster and memes to spread virally, old information is not so much forgotten as pushed to the fringes. For instance, Google "Ryan Gosling" and three of the top six hits will be images from The Notebook, even though the movie is seven years old and Gosling has starred in a movie every year since (including 2010's Blue Valentine, which won two awards and was nominated for seventeen more). Since old information can be accessed as easily as new, the web is inherently nonlinear.
With a blockbuster movie, a prime-time television show, or even a print book, the dialogue is one-way—from artist to audience. But the web has given consumers a voice and it rewards them for using it. If web content does not facilitate audience participation, surfers will take their attention elsewhere. The web is by nature interactive.
Literary writing is largely absent from the web because designers have not yet found ways to humanistically display longer works. The scroll bar on the side of a browser window actually represents a technological regression (think Egyptian papyrus scrolls). Longer documents, unless intuitively organized, lose all the advantages of print books (the abilities to instantly jump from beginning to end and to see how far you have progressed in the text). Because screen reading can be somewhat uncomfortable, web writing tends to be digestible in shorter chunks (e.g., the length of a blog post).
For my creative writing senior project, I'm writing a prose-poetry hypertext novella—a book-length work organized as a website. My novella tries to takes advantage of the above properties. It's partly an homage to my pre-digital childhood and partly a collage of memories (our present encompasses our past). Think of it as a fictional Wikipedia or a digital choose-your-own-adventure—it's an experiment in nonlinear storytelling.
I'm hoping these guys will publish it.