How to be valuable
I frequently do design exercises with engineers, designers, or product people I’m interviewing to see how well they think about systems. The exact prompt or exercise changes depending on the role, but no matter how you slice it, building a good system always ends up boiling down to managing three constraints: time, space (or physical materials), and cost.
You could argue that time is the only foundational resource here, since you can spend time to get more space or money, but since your time isn’t necessarily liquid or infinite, you can’t always make that trade.
The more I thought about it, the more I realized that building a good product is remarkably similar — to be valuable, it has to save its users time, space, or cost. I actually think this similarity between good computer systems and good products is why we’ve seen tech companies become so valuable — software does all three historically well, and products that fit into multiple categories are multiplicatively valuable.
This approach is useful for building new things from first principles or for understanding and improving market fit for existing products. Here’s what I look for to find value.
How does this save time?
I find time is arguably the most important resource, so I gravitate towards starting here. I think it’s tempting because it’s concrete and obvious: there’s value in doing something faster. If a bunch of horses at the track all go 40 mph, and you can make yours go 50 mph, you’ll win the race.
Before Google existed, finding sites on the internet was a painstakingly slow process since the web was indexed by hand. Now you can find exactly what you want in fractions of a second, with the most relevant results at the very top, and Google is worth almost $1.8 trillion. Email and text messages have supplanted letters and telegrams. Alquist uses 3D printing to build houses in three months, two-day shipping is Amazon’s selling point, and an entire industry of manufacturing robotics sprang up because they produce goods in a fraction of the time that humans can.
The most interesting version of saving time, in my opinion, is thinking about radically new solutions that can solve problems at speeds their predecessors couldn’t imagine. To borrow from Henry Ford, build a car, not a faster horse.
My favorite model for this is the history of computing. The computer was totally arcane in its early days, and if you weren’t a devoted student, you weren’t using one. Then in 1973, Xerox PARC built the first modern graphical interface. Their computer, the Alto, featured windows, icons, menus, and the cursor — all the things that made using a computer simple and commonplace. It didn’t make people faster programmers, but bypassed programming entirely and made people faster accountants or actuaries or business people by giving them programs they could learn within a matter of days. The Alto wasn’t a commercial success, but two small companiescalled Apple and Microsoft popularized the operating system and home computer, and the rest is history.
Education can be an insurmountable barrier to entry, and some of the most valuable products I know flatten the learning curve as a way to save time. They grant access to new knowledge or concepts (VisiCalc), answer difficult questions (Flexport), or provide outsourced experience (AWS). I’d even argue that Facebook belongs here too, which began its empire by instantly answering the burning question, “Is my crush single?” — or as a friend pointed out, the darker but equally profitable, “What is my ex doing right now?”
There are plenty of ways to identify time as a constraining resource. You can look for industry- or task-specific ways to save (e.g. higher bandwidth communication is a version of going faster), or come up with other equally fundamental questions, but these are a starting point.
Can it do the same task faster?
Can it reduce the amount of time spent waiting?
Can it multi-task to accomplish more in the same amount of time?
How does this save space or materials?
I use the term “space” loosely to refer to the physical constraints that people face. If time is the abstract concept, materials are the tangible concept. Mass, volume, and density are all relevant properties of matter here.
If you can do something in 1/10th the space or with 1/10th of the materials, you’re in business. Investors were tantalized by Theranos’ claim that blood work could be done with a tiny drop of blood rather than a whole vial. The medical industry as a whole likes getting smaller. Surgical robots and scoped instruments enable heart surgery through millimeter-sized incisions instead of having to crack open someone’s entire chest. Fentanyl is 100x more powerful than morphine, so you only need a tiny amount. All we’re missing is cancer-fighting nanobots.
One of the reasons the tech sector has boomed is because it digitizes physical goods and processes. Electrons take up considerably less space than paper. Making something digital is one of the easiest ways to save materials, plus digital materials are much easier to handle and reuse. AutoCAD is a lot cheaper than building a bunch of prototypes. CGI has saved Michael Bay millions of dollars on dynamite. Instagram is a heck of a lot easier to move around than photo albums full of 70 lb photo stock.
Some questions to help identify space or materials as a constraining resource:
Can it do the same task with fewer materials or in less space?
Can it make something lighter or more portable?
Can it turn physical goods into digital goods?
Can it consume less fuel?
How does this save costs?
Costs are universal. You can’t get something for nothing; in business, you’ve got to spend money, and in physics or chemistry, there’s always an energy cost.
The most direct way of saving costs is to offer the same product for cheaper. The IBM RAMAC 305, the first commercial computer, stored data on a disk drive that you could lease for $3,200 a month and held about 3MB. While that was later expanded to 5MB, that initial price comes out to about $1M per gigabyte of storage. Today, that same gigabyte will run you about 2 cents. Affordable storage has quietly defined progress since the birth of the hard disk drive; without it, there’s no internet, no personal computers, and no smartphones.
Alternatively, you can offer a much better product for the same cost. For $20, you can call a 1–800 number to get a taxi cab to the airport and ride in a sticky sedan from 1999 by an angry man demanding cash, or you can open Uber or Lyft and summon a brand new car to your doorstep that has bottled water and free candy in the back. Which sounds better to you?
Another viable course is to prevent people from having to buy the same thing twice. The space shuttle and its single-use rocket booster cost $1.5B per launch, or $54,500 per kg put into orbit. SpaceX’s Falcon 9 has the distinct advantage of not having its propulsion system incinerated during launch and costs $2,700 per kg. Geico prevents you from having to buy another car if someone totals you. Ring makes sure your home doesn’t get robbed. The entire repair industry capitalizes on this value, because nobody wants to buy another air conditioner or dishwasher if they can help it.
If pennies saved and pennies earned are equivalent, then helping someone make money falls under this category. This is actually advice you’ll find Paul Graham handing out on Twitter, and is the value proposition that has driven Stripe, one of his favorite startups, to meteoric heights.
Some questions to help identify cost as a constraining resource:
Can it do the same task for less money?
Can it do something better for the same cost?
Can it help people make money?
Can it reuse resources?
Can it distribute or share costs across a community?
A closing note on this framework
My last note is one of caution: this framework won’t prevent you from coming up with a solution in search of a problem, nor will it prevent you from optimizing one category at the cost of another. Washboard wanted to save time by mailing you bags of quarters for laundry, at a cost of 33 cents per quarter. Juicero wanted to save time by sending you pouches of precut fruit and vegetables for juice, at the cost of a $400 IoT juice press (which turned out not to work as fast as squeezing the pouch by hand). Reefill wanted to save you from both the repeated cost and wasted materials from buying bottled water by using their subscription-based service for tap water… which, as you might guess, is already available from the tap. They also overestimated people’s desire to carry around a reusable bottle at all times, which feels like it takes up more space to me personally than a bottle I can throw away once it’s empty.
Washboard was pushed out of business by on-demand laundry services that realized the pain point wasn’t looking for quarters, it was doing laundry in the first place. Juicero and Reefill shut down when they couldn’t provide more value than the incumbent ways to get juice or water.
The greatest products define value through the lens of their user, and are built with empathy and understanding. This framework is a good place to start, but don’t forget who your product needs to be valuable for.
Thank you to Josh Cole and Chris Granger for reading and editing this post!
