One thing.
Startups, focus, and the virtue of just doing one thing.
Big companies are the absolute worst. They’re inefficient. They’re timid. They’re slow. But worst of all, when confronted with the need to do something, before they do anything, they need to consider doing literally everything.
I’m sure you’ve seen it. BigCo needs to launch a new business line? First thing it does is an exhaustive, unending market study to size exactly what new product or service maximizes revenue, growth, margin, and share price.
A RetailCo needs to improve its in-store experience? First thing it does is hire a consulting firm to run a 12 month engagement to force 90 different VPs to prioritize - between their kitchen remodels and bass fishing trips - 381 potential initiatives with tidy little Harvey Balls1.
A ServicesCo needs to better streamline a broken set of processes? First thing it does is gather feature requirements from a dozen different stakeholders and then allocate 20 software engineers to spend 24 months building MegaAwesomeSolveItAllTool™, that will at best suck and at worst never be completed.
Everyone good and triggered now? Great. Solid.
Why I love growth companies is that, generally speaking, they just don’t operate like this. I’ve written before about the importance of a bias to action; essentially that what matters in startups is speed of execution. And that’s absolutely true. But what’s equally powerful, when faced with the daunting task of putting out 101 on-fire garbage cans is focus. That in a world of unlimited problems and severely limited resources you’ll help yourself immensely by forcing yourself, your team, and your company to focus on doing just one thing.
Get just one paying customer
Starting a new business is hard, with a whole lot of questions to answer. What should you name it? What should the logo be? What should you price at? What’s your refunds policy? Should you offer free shipping? What’s your snacks budget? Indeed, it’s the volume of these potential problems that grinds so many new businesses to a halt well before they get off the front page of a slide deck.
So, what if instead of drowning trying to answer all these questions, there was a better way to focus your efforts and successfully launch your business? Shockingly, there is. And that’s simply doing everything in your power to find one customer willing to pay more than zero dollars for whatever product or service you’re trying to sell, and then letting them tell you the next problem you need to solve. Case in point, the Shake Weight guy has a 120ft yacht in the Bahamas called “Iron Fist” all from a piece of exercise equipment you basically can’t show on network TV before 11pm. You better believe that man did not tie himself in knots with a 40 page feasibility study.
In 2017 DoorDash Drive - its white-labeled, 3rd party delivery offering - was in its infancy. It didn’t really have a team, and it didn’t really have a fully-baked product. That reality, however, did not stop us from selling the fine folks at Edible Arrangements on a massive Valentine’s Day promo, where DoorDash would power the delivery of their heart-shaped cantaloupe skewers to hopeless romantics all across the country. And let me tell you, it was, gate to gate, a hot, steaming nightmare. The platform crashed. Communication broke-down. Some drivers were assigned 19 Pineapple Passion Peonies™ while others were given just a single Sexed-Up Strawberries™. Noses were bloodied, eyes were blackened. My wife had to buy me a Cheer-up Cherries™.
Years later I - like an absolute %#^ing idiot - asked our COO during a leadership team meeting whether the learning from Edible Arrangements had been to be more diligent in our future new business launches.
The answer was, of course, a resounding ‘no’ (coupled with the rest of the team slowly turning off their Zoom videos like a bunch of old timey French villagers shuttering their windows just as the air raid siren goes off). On the contrary, by putting a new offering out to one new customer we’d learned an astounding amount about how we should sell, how we should price, how the logistics engine should be built, and how support should work. And that no number of months sharpening our pencils on a business plan could have come close to that volume of insight. And of course, as usual, the man was right. DoorDash Drive is now a zillion dollar business and I am here writing the 492nd most popular marketplace Substack.
Solve just one problem
I remember in my early days at DoorDash being hauled in front of its management team to explain how I was going to improve our dismal customer support scores (CSAT)2. And like a good ex-consultant I built a beautiful waterfall chart showing the 30 initiatives I was going to launch to boost us from ‘pretty bad’ up to ‘pretty ok’ over the next 3 months.
Dear reader, the feedback was not the Lando Norris champagne explosion I had hoped for. No, instead there was skepticism, bordering on contempt. “Steve, I’ll be honest, when I see plans like this, with a thousand little improvements, they never, ever, ever work.”
“If you could only do one of these, which would you do, starting tomorrow?”
I was aghast. ‘Just do one thing?’ I thought. That’s so chaotic. So unrefined. So undignified.
“Well, I guess I would try to steer more contacts from email support (which sucked) to phone support (which sucked less).”
“Okay, go do that, come back here in 2 weeks with an update.” Meeting adjourned.
And you know something, we went out, we focused on implementing that one change, and it showed a pretty meaningful improvement. We came back before the tribunal two weeks later, and, to our amazement, our inquisitors all had an expression on their faces that I was later informed was ‘happiness’. “Good work, go pick the next thing on your list, do that, come back in two weeks.”
The above was a great example of how to solve the ‘peanut buttering’ problem, the phenomenon where in trying to solve a massive problem you spread your scarce resources so thin that you exert maximum effort but make no real impact.
Solve every problem in just one place
Not sold? Alright, well there’s another way to solve the issue of peanut buttering; a concept called the Golden Market. The premise is that instead of solving a small part of a big problem (as above) you instead solve all of a big problem in a small place. Why? Because if you can channel all of your limited firepower at a much smaller surface area you stand a much better chance of making an impact. And if you do make an impact you’ve effectively changed the nature of your problem from a very challenging one of strategy (what’s the right solution?) to a much easier one of scale (how do we scale our solution to 100%?).
So when then do you opt for this approach? The Golden Market is especially helpful when you’re using manual human inputs to drive changes to manual human outputs. For instance, in trying to evaluate the impact of our customer service agent training on CSAT we started not with our entire 18,000 person workforce but instead just a single 400-person call center. In that way we could work with the team live, we could deliver more personalized coaching, we could spend more time answering questions. The result? We boosted CSAT 20 percentage points. This worked similarly for our logistics team in trying to understand how to bring down order cancellations. They focused on just a single, very small city where they could knock down the root causes of every individual order cancellation until they achieved a cancellation rate of basically zero. It was then just a much easier exercise of leveraging their learnings to deploy targeted, scalable solutions.
Build just one feature
I remember when DoorDash hired a new head of product in late 2017 I’d thought - as the perennially under-resourced head of CX - that my prayers had been answered. Our ill-equipped global support team was finally going to get all the tools and systems we’d dreamt of such that we’d never make another mistake again. I rocked up to our first meeting, and started in on my plans: “Okay, so we have to build Support Megasystem™ 1.0 which will do everything for everyone, perfectly, all the time.”
He took a beat.
“Okay, but like, what’s the one thing we can do tomorrow.”
My hope was smashed, like a kid who’d just dropped his ice cream cone. What was he talking about? I’d thought product was all about being visionary?
I thought for a while. “Well, I guess it would be great if agents could easily trigger an order redelivery in dispatch.” While not a massive issue, our inability to quickly redeliver a botched order to a customer who was taking incoming fire from hungry friends and family on a Friday night had been a longstanding gap.
“Okay, let’s do that then,” he said.
And a few weeks later, ding, a new button showed up in dispatch and our support team could do something new. Customer happiness went up. The townspeople rejoiced. And then we went back and did the same thing again, and again, and again.
What did I learn from this experience? A number of things, actually:
Prioritize your one thing; engineering is most-often a company’s most expensive, in-demand resource so when you finally find yourself with an opening to get something on the roadmap make sure you have a single, crisp, thoughtful ask. In a world where a product manager can always say “uh, this needs more thinking” and send you to the back of the line, it dramatically increases your odds of getting what you need to come prepared.
Momentum in product is real; there is value in knocking out small, quick wins at accelerating velocity and showing iterative impact. That impact shows there’s an ROI to investing in your asks, motivates more product investment, and so on.
Avoid the white elephant, the magic product solution that will solve all your problems but takes 12-24 months to get live. More often than not it fails to fully solve your customers’ problems, it takes too long, it gets too expensive, and ends up getting scrapped.
Build just one business
Finally, at the very highest level, where a lot of early-stage companies err is in getting too excited. You’ve got a great team, you’ve raised a bunch of money, you’re ready to conquer the world and… you immediately find yourself building a half-dozen businesses at once. You’re a SaaS company, you’re in fintech, you’re building product extensions, you’re expanding to new verticals, you’re launching Thailand, you’re liquidating your bitcoin to buy your .ai domain3.
Startups are marathons and as such the best ones diligently deploy their limited capital to methodically prove out the viability of their business in stages. Smarter people than I have spilled much ink on this topic, but, for example, at your seed stage you’re trying to prove that your business works in one market, at Series A that it works in multiple markets, at Series B that you have a repeatable growth engine to launch 30 markets, at Series C that you can win a country, and then at Series D and beyond that you can successfully expand into adjacent products, verticals, markets etc.
Like a top pro football team that’s focusing entirely on the current play, drive, and quarter, great operating teams focus on their current quarter, year and funding round, ensuring they’re doing only those things needed to achieve their current objective and graduate to play again with a new slug of capital.
The Harvey Ball, a circle with shaded quadrants meant to show progress to a specific criterion, was popularized in 1921 by famed arsonist Harvey T. Ball, who methodically burned down each of his four neighbors’ homes at the end of his Sheboygan, WI cul-de-sac.
You know that thing where a customer service agent does a great job and says “I’d really appreciate it if you’d hold for a brief post-call survey” and you say “absolutely Teresa, I’d be happy to”, and then you immediately hang-up because you’re just up to your eyeballs in Big Bang Theory and its rolling off Netflix in 2 weeks.
Not because there’s an AI play but because you want to launch in Anguilla.
YES! Your point about coming to the product team with too many requests (which sends you right to the back of the queue) is spot on!