The Need for Speed*
I caught up with a friend recently who has been pondering a startup for the past year or so. I still don’t know the exact idea; he wants to keep it secret lest a competitor take it and run with it.
Of course, in the past year, at least three other companies have started in his general space. Are they competitive? I have no idea, but that almost misses the point. My friend has missed out on a year of learning and execution. He is still pondering and talking while others are out building and making money!
Over the years, I’ve learned to value execution way more than ideas. Ideas are a dime a dozen; ideas without anything behind them (even if just a blog post!) tend to be worth very little.
But it’s not just execution that matters. Another friend has been grinding away at his company for well over a decade now without any breakthrough (or, indeed, making any money). It’s not just startups that can suffer from execution without a breakthrough—Microsoft has been working on Bing for well over fifteen years now without any meaningful breakthrough in market share, either. I’ve done the same with some of my past companies—challenging work and long hours, and sometimes without financial success.
Clearly then, simply working hard is not If not just execution, what then makes the difference?
Speed and Accuracy
Before Sam Altman was famous for leading OpenAI, he led Y Combinator for many years. For those of you who don’t know, Y Combinator is a startup incubator famous for companies like Airbnb, Instacart, and Dropbox.
The clip below is worth watching—over 5,000 companies have gone through Y Combinator, so the data set is meaningful. The short version—Y Combinator found that execution cadence was the most significant predictor of success:
As Sam points out, execution speed leads to iteration speed—continuously learning what works and what doesn’t and making improvements.
All things being equal, the company that moves and learns faster will succeed over the slower-moving company.
However, speed alone is not enough. Accuracy matters—and by accuracy, I mean really understanding and solving the customer problem. There are a number of different techniques for this; my favorite is Clayton Christensen’s “Jobs to Be Done.” methodology. This approach really focuses on understanding not what features and bells and whistles a customer might like, but rather, what “job” they are trying to do.
For example, streaming music services like Pandora and Spotify have almost entirely supplanted portable music devices like iPods (indeed, the iPod is no longer even sold!). Streaming music solves the “job” of “getting me energized for my workout with music” much better than futzing around with creating playlists, uploading and syncing music from my CDs, etc.
If you execute fast but are solving the wrong problem, you end up with many of Microsoft’s new AI-driven Copilot features. Microsoft was able to ship these capabilities with astounding speed, particularly for a company of its size. Some of these features, such as Github Copilot, have proven to be incredibly useful. Others, not so much:
I don’t want to chat with Windows to make it secure from hackers; I just want it to be secure!
As Mark Benioff of Salesforce recently quipped, Copilot is the new Clippy.
Now, Mark is out pitching the new AI tools his company is selling, so while this is mainly competitive ribbing, it still struck home. Mark is wrong, though, to dismiss Microsoft. As the old saying goes, “buy version 3 of a Microsoft product”. Microsoft is now on a swift learning curve; that proverbial version three may arrive sooner than Mark thinks!
Exponential Returns
Financially, speed and accuracy can yield incredible results due to something known as multiple expansion. Mathematically, this is a consequence of the Gordon Growth Model, but the insight is very straightforward.
Let’s say you have some shares of stock from your employer. The price of that stock multiplied by the number of shares is the nominal market capitalization of the company—or, more simply, “how much it’s worth.”
The most helpful way to look at the stock price then is to think of it as if you owned the entire company. To keep the math simple, let’s say the company makes a million dollars of profit every year. Thus, if you owned the company, you would make a million dollars a year. Pretty nice!
Now, let’s say someone wanted to buy that company from you. How much would you sell it for? Clearly, you’d like more than a million dollars! After all, you’d get a million dollars just by sitting tight and going to the beach for a year. What about $10 million dollars then?
Maybe!
As you reflect on this thought experiment, consider all the factors that were flowing through your mind. Well, what if the business is shaky and might not last ten years? What if it were actually shrinking and next year would only be making half a million in profit?
In that scenario, you’d probably quickly decide to sell for $10 million!
But what if the company was growing—and growing fast! Let’s say it was doubling in profit every year! In that case, you would quickly come to the conclusion that $10 million was far too little—after all, you would earn $10 million in profit in just over three years.
Countless people, from Wall Street firms to individual stock traders, try to figure out stock prices and trends every day, so there is no way to do justice to the full topic in a short blog post. However, I don’t have to—the important takeaway for our purposes is that fast-growing companies are worth a lot more than slow-growing or shrinking companies. And due to the mathematics of compounding growth, the returns can be exponential: doubling the growth rate can more than double the stock price.
Or, more simply, the faster you can grow a company, the faster you will make even more money.
How do you grow a company faster? With greater speed and greater accuracy in addressing customer needs.
Easier Said Than Done
That is, of course, a very lovely mathematical exercise, but what does it mean in practice?
I’ve been fortunate in my career to have worked with many different companies, from the very small to the very large. Many of these companies, particularly the technology companies, give their employees stock in some form as part of their compensation. If the company is successful, the value of the employee stock would significantly exceed the employee’s cash compensation.
Yet, despite this emphasis on stock compensation, it is exceedingly rare for me to see employees and executives executing in a way consistent with making more money on their stock!
I find the disconnect here fascinating.
The default instinct I see many people have is that to double the stock price; you have to double the profit. That seems like a tall hurdle for many people. But thanks to the compounding mathematics of growth, it is actually much more manageable. For many companies, simply increasing profit by ten to twenty percent—if done quickly—could double the value of the stock.
That is a lot more achievable!
Just about every function in an organization can contribute to speed and accuracy to gain an extra ten points of growth.
In sales, how long is the typical sales cycle? Days? Months? Could it be done faster? What would it mean to get an initial sale within minutes? This is a deep topic worth its own post, but in brief—it is usually much easier to upsell an existing, happy customer than it is to acquire a new one. Therefore, what is the simplest and quickest offering you can provide to turn a prospect into a customer? There is a reason Costco offers free food samples! What is your equivalent?
In marketing, how much reach would you get with paid advertisements versus viral but free social media marketing? Do you need to have an expensive, time-consuming, and highly produced photo shoot, or would a more authentic video blog—or even AI-edited video—do just as well?
In engineering, what percentage of your time is spent on work that directly benefits the customer versus everything else (legacy, infrastructure, boilerplates, etc.)? In software, there are a plethora of new services, from Vercel to Fauna to Clerk, that simplify undifferentiated but necessary capabilities like authentication and database multi-tenancy. Why spend hundreds of thousands of dollars and months of engineering time on a custom multi-tenant database solution when you can buy it from Fauna for $150 / month and get it in minutes?
Of course, this is a software example, but the same question applies to manufacturing and services. Do I really need a custom sales contract, or would a standardized form suffice?Last but not least, in the product team, the better you understand your customers, the better your company’s offerings will be.
Of course, none of these things can guarantee success, but they can definitely increase the likelihood of success.
As the final litmus test, no matter what your job is within a company, ask the question: if you started up your own company to compete, would you do the same work in the same way? Or would you do it smarter, faster, better, and more customer-focused in some way? If the latter, why not do it that way at your current company?
If you own stock in your company, figuring that out is the quickest way to increase its value and thus make more money for yourself.
Bottom line: speed and accuracy are everything.
In the inimitable words of Ricky Bobby, who wants to go fast?