The evolution from a one-product to a multi-product company can be excruciating for any high-growth startup, especially—counterintuitively!—when the first product has been tremendously successful.
There are certainly examples of tech companies that maintain a single-product focus well into maturity. Many of these businesses eventually take a platform approach to solving adjacent customer problems—think Dropbox or DataDog—so the distinction between one-product and many blurs. However, for every successful single-product success story, there are countless examples of failures to translate the first hit into a second.
Here’s Steve Jobs describing the predicament, which he called “Second Product Syndrome”:
There is a classic thing in business which is the Second Product Syndrome, if you will. And that is, companies that have a really successful first product, but they don’t quite understand why that product was so successful. And their ambitions grow, and they get much more grandiose, and their second product fails. Believe or not, Apple was one of those companies. The Apple II, Apple’s first real product in the marketplace, was incredibly successful. And the Apple III was a dud. And so I’ve lived through that, and I’ve seen a lot of companies not make it through that.
That quote is from the documentary The Pixar Story (2007). Jobs was contrasting the Apple III with Pixar’s second movie, A Bug’s Life. He argued that Pixar was the rare example of a company that managed to avoid the pitfall. (But did they?1)
This problem isn’t unique to tech companies: investors are dubious of a VC firm’s second fund, TV network executives grumble about “second season downfall,” and musicians fret over “the sophomore slump” (or even the “third album curse”).
Why is it so hard?
Regression toward the mean, probably. Sustained performance at a high level is tough to maintain. After a breakthrough product launch or a #1 chart-topper, there’s nowhere to go but down. Not only that, everyone guns for you, expectations ramp up, and competitors abound.
We learn the wrong lessons from the first success. Jobs said that companies don’t truly understand why their first product succeeded and, as a result, carry the wrong lessons into the second. Sometimes executives credit success to their brilliance and assume their magical touch will lead to victory again. Other times managers don’t recognize how much market conditions outside their control played a role or how they may have benefited from competitive dynamics that have eroded. And what about plain luck? Take the time to understand why your first product was successful. That process can be humbling and may take multiple customer cycles.
The timing has to be right. Startups that are too slow to launch new products in adjacent problem areas or customer segments can get disrupted by upstarts or incumbents. But moving too fast to launch a second product may overwhelm a startup prematurely and divert attention from the still-nascent first product before it has traction.
Overconfidence can hurt. Companies can let overconfidence get the best of them and assume everything that led to success the first time will carry over to the second product. It’s easy to forget that the team that launched the first product was scrappier, had fewer resources, no customer obligations, and (usually) nothing to lose. It’s impossible to approach your second product exactly the way you built your first.
We burden second product teams. Startups launching their next product may also hinder product teams with complex integration requirements, back-office obligations, and fears of cannibalizing the first product. These are distractions the original product team never had. Here’s former Apple employee Steve Hix:
A huge amount of engineering effort went into making sure that the Apple II emulation mode of the Apple III was restricted to a subset of the Apple II’s expandability. Some in marketing were apparently certain that otherwise the $3,000+ Apple III would cannibalize sales of the much less expensive Apple II. (That never made sense to any of us.)
The perfect becomes the enemy of the good. Whereas Apple pushed the Apple III to market without proper testing, other teams may make the opposite mistake: over-engineering their second product to perfection, consequently losing out on the ability to learn fast and adapt. Even when the company is willing to make feature or quality tradeoffs for velocity, their customers—now accustomed to something more polished—may be displeased.
Of course, many of these are challenges that product teams at big companies deal with every day. But enormous incumbents compensate by throwing their weight around through marketing, earned media, and cross-promotion in ways upstarts can’t. The result is a perfect storm of misery: big company bureaucracy without the market muscle.
All of this is to say that building second products is damn hard.
🦘 I admire Atlassian and enjoyed this piece about how they grew the company and grappled with Second Product Syndrome: Unpacking 5 of Atlassian’s Most Unconventional Company-Building Moves. Atlassian launched their second product, Confluence, less than two years after shipping Jira. They attribute part of their success to quick lessons in cross-merchandising, cross-selling, and upselling. Of course, Atlassian’s HipChat would later lose to Slack and Slack’s single-product laser focus. As former Atlassian president Jay Simons puts it:
History is painted with some examples like this one where you’ve got Slack just ruthlessly focused on one particular thing and moving forward and moving faster and innovating, while we were trying to do a number of things simultaneously.
Atlassian eventually made the difficult decision to cede the space entirely, divesting HipChat to Slack. I imagine that having a portfolio of other profitable products made this decision easier to stomach. Again, this stuff is hard.
📚 I’m reminiscing about the mid-nineties recently after the news that Jeff Bezos is stepping down as Amazon CEO. Bezos’ annual letters to shareholders have always been must-reads for PMs,2 and his goodbye letter follows that trend:
Invention is the root of our success. … We’ve done crazy things together, and then made them normal. If you get it right, a few years after a surprising invention, the new thing has become normal. People yawn. And that yawn is the greatest compliment an inventor can receive. … Keep inventing, and don’t despair when at first the idea looks crazy. Remember to wander. Let curiosity be your compass. It remains Day 1.
acknowledge that Bezos’ complicated labor, social, and environmental legacy leaves me unsettled. Still, no one can doubt Bezos’ dogged adherence to a vision, his relentlessness as a leader, and his ability to inspire his people to create and invent.
Of all the recent pieces written about Bezos, Ben Thompson’s is the best.
There’s a new book about Amazon’s product development process called Working Backwards, named after Amazon’s famous “press releases before products” approach. I haven’t read it yet, but here’s a good interview with the authors.
Brad Stone’s second book about the company, Amazon Unbound, comes out in May. If it’s even half as good as its predecessor The Everything Store, it’ll be a must-read.
Keep chasing yawns.
🔴 Let’s stay on mid-nineties nostalgia. My first real internet job was at CNET: The Computer Network. I moved to San Francisco in 1996 to join the company as one of the earliest software engineers. It was a formative time for the company: an IPO and the transformation from TV network to web powerhouse. I sat in the San Francisco headquarters-slash-studio in the so-called “online room,” where I shared a workspace with about fifteen other people and the website. I mean that literally: we served all of CNET’s traffic off three Sun SPARCstation 20 “pizza boxes” racked in an open closet a few desks over from mine.
CNET was the first job I ever truly loved, and I was fortunate to have been there through the heady dot-com days. It set my career on the right trajectory with a strong brand and network. I experienced disciplined, systems-oriented software engineering,3 a rarity in early dot-coms. It was also at CNET that I first encountered talented designers and product managers. I probably would have had a long career there, but I ended up working on the team that built the Snap search engine. We were spun out of CNET into our own company after an NBC investment, and we would eventually become NBC Internet, where I served as CTO. I like to say I never left CNET; CNET left me. It will always hold a special place in my heart.
So naturally, I was delighted when CNET co-founder and eventual CEO Shelby Bonnie started a newsletter chronicling CNET’s history: No Blue: To CNET and Beyond. The title is an inside joke: those of us who worked in the main headquarters were essentially props in a TV studio, appearing in the background of shots. As a result, there was one important rule: we couldn’t have anything blue on our desks because it clashed with the brand. It’s a fun read for anyone who wants to peer back into the early days of a successful upstart, especially those who are curious about the dawning of the web and the dot-com boom.
(Everyone was confused about the pipe “|” in our name because they couldn’t find it on their keyboard. I distinctly remember complaints that emails kept bouncing because they stuck the pipe in the address. That’s why the company eventually switched from “c|net” to “CNET”4.
By all objective measures, A Bugs Life was a blockbuster. Grossing more than $363M worldwide, it was no Apple III-style flop. But with the benefit of history, it looks more like an example of Jobs’ Second Product Syndrome than an anomaly. It ranks 20th of 22 on Pixar’s list of worldwide box office revenues (I’m omitting Soul, which is too new). A Bugs’ Life also took the longest to bring to market. Three years passed between Toy Story’s release and A Bugs’ Life, the most time that has passed between any two Pixar films in the company’s history. Perhaps Steve Jobs’ “exception that proved the rule” was an example that proved the rule. ↩
Here’s a PDF of every Bezos letter over the years. ↩
Most of CNET’s early engineers were veterans of Bell Labs and Bellcore, and based in New Jersey. One of the earliest successful CMS platforms—Vignette—was a commercialized version of software we developed. CNET made a bundle on their IPO. ↩
After a career at “c|net,” “Snap!,” “Yahoo!,” and “JotSpot,” I recommend avoiding clever capitalization or punctuation tricks. ↩