超高速增长买不来：Atlassian 前营销一号位谈产品、口碑和选对下一家公司

Dietrich begins by saying the executive track is not for everyone, because it demands trade-offs against family, hobbies and other passions. When she decided early on that she wanted to be a CMO, she worked two hours later than her colleagues every day, reasoning that this added up to years of extra experience. She volunteered to run a product beta and do release engineering after a product head left before a major launch, which taught her how other departments worked.

She argues there are no shortcuts to knowing a lot, and cites VC Tomasz Tunguz, who woke at 5 or 6 AM to write several blog posts a week for ten years. Beyond personal effort, she lists five things to get right: tie your responsibilities to revenue, learn to speak the language of the CEO and board, take tours of duty in other functions, build strong relationships by doing good work and helping people, and choose strong companies.

The last point matters most for speed. At Oracle her team grew from five to seven over five years, while at Atlassian it grew from 15 to 100 in four. She also notes that great companies bring you great colleagues: she landed at Atlassian because she had worked twice before for its president, Jay Simons.

Asked how to spot the next big company, Dietrich calls it the billion-dollar question. After Atlassian she worked at Classy, which was acquired by GoFundMe and was fine but no Atlassian, and she spent five years working out how to pick the next one. Early in a career, she says, the simplest route is to go after big, high-momentum names such as Tesla, AWS, Salesforce or OpenAI, and to keep applying year after year if you have to.

Later in a career she prefers later-stage companies because their momentum is easier to read. Her Post-it checklist covers the Rule of 40, the quality and recency of investors, Net Promoter Score, net dollar retention, growth and burn rate, whether the company leads its market in analyst rankings, and its Glassdoor reviews. She points to Snowflake's net dollar retention of well over 100 percent as the benchmark for a healthy business, since existing customers alone nearly doubled revenue.

Her other Post-its work as a small life philosophy: "More Yoda, less Wonder Woman" means asking better questions, "Hell yes or no" filters opportunities by the energy they give you, and "Worrying is wasted energy" turns fear into a list of what you can control.

For people out of work, Dietrich says the market is tough and supply is high, so endurance and grit matter. The best executives she knows have had down periods or been fired, and what they share is that they got back in the ring. She also warns against taking a bad job or company, since each logo prepares you for the next role.

She describes chasing waves in her own career, moving from nonprofit work into tech, then B2B, SaaS, dev tools and collaboration, and now AI. Citing *The Millionaire Next Door*, she notes that industry choice matters: the best antique store makes a profit of one or two percent while the worst software companies can make far more. Her formula is to work your network, stick with it, and keep growing your skills.

For people who still have jobs, she says some of her biggest career growth came in downturns such as 2008, because others take their foot off the gas. With fewer people hired, there are open projects to raise your hand for. The catch is that you should not fixate on title and salary now, since you are building value to monetize on the other side.

Dietrich says she has spent hundreds of millions of dollars on awareness advertising. At Oracle that meant airports, the front page of the Wall Street Journal, the back page of The Economist, Iron Man sponsorships and arena deals, and at Atlassian a few million dollars on campaigns for Jira and HipChat. Her advice to founders is that product quality comes first: Oracle spent heavily and was still disliked by many people because of bad sales experiences and a stagnant product.

Good brand advertising must also be sustained. Oracle's boring red-bar template made it memorable, and market tests showed two to ten times more people recognized the brand without the logo than recognized competitors such as SAP or IBM. Snowflake is credited with heavy advertising, she says, but for many years it was a single billboard on Highway 101, which was consistent and strategic.

Atlassian's HipChat campaign, an *Office Space* spoof, did not help because the product had uptime and feature problems and Slack pulled ahead. The advertising only amplifies the product experience. The ads she remembers most have a twist of humor or truth, and many of her favorites, like New Relic's data nerd campaign or Salesforce's customer features, put customers at the center.

Asked what the fast-growing companies she advises have in common, Dietrich names three things. First is an amazing product people love, with ChatGPT as the most hypergrowth product ever. Second is organic, inbound and viral word of mouth, since you cannot pay enough to grow at those rates and stay viable. Third is "riding the lightning": these companies jump from 10 to 50 to 100 to 200 people so quickly that they need to hire leaders who have already seen the next stage.

She gives examples of products that spread naturally. Whiteboards were the most uploaded asset in Jira before Miro, and one person using Miro invites others onto the board. She has used 1Password for over a decade, from work to a family plan, and she became an advisor to Productboard after seeing a 1Password executive roll it out company-wide.

Founders without viral growth should not despair. Dietrich says it never felt easy at Atlassian, and the work was a slog of new features, customer listening and strong SEO. SEO was the number one marketing motion, and paid search generated roughly 15 to 25 percent of leads. Hypergrowth, she says, is consistency, customer obsession and incremental improvements, not a magic bullet.

To accelerate word of mouth, Dietrich recommends first empowering people inside the company to be thought leaders. Atlassian's founders were on developer boards early, and CEO Scott Farquhar was still taking part in dev forums when the company was around 150 to 200 million in revenue, before it created evangelist roles. She sees the same pattern at Weights & Biases, where technical staff write and engage with the community, and at HubSpot with its helpful content and certifications.

The second ingredient is a strong content strategy, often inside existing online and open source communities. The third is building your own community, because no one sells your product better than happy customers. She recalls events where customers and prospects were mixed so that customers spoke to each other, and Atlassian's user groups, where the company simply paid for beer and pizza and local leaders ran the meetups.

Lenny adds that the Gusto founder sat with the first 200 customers as they ran payroll, and Dietrich agrees this is the secret sauce: build processes that do not scale first. She warns that insight into customers' daily struggles gets harder as companies grow and product managers see only administrators or a few users.

Atlassian was unusual because its founders never took outside money apart from one late financing to give early employees liquidity, so no investors could overrule them. They redirected what would have been a sales budget into engineering and product. Dietrich shows a chart in which Atlassian had the lowest sales and marketing spend and the highest product spend among peers, and notes that the ratio still held in 2020.

Sales is the most expensive vehicle, she says, so the Atlassian model sold mainly to existing customers, helping with renewals and later enterprise deals, and rarely prospected for new ones. Most companies cannot copy this because investors expect accelerating growth and believe adding sales reps adds predictable revenue. The shift toward account-based marketing and intent signals tries to engage salespeople only with likely buyers. At Airtable and Miro, sales steps in after a threshold of users is already paying, sometimes 20 to 40 people at an account.

Her advice is to read the signals. If customers can get value within days, product-led growth works, and developers in particular dislike salespeople and like to research. If the product needs customization or the buyers do not buy online, sales is needed, though hiring too early is a drain. She likes a hybrid in which sales development reps become sales engineers who guide trial users, and Lenny notes that Retool had to go sales-led because customers needed hand-holding.

The main growth levers are consistent across companies: add a product, add a segment or vertical, go global, or add a channel such as partners. Dietrich says none of these can be a single department's goal. Moving upmarket requires marketing leads, product roadmap features, customer research, support staff who understand enterprises and salespeople who sell differently. When leads fall short, marketing or sales gets blamed, but it is usually a company strategy problem.

The change can come from the CEO or from the whole C-suite. With first-time founder CEOs, a junior marketer may not be believed, and an outside advisor can confirm the issue and help set company OKRs. At larger companies a seasoned CMO or head of engineering has the social capital to push back, for example to ask for time to address technical debt.

CMOs mostly get fired, she says, because they are not focused enough on revenue, and CEOs feel they lack a partner who knows what drives it. Marketing is a big spend category where much of the money does not convert within a quarter or even a year, so CMOs need strong metrics, solid predictions and the ability to talk like the CEO and board. She believes chief product officers face a similar fate because both are strategy jobs with results that are hard to measure, and she describes a CFO who got a CPO fired and then saw his own plan fail.

Asked about bundling, Dietrich says that in her view companies move from one product to many to diversify revenue, sell into a friendly base and expand their addressable market toward a billion dollars, so bundling is a valid growth strategy. But for product-led companies it is a poor land strategy, and Atlassian's bundled entry offers slowed the motion, so it returned to landing with a single high-velocity product.

The reason is that a customer evaluating several products cannot buy quickly online. Someone who needs issue tracking can try Jira and be ready in about seven days, while a bundle of Jira, Confluence and HipChat adds questions and leads to fewer conversions. On competing with an all-in-one such as Microsoft, she says the best-of-breed approach works only if the product is so much better that it justifies the extra cost, especially as CFOs squeeze budgets.

She closes the point with the secret of product-led growth: you can test it. Atlassian tried bundles, no bundles, different bundle contents and different trial lengths, and she says data-led insights beat any pundit on a podcast.

In the lightning round Dietrich recommends the Tao Te Ching in the Stephen Mitchell translation and *Never Split the Difference*, a negotiation book by a former CIA hostage negotiator. Its lesson for her is to get so deep into the other person's shoes that you can find a win for both sides. She links this to *How to Win Friends and Influence People*, which makes the same point that people want you to think about them.

Her favorite interview question is "How many people have you fired, and tell me about each experience?" Because she mostly hires managers of managers, how many people someone has fired shows how long and how widely they have managed. How candidates talk about layoffs, performance exits and restructures reveals their compassion and how they act under pressure.

She also admires the Amazon practice of writing the press release first, because it avoids handing marketing a "bag of doorknobs" when customers want the door. She advises meditators to stick with it, use guided sessions and accept there is no right way, and says she would love to work with more AI scale-ups through her blog at carilu.com or LinkedIn.