先搭骨架再化妆：Deel 从零做到近三亿美元年经常性收入的打法

Meltem describes Deel as a payroll, HR and compliance platform for global teams that helps companies hire employees and contractors across countries. She joined in July 2020 when Deel was below $1 million ARR, as the second marketing hire and roughly employee 19 or 20. By January 2021 the company was at $4 million ARR, finished 2021 at $57 million, reached $100 million in April 2022 and started the cited year at $295 million ARR. She emphasizes that Deel has also stayed EBITDA positive, which the team is proud of alongside the ARR growth. Lenny notes that Deel and Ramp have each held the fastest-growing SaaS crown at different points.

Lenny asks why Deel hired her around cheap growth channels and what worked. Meltem says most B2B companies exist because there is an active problem, and buyers want that problem solved rather than being sold to. The cheap channels are therefore places where people ask questions: SEO articles, Reddit communities, partnerships, existing peer communities and Quora. When a company answers those questions and presents a fitting solution, it can acquire customers without paying for advertising.

On Reddit, Meltem set up keyword tracking so the team could see when people asked relevant questions and then reply with useful advice. She did not pay for that visibility; the cost was time. Those answers, combined with word of mouth, live in public digital places and can keep being found by later searchers. Early on, this was a meaningful part of Deel’s growth even though each individual interaction looked small.

The Reddit tactic had a clear ceiling. Meltem warns that a subreddit of about 1,000 people will not produce 5,000 customers; the team might win 10% of the people with the problem. She also cautions against blind SEO: if nobody is asking the question on Google, well-written content will not be found. The audience size and existing demand must justify the investment.

Lenny connects this to companies that piggybacked on Craigslist, including Airbnb, Uber and Lyft, but Meltem’s rule is stricter than access. A team cannot enter a community and announce its product; it has to add value or members will ignore or remove it. The question-answer approach works only when the company behaves like someone people actually want to talk to. That means giving a real answer before mentioning the product.

Meltem set up keyword tracking on her first day in the role and Deel still answers questions on Reddit and Quora. Over time those channels fell from roughly 15% of the funnel to less than 5% as other channels grew, but the absolute number of people reached continued to rise. More broadly, about 50% of growth today comes from non-paid channels such as partnerships and SEO, down from roughly 80% to 90% early on. The interactions were never automated and still are not; team members and co-founders answer personally, and whoever sees a question first flags it, with sometimes several Deel people answering the same Twitter thread.

Beyond public forums, Meltem points to closed communities such as Slack groups, Discord servers and programs like Y Combinator. They are harder to access, but they can be valuable if a company can earn entry. The same rule applies once inside: contribute value first or the community will reject the pitch. Closed communities are not a shortcut around trust.

Meltem says the biggest SEO mistake is stuffing keywords and assuming that is enough. The better test is whether the Google search is over after someone reads the page. Search engines want to send people to the quickest satisfying answer, so a page that makes readers bounce back to Google is failing. Writing for that outcome produces content people actually want to read. She adds that SEO never stops being time-consuming, and senior people sometimes think the work is below them, which leads to corner-cutting and weak resources.

Deel’s content team uses a traffic-light system to choose topics. The team finds up to 700 related keywords and ranks them by volume, then reviews each keyword’s intent. Green means the searcher is likely to want Deel’s solution, yellow is a 50:50 buyer, and red means the searcher is not a buyer at all. The team works green keywords from highest to lowest volume, then yellow, and often never reaches red.

Before writing, the team studies what Google currently serves and what the searcher is trying to understand. Meltem’s example is EOR: Google interprets that query as enhanced oil recovery, not employer of record, so Deel cannot simply decide to rank for the acronym. The team uses the related questions at the bottom of Google results to map the next questions a reader will ask. It then uses tools such as Clearscope to check keyword placement and reading level, aiming for language a fifth grader can understand.

The content operation is about eight people, led by a director and supported by an operations owner who manages freelancers, briefs, fact-checking, publishing and tracking. Lenny notes that this roughly eight-person team accounts for about half of Deel’s growth, which Meltem confirms. Product-line experts write in their areas, and a newer group covers video, education and other formats. Deel used to publish about 10 net-new articles a week; now it publishes about five new articles and five updates because regulations change and old pages must stay current. Meltem also cautions that SEO is not for every company: it fits B2B problem-solving, but a lipstick brand should probably focus on Instagram and influencers instead of trying to rank for generic buying queries.

Meltem’s early prioritization advice is to build the skeleton before putting on the makeup. The first questions are whether the company has a website, whether it is fast, and whether search engines know it exists. Next comes whether people can find it and whether content is needed to make it findable. Only after those basics should a team consider spending money on paid ads, because paid traffic cannot rescue a site that loads in four-plus seconds.

Positioning is part of the skeleton. Meltem says B2B companies often write one-liners that sound good but could be pasted onto another company’s website. If the statement works just as well for another business, it is not clear enough. The copy team’s job is to make people understand what the company actually does, not to hide behind abstract language about doing complex things so customers can focus on what they do best.

Deel’s website was hard-coded at first, so the team moved it to a platform Meltem could edit and use for continuous A/B testing. They tested value propositions against each other, including problem-first versus solution-first framing and time-savings versus cost-savings messages. Rapid testing mattered because the team needed to learn which explanation made people act. This sequence kept growth work grounded in a usable, understandable website before scaling acquisition.

After covering the major ad platforms, Deel invested in the long tail. Meltem describes review sites, smaller outlets, newsletter ads and podcast ads as individually too small to matter much, but collectively about 30% of lead flow. Running 10 podcast ads may not produce 2,000 customers from one show, but the results add up. The paid team spends as much time setting up and running those smaller platforms as it does on Facebook and Google, and that diversification is deliberate.

Meltem is not a fan of early awareness campaigns for B2B businesses. Proper awareness work takes strategy and creative time, and the creative may not resonate, so a team can lose a month before learning anything. B2B companies usually have a bottom of funnel full of people ready to convert today, and she estimates it can take six to eight months to tap that demand before awareness makes sense. Awareness is not forbidden, but it is often the shiny campaign that distracts from nearer revenue.

Her example of awareness done later is Notion’s out-of-home advertising. By the time Notion ran those ads, many tech workers already knew the product and recognized its interface. The ads could therefore reinforce recognition rather than introduce an unknown brand. Meltem’s hypothesis is that the same layout on day one would have confused viewers because the familiarity had not been built yet.

On paid growth, the team focuses on both optimization and messaging. Creative fatigue is real, so ads that work for three weeks need refreshing, and Deel updates most ads monthly except for a few direct Google ads. Messaging also has to keep up as the product evolves, even if an ad is still performing. Most importantly, Deel measures what happens after the lead: conversion to qualified opportunity, closed-won revenue and one-year payback, with a data-team dashboard tracking those rates in real time.

Meltem says partnerships should be chosen by trust, not just audience overlap. There are two groups of potential partners, and a shared audience is only a decent starting point. The real question is whether the audience goes to that partner for guidance on the problem the company solves. For Deel, venture capital firms were a strong partner because portfolio companies ask their VCs what other companies use and trust; another tech company might share the same audience but never be asked that question.

The questions Deel answers are often nitty-gritty compliance and tax questions. An HR, finance or legal leader hiring abroad does not know what they do not know, so Deel keeps in-house experts who can answer with context from both the home country and the hiring country. Regulations change constantly, which means an answer from a week ago may need an update. Tax treatment also varies by country, role and whether the person is an employee or a contractor.

Meltem shares an earlier-role story from the start of Covid, when the US government launched the PPP program. Customers did not know whether they qualified or how to apply, so the team studied the rules and built a resource to help them. The resource became good enough that the IRS linked to it as a place to send people with questions. It drove a large but short-lived growth spike during the few weeks when PPP applications were urgent, and the traffic faded as the program did.

Meltem stresses that acquisition channels fail if the product does not meet expectations. She says she was spoiled at Deel because the product was strong from the day she joined after the Series A. The engineers and product team could actually deliver what they promised, and early pitching sometimes had to tone down claims because prospects did not believe Deel could do so much. A great marketer cannot save a product that disappoints, especially in B2B where customers leave and spread the word.

Her advice for people joining a young company is to ask about the team breakdown. When Meltem joined Deel, most of the team was product and engineering, which signaled that the core would be solid; marketing, sales and data were built around that core later. She warns that an early B2B company with six salespeople and two engineers may not keep a strong product for long. An outsourced dev shop that is expected to handle the product while the company sells is another warning sign.

Deel started before the pandemic, but Covid forced many companies into a remote-work pilot that validated the broader vision. People learned they could work well with colleagues in other countries and became more comfortable hiring the best person regardless of location. When people asked whether Deel would suffer as Covid ended, Meltem’s answer was that Deel is not a remote-work platform but a global-work platform. Customers might return to offices, but they now had offices in Germany, Canada, the US and France, so the need to hire and pay globally remained.

Meltem says word of mouth was important only very early and quickly became a smaller share of discovery. The deeper driver was urgent demand: companies needed payroll and a legal way to hire contractors overseas. Deel kept putting itself in front of those buyers and offering to help them do legally what they were already trying to do. The growth came from matching a persistent operational need, not from a single channel.

When hiring the early growth team, Meltem asks candidates which KPIs they are willing to commit to. A candidate who will only own lead volume is not enough; she wants people who will commit to closed-won revenue and the full funnel. She is also wary of hiring by big-company brand alone. If someone joined a huge company when it already had 5,000 people and every resource, they may not know how to operate with 10% of those resources.

Deel uses the idea of “little hands” to describe willingness to do the smallest, least glamorous work. Meltem says the phrase is loosely translated from French and means that no matter who someone is or where they sit, they get into the nitty-gritty. Interviewers ask whether a candidate will do the tiniest jobs; answering that they would build a team for that is not the right first response. The right first response is yes, of course.

To test revenue commitment, Meltem looks at the KPIs a candidate has actually owned in the past and asks what KPIs they have today versus what they think they should have. A case study can also ask for a strategy at $0, $10,000 and $100,000 to see how the candidate thinks about scaling and what they will commit to at each level. If someone at $0 can only propose social media monitoring and throws up their hands, that signals they will not be comfortable with bottom-funnel accountability. At Deel, growth teams are ultimately held to revenue goals rather than leads or traffic alone.