Meltem Kuran Berkowitz
Head of Growth at Deel
Where to hear and read about Meltem Kuran Berkowitz1 episode or test
Meltem Kuran Berkowitz is Deel's head of growth, who joined as the second marketing hire in July 2020 and helped take ARR from under $1 million to $295 million while the company stayed EBITDA positive. The thing most worth learning from her is to answer buyers' real questions for free, on top of a product that delivers, and to tie every channel and hire to revenue before spending on ads or awareness.
The whole point of someone asking a question isn't to be sold a solution. It's like, I just need an answer.
An inside look at Deel’s unprecedented growth | Meltem Kuran Berkowitz (Head of Growth)
Meltem Kuran Berkowitz: the story and the lessons
1Meltem Kuran Berkowitz and the rise of Deel
Berkowitz joined Deel as its second marketing hire in July 2020. She was employee 19 or 20, and ARR was under $1M. Before Deel she led marketing at Bench Accounting.
Her role now covers paid ads, content, product marketing, community and brand.
Deel's ARR went from under $1M to $295M in about two and a half years. It reached $4M by January 2021, $57M at the end of 2021, $100M in April 2022 and $295M at the start of 2023, while staying EBITDA positive.
Ramp and Deel each held the "fastest growing" crown at different times.
She found a product that did what it said before she marketed it. She joined right after the Series A expecting duct tape in the back. She found engineers whose products did what they promised and more.
Early on, the marketing team had to tone down its claims because buyers did not believe the product could do all that. Her view is that B2B buyers won't put up with bad products, and no channel works if the product misses expectations.
She does not call Deel's growth a channel story. Deel started before the pandemic, and Covid pushed companies to test the global hiring it was built for. Word of mouth stopped being a large share of discovery within the first few months.
Buyers needed payroll or a way to hire overseas contractors, and Deel kept showing up in front of them. As Covid ended, her answer to doubters was that Deel is not a remote work platform but a global work platform.
2How Berkowitz uses free channels to reach buyers
She answers buyers' questions on Reddit and Quora from day one. Keyword alerts track questions about problems like hiring internationally, and real people on the team reply. Nothing is automated, and it costs only time.
The alert tool was a "janky" one someone else built, and she only plugged in the keywords. One answer stays online, so later readers keep finding it.
She solves the asker's problem first and mentions Deel second. The weakest reply is "we solved that, check out our website." The strong one explains the solution, says what is and isn't doable, then lets the person choose Deel or someone else.
Everyone at Deel, co-founders included, can see the questions, and the first to jump in flags it. On Twitter, once three people from Deel have answered, the rest stop.
She sizes a channel's audience ceiling before investing in it. A subreddit of about 1,000 people won't deliver 5,000 customers. It might deliver 10% of those who have the problem.
Content for a question nobody asks Google gets no traffic either. Her test for joining any community is whether people would want to talk to you.
She treats closed communities as harder to enter but less crowded. Slack and Discord groups and program networks work on the same value-first rule. Deel has partnerships with places like Y Combinator, where members get access after qualifying.
Posting without adding value gets you removed fast. Conversations in DMs or closed Slack don't stay visible, unlike public answers.
She picks partners that buyers trust for the specific question. Venture capital firms worked for Deel because new portfolio companies ask their VC which platforms to use right after a funding round.
A tech company with the same audience may never be asked that question. In her words: "just because you share an audience doesn't mean your audience goes to your partner for guidance."
Free channels shrank as a share of growth while still growing. Non-paid channels were 80% to 90% of growth early on and are about 50% today. Reddit went from about 15% of the funnel to under 5%.
Other channels outgrew them, but Reddit and Quora still reach more people each year.
3Berkowitz's SEO and content system at Deel
She judges every article by whether the reader's Google search is over. Search engines want people to get an answer quickly, so the test stops padding and keyword stuffing. Deel's employer of record page covers what an EOR is and isn't, plus its downsides and when not to use it.
She calls stuffing a keyword five times the biggest mistake. A keyword with 10,000 monthly visitors is not enough reason to write about it.
She ranks up to 700 keywords by buyer intent with a traffic light. Green is high intent, yellow is about 50:50, and red is a searcher who will never buy, like a student writing an essay. The team writes greens first, highest volume to lowest, then yellows, and often never reaches the reds.
The review is slow work and the team accepts that.
She reads what Google already serves before picking a keyword. Searching "EOR" returns enhanced oil recovery, so ranking for it as employer of record is hopeless. The "next questions" box shows what readers ask next.
Clearscope then grades the draft, and the target is a fourth or fifth grade reading level.
She refuses to cut corners on SEO. Tools save time, but keyword research done one by one never stops being time-consuming. Skipping it produces content nobody wants to read.
Senior people often treat it as beneath them. Her line: "it never stops being time-consuming."
Deel's content team publishes new articles and updates old ones every week. The team of about eight moved from about 10 net new articles a week to five new and five updated, because regulations change. It is roughly attributed half of Deel's growth.
A fact-checking team reviews older pieces, since a two-year-old article must still be current. The operation began in English and now runs in other languages. One operations person handles freelancers, briefs, fact checking, publishing and tracking, while specialists own product lines.
A thorough guide at a past company earned a link from the IRS. When the PPP program raised questions about who qualified and how to apply, the team learned the system and wrote a full guide for its own customers.
Traffic was huge for a few weeks, then died as PPP did. Her takeaway: when you answer questions others don't have time to answer, "they're going to send people back your way."
She tells consumer brands to skip SEO. SEO pays only where people go to Google to solve a problem, which covers most B2B. Shoppers for lipstick look on Instagram and with influencers.
A brand's own site is not an objective third party, so it rarely ranks for "best of" searches.
4Berkowitz's order of spend: site, paid ads, then awareness
She builds the skeleton before the makeup. She asks in order: is there a site, is it fast, can search engines find it, can people find it. Only then does she ask whether there is money for paid ads.
Paid ads fail if the site loads in four seconds or more.
She wants a one-liner that could not describe any other company. "We do the complex things so you can focus on what you do best" could describe about 90% of B2B firms. Deel's hard-coded site moved to a platform the growth lead could edit.
The team then A/B tested value propositions against each other: problem first, solution first, time savings and cost savings.
She adds many small long-tail ad outlets after the big platforms. Review sites, newsletter ads, podcast ads and niche sites with 50,000 to 100,000 monthly visitors add up to about 30% of Deel's lead flow.
The paid team spends as much time on these as on Facebook and Google. Ten podcast ads add up, though one alone will not bring 2,000 customers.
She refreshes ad creative every month. An ad that works for three weeks starts to tire people out. A few Google ads that stay straight to the point are the exception.
Messaging also has to keep up with how the product has evolved, even when an ad performs well.
She judges each channel on revenue a year out, not lead volume. A real-time dashboard built with the data team tracks lead to qualified opportunity to closed deal, plus revenue per customer one year out.
It sets how much Deel can spend to win a customer and shows when a channel has peaked, which settles the argument when sales say a channel's leads don't convert.
She holds off on awareness campaigns until the bottom of the funnel is tapped. Awareness work needs strategy and creative teams, may not resonate, and can burn a month. She expects six to eight months minimum of tapping bottom-of-funnel buyers first.
Her hypothesis is that Notion's late billboards worked because people already knew the product and its distinctive interface.
5How Berkowitz hires growth people and vets teams
She hires growth people who will commit to closed revenue goals. She asks which KPIs candidates committed to before and which they think they should have had. Someone who never held a revenue goal probably won't commit to one now.
Her case study asks for a strategy at $0, $10,000 and $100,000. A candidate who says that at $0 they can only monitor social media won't commit to bottom-funnel metrics until given every resource.
She hires for "little hands" and checks when big-company hires joined. She asks whether the candidate joined a large company early or arrived when it had 5,000 people and full resources. Someone used to ten times the resources may not do the dirty work.
The first answer to a small task should be "Yes, of course I'll do that," not "I would build a team for that."
She judges a young company's product by its team mix. Anyone considering a young company should ask for the team breakdown. Deel's was mostly product people and engineers, with marketing, sales and data built around them.
Six salespeople and two engineers is a bad sign, and an outsourced dev shop building the product is worse. Sales and marketing are there to carry the product, not patch it.
An inside look at Deel’s unprecedented growth | Meltem Kuran Berkowitz (Head of Growth)
Deel's head of growth explains how a B2B company grows by answering buyers' questions in communities, SEO and partnerships before paid ads and awareness campaigns. Deel went from under $1 million ARR in July 2020 to $295 million at the start of 2023 while staying EBITDA positive.
Berkowitz joined Deel as its second marketing hire in July 2020. She was employee 19 or 20, and ARR was under $1M. Before Deel she led marketing at Bench Accounting.
Her role now covers paid ads, content, product marketing, community and brand.
Deel's ARR went from under $1M to $295M in about two and a half years. It reached $4M by January 2021, $57M at the end of 2021, $100M in April 2022 and $295M at the start of 2023, while staying EBITDA positive.
Ramp and Deel each held the "fastest growing" crown at different times.
She found a product that did what it said before she marketed it. She joined right after the Series A expecting duct tape in the back. She found engineers whose products did what they promised and more.
Early on, the marketing team had to tone down its claims because buyers did not believe the product could do all that. Her view is that B2B buyers won't put up with bad products, and no channel works if the product misses expectations.
She does not call Deel's growth a channel story. Deel started before the pandemic, and Covid pushed companies to test the global hiring it was built for. Word of mouth stopped being a large share of discovery within the first few months.
Buyers needed payroll or a way to hire overseas contractors, and Deel kept showing up in front of them. As Covid ended, her answer to doubters was that Deel is not a remote work platform but a global work platform.
She answers buyers' questions on Reddit and Quora from day one. Keyword alerts track questions about problems like hiring internationally, and real people on the team reply. Nothing is automated, and it costs only time.
The alert tool was a "janky" one someone else built, and she only plugged in the keywords. One answer stays online, so later readers keep finding it.
She solves the asker's problem first and mentions Deel second. The weakest reply is "we solved that, check out our website." The strong one explains the solution, says what is and isn't doable, then lets the person choose Deel or someone else.
Everyone at Deel, co-founders included, can see the questions, and the first to jump in flags it. On Twitter, once three people from Deel have answered, the rest stop.
She sizes a channel's audience ceiling before investing in it. A subreddit of about 1,000 people won't deliver 5,000 customers. It might deliver 10% of those who have the problem.
Content for a question nobody asks Google gets no traffic either. Her test for joining any community is whether people would want to talk to you.
She treats closed communities as harder to enter but less crowded. Slack and Discord groups and program networks work on the same value-first rule. Deel has partnerships with places like Y Combinator, where members get access after qualifying.
Posting without adding value gets you removed fast. Conversations in DMs or closed Slack don't stay visible, unlike public answers.
She picks partners that buyers trust for the specific question. Venture capital firms worked for Deel because new portfolio companies ask their VC which platforms to use right after a funding round.
A tech company with the same audience may never be asked that question. In her words: "just because you share an audience doesn't mean your audience goes to your partner for guidance."
Free channels shrank as a share of growth while still growing. Non-paid channels were 80% to 90% of growth early on and are about 50% today. Reddit went from about 15% of the funnel to under 5%.
Other channels outgrew them, but Reddit and Quora still reach more people each year.
She judges every article by whether the reader's Google search is over. Search engines want people to get an answer quickly, so the test stops padding and keyword stuffing. Deel's employer of record page covers what an EOR is and isn't, plus its downsides and when not to use it.
She calls stuffing a keyword five times the biggest mistake. A keyword with 10,000 monthly visitors is not enough reason to write about it.
She ranks up to 700 keywords by buyer intent with a traffic light. Green is high intent, yellow is about 50:50, and red is a searcher who will never buy, like a student writing an essay. The team writes greens first, highest volume to lowest, then yellows, and often never reaches the reds.
The review is slow work and the team accepts that.
She reads what Google already serves before picking a keyword. Searching "EOR" returns enhanced oil recovery, so ranking for it as employer of record is hopeless. The "next questions" box shows what readers ask next.
Clearscope then grades the draft, and the target is a fourth or fifth grade reading level.
She refuses to cut corners on SEO. Tools save time, but keyword research done one by one never stops being time-consuming. Skipping it produces content nobody wants to read.
Senior people often treat it as beneath them. Her line: "it never stops being time-consuming."
Deel's content team publishes new articles and updates old ones every week. The team of about eight moved from about 10 net new articles a week to five new and five updated, because regulations change. It is roughly attributed half of Deel's growth.
A fact-checking team reviews older pieces, since a two-year-old article must still be current. The operation began in English and now runs in other languages. One operations person handles freelancers, briefs, fact checking, publishing and tracking, while specialists own product lines.
A thorough guide at a past company earned a link from the IRS. When the PPP program raised questions about who qualified and how to apply, the team learned the system and wrote a full guide for its own customers.
Traffic was huge for a few weeks, then died as PPP did. Her takeaway: when you answer questions others don't have time to answer, "they're going to send people back your way."
She tells consumer brands to skip SEO. SEO pays only where people go to Google to solve a problem, which covers most B2B. Shoppers for lipstick look on Instagram and with influencers.
A brand's own site is not an objective third party, so it rarely ranks for "best of" searches.
She builds the skeleton before the makeup. She asks in order: is there a site, is it fast, can search engines find it, can people find it. Only then does she ask whether there is money for paid ads.
Paid ads fail if the site loads in four seconds or more.
She wants a one-liner that could not describe any other company. "We do the complex things so you can focus on what you do best" could describe about 90% of B2B firms. Deel's hard-coded site moved to a platform the growth lead could edit.
The team then A/B tested value propositions against each other: problem first, solution first, time savings and cost savings.
She adds many small long-tail ad outlets after the big platforms. Review sites, newsletter ads, podcast ads and niche sites with 50,000 to 100,000 monthly visitors add up to about 30% of Deel's lead flow.
The paid team spends as much time on these as on Facebook and Google. Ten podcast ads add up, though one alone will not bring 2,000 customers.
She refreshes ad creative every month. An ad that works for three weeks starts to tire people out. A few Google ads that stay straight to the point are the exception.
Messaging also has to keep up with how the product has evolved, even when an ad performs well.
She judges each channel on revenue a year out, not lead volume. A real-time dashboard built with the data team tracks lead to qualified opportunity to closed deal, plus revenue per customer one year out.
It sets how much Deel can spend to win a customer and shows when a channel has peaked, which settles the argument when sales say a channel's leads don't convert.
She holds off on awareness campaigns until the bottom of the funnel is tapped. Awareness work needs strategy and creative teams, may not resonate, and can burn a month. She expects six to eight months minimum of tapping bottom-of-funnel buyers first.
Her hypothesis is that Notion's late billboards worked because people already knew the product and its distinctive interface.
She hires growth people who will commit to closed revenue goals. She asks which KPIs candidates committed to before and which they think they should have had. Someone who never held a revenue goal probably won't commit to one now.
Her case study asks for a strategy at $0, $10,000 and $100,000. A candidate who says that at $0 they can only monitor social media won't commit to bottom-funnel metrics until given every resource.
She hires for "little hands" and checks when big-company hires joined. She asks whether the candidate joined a large company early or arrived when it had 5,000 people and full resources. Someone used to ten times the resources may not do the dirty work.
The first answer to a small task should be "Yes, of course I'll do that," not "I would build a team for that."
She judges a young company's product by its team mix. Anyone considering a young company should ask for the team breakdown. Deel's was mostly product people and engineers, with marketing, sales and data built around them.
Six salespeople and two engineers is a bad sign, and an outsourced dev shop building the product is worse. Sales and marketing are there to carry the product, not patch it.