top of page

Mode Media: The Story That Was Never Told

  • May 27
  • 32 min read

A Forensic Look: What really happened at Mode Media? Correcting how a $1B Silicon Valley unicorn was brought to its knees in six months and the untold story of how ten years later Mode in Japan not only survived but grew to operating profitably.


By Samir Arora



Thailand, September 2016. It is a long winding road from the ferry on Koh Pha-ngan. The road snakes up the moon island, shedding the tourists behind, until all you see is a glimpse of old Thailand — the water, the trees, the white sand. L’Alcove sat on the edge of the beach, the steaming sand waiting for the sun to set. The bay, a mirror. A guitar being tuned near the bar, a fish being grilled, the smell of Provence overpowered by spices in the kitchen, a warm glass of white Bandol with ice melting on the low wooden table. Nomadic before the remote norm.


I sat with my colored pens on my big Strathmore 300 watercolor paper notepad, as usual, in between troubled times, visioning my next future in AI and the wheel of life for health. The heat was heavy, and I had no choice but to surrender to the utter calm and sit in total peace.


Then my phone rang. +81. Japan.


It was Yusuke, the CEO of Mode Japan, that I had hired back in 2014. I hadn’t spoken to him in a long time. I picked up. He told me — Samir-san, they are shutting down Mode Media. I need your help.


My life flashed through me. The company I had helped found in 2002, and run as CEO from 2005 for a decade till I stepped down as chairman in March 2016. The tireless work of hundreds of employees, over 12,000+ creators/publishers and all major brands to build one of the largest digital companies was gone. The press was already writing the story.


It hit me that no one from Mode in the US had called me. I was learning that the company I had founded was being killed by my CEO in Tokyo eight time zones away who simply could not understand how this could be happening and reached out to ask for my help. I said yes, I’ll be there tomorrow.


The sun was going down over the bay. The string lights had fully come on. The boats were dark silhouettes now. I put down my iPhone. I took a deep breath. The kind you take before you go to a war. Not moving. A long exhale. Then I packed and flew to Japan.


This is the story that has never been told publicly. Why now? Ten years later, Mode in Japan is still operating, and has been taking care of its Japanese shareholders/investors and serving its content creators, publishers and employees since then. Lots has been written about this. But what really happened never got told. It took me a while to find out what went wrong. This is what actually happened.


EDITOR’S NOTE: Updated May 19, 2026 with documentary substantiation added. See end of piece.


What was Mode Media


Through its humble beginning in Silicon Valley with Accel (Jim Breyer, who introduced me to Theresia Ranzetta) and DFJ (Tim Draper) it had a phenomenal rise as the first vertical media company focused on women, then grew to multiple verticals based on the idea that the next wave of content would be driven by digital creators. By 2015, Mode Media reached more than 500 million global monthly visitors and was the sixth-largest publisher in the United States by reach, and third-largest in digital video streaming. It served eighty-five of the top one hundred US brand advertisers. Annual revenue was over $100 million. Total capital raised across the company’s twelve-year life was approximately $200 million, building a company valued reported at over $1.2 billion.[1]


A confidential S-1 was being prepared with Goldman Sachs as lead banker. A former Apple CFO led the finance team. A signed letter of intent was signed with Time, Inc. as a strategic investor.


What was reported but not understood was that the company had posted twenty consecutive quarters of year-over-year revenue growth from Q1 2007 through Q4 2011. And not ever reported externally, gross margin moved from under 10% ad-network levels in the early years to 62% premium tech platform (Facebook) levels by Q1 2015 (with content margins over 70%). North America reached EBITDA breakeven in Q4 2010 and full-year breakeven in 2011. Next, Mode in Japan reached breakeven in Q4 2011 and the full business operations achieved EBITDA profitability in 2012. This was all through one of the toughest recessions from 2007. (Source: audited company financials used for the confidential S-1 with lead Goldman Sachs, filed with the SEC)


Then the mid-2010s industry-wide technology-driven transition to programmatic advertising suddenly affected every major online media company. While I was there, through 2015, Mode added programmatic to its revenue mix with lower margins, but also succeeded in transforming by expanding to video, streaming, and premium content. A few quarters into the recession — most companies responded with layoffs and cuts too late. As TechCrunch reported in 2017, Samir took a 60 percent salary cut and protected the team through the downturn rather than letting people go. The senior team took graduated cuts below mine. The lowest-paid employees took zero. Nobody at Mode was laid off. The creators and publishers expanded. We took market share.


The result. We passed iVillage, NBC Universal’s women’s network valued over a billion — the number one in the space. Operating expenses went down and stayed nearly flat from 2008 through 2011 while revenue grew more than five times. Gross margin climbed to the top technology platforms level. Reduced expense and grew reach and revenue through the same period when iVillage and most of the category were contracting. By the end of 2011, the lead Mode had taken in 2007 had compounded into something no one in the category ever closed.


In contrast, when the team was terminated without severance in September 2016, and creators were left unpaid, that was a different decision made by completely different types of people. The leadership team that had survived 2008 by coming together was not the team that ran things in 2016.


At its heart, Mode was about the 12,000+ creators and influencers — writers, bloggers, photographers, designers, food and lifestyle editors, health, parenting, and women in media who had built audiences in their own voices and chose Mode as their platform. It had a few large publishers, even Oprah’s digital business at one point. And the first creator that grew to make a million in a year. That focus was the differentiation from Hearst and Condé Nast to the Facebooks that were user content driven. It was the reason the large brand advertisers came to us. It was the reason the audience came. It was the core reason the company existed.


Since Apple, all my platforms have empowered people from app development to the first web site builders, to creators and influencers. We didn’t just see the trend, we helped create the trend that these voices not only mattered, they were the creators and curators of the new media.


In November 2007, Jeff Jarvis — the media analyst who had been arguing for years that the future of media was networks rather than products — wrote about Glam in his column. “I have been arguing for as long as anyone would listen that the future of media is less about products and more about networks. It’s so nice to be proven right.” He documented the model directly: Glam had overtaken iVillage as the largest women’s network in the United States within eighteen months of launch, with 23 million uniques and 600 million monthly page views. Jeff Jarvis documented the creator economics: some Glam creators were making multiple six figures a year. He named what mattered most:“So in the end, Glam is really a platform. That’s the key… I say media companies should be asking what Glam would do. WWGD, the sequel.”[2]


The deep operating fabric of a company, its tradecraft — the product craftsmanship, customer relationships, the partners trust, the team’s institutional memory — is the work nobody sees, and it lives in specific critical people.


The Other Plan


Hubert Burda Media, the German publishing group, had grown its position to become a larger investor and was on the Mode Media board. Burda invested in several late growth stage rounds before September 2016 — growing to owning 15% of Mode. In 2016, after Samir And Marc left, by May 2016, Burda had grown to a majority of preferred and 52% of the fully diluted stock, effectively in full control.[3]


Martin Weiss joined Hubert Burda Media in January 2015 as Managing Director of Burda Digital and Burda Principal Investments. As the strategic disagreement sharpened through 2016 and Burda’s ownership escalated, Burda tasked Weiss with leading active board management of Mode and company plan— the effort that VentureBeat reported was internally called the “Mode Burda Plan.”[3]


Burda’s €1.13 billion National Digital Brands segment in 2015 was overwhelmingly e-commerce and marketplace revenue — Cyberport and computeruniverse (tech e-retail), HolidayCheck (travel booking), XING (professional network and recruiting, €122.9 million in 2015), BurdaDirect (CRM and commerce services), and Jameda (doctor appointments). The true digital media line — BurdaForward, operating Focus Online, Chip, Huffington Post Germany and similar properties — was a minority share of the segment. At $100 million revenue with 12,000+ creators, premium video, and an IPO process underway with Goldman Sachs, Mode Media’s standalone scale was comparable to or larger than Burda’s entire owned editorial digital media business at that time.[4]


Back in July 2015, at a Burda-hosted retreat at Lake Tegernsee in the Bavarian Alps — a lake locally nicknamed Bonzen-See for the wealthy clientele its shoreline historically drew — the Burda leadership drove a process to plan on how to react to the industry shift to programmatic advertising. The proposal required focus on a small number of very large publishers with volume, move away from the network of creators, and put in place the management processes needed to implement this strategy. The meeting at Tegernsee was the formal beginning of the nine-month conflict that followed that has not been talked about.


For context: 2014–2015 was the digital media bloodbath. Programmatic display CPMs collapsed across the category. Yahoo’s display ad business fell sharply and never recovered, selling to Verizon in July 2016 — down to 10% of its valuation. BuzzFeed and others missed their 2015 revenue target and cut 2016 forecasts. In contrast, Mode went into careful transition, changing its product mix and by Q4, 2015 had stabilized and reached its highest gross margins.


The strategic disagreement was not over whether to participate in programmatic. The successful transition was already in motion with the team. The disagreement was over the impact on the highest quality creator network — the very differentiation from Hearst and Condé Nast, the source of audience trust — who had the highest quality but too low volume and were the reason top brand advertisers came to Mode rather than to other commodity programmatic ad networks. A significant move to programmatic would have reduced revenue and margins.


The disagreement was also financial. Mode’s gross margin profile by 2016 resembled a premium owned-and-operated digital media company at Facebook levels, not the ad network it had been. A significant move to commodity programmatic would have reversed that. It was clear that Burda wanted to directly take control and select leadership that would support their plan.


A strategic redirect that looks correct at the surface that could collapse the financial stack the company spent ten years building. The numbers need to be the first conversation in strategy, not the last.


The heart of the issue inside Mode’s top executive team was that the Burda Plan would reverse that position. The strategy seemed sound on surface, but the detailed downstream mattered that this would lower Mode’s revenue and collapse gross margins and profitability, crippling the company. When we returned, we knew that though Burda meant well, their team’s strategy of how they wanted to help lead Mode vs. its Silicon Valley management to create processes would significantly hurt the value of the company. The non-Burda majority of directors and the venture investors had seen this before.


In September 2015, the Mode Media board — with the exception of the Burda-appointed directors — unanimously approved rejecting the Burda Plan, supported Samir and his team leadership and also asked him to engage Goldman Sachs to help.

From September 2015 through March 2016 — me, my team with our board directors, the venture investors, worked continuously on the company’s direction and opposed their Plan. This was not personal — months of sustained corporate-governance to keep Mode on its course — for the shareholders, company, employees, and the 12,000+ creators and publishers whose livelihoods depended on us. During the meeting at Tegernsee, I went to have lunch with Hubert at his home in Munich, and he asked me, “Will this work?”, to which I replied, “as long as I am there I will do everything to make it work.” We did everything in our power to protect our employees, creators and investors.


On September 27, 2015, the Silicon Valley venture investors signed and sent a formal written warning opposing the Burda Plan stating “it would hurt the value of the company”. The document was signed: Mode Media Venture Capital Investors. The text was later surfaced by VentureBeat.[3]


When during a major technology-driven shift, a founder-operator and one of leading venture board members are aligned on the future, what comes next decides everything.


Unlike Mode, when Vinod Khosla was in the same position with Sam Altman at OpenAI in November 2023 — he helped reverse the decision and OpenAI continued and went on to change the world.


When Burda subsequently moved to take control of the board in early 2016, the same investor group opposed the takeover and Marc stated his support for founder-entrepreneur operator in these situations as the heart of what he believes in as it would hurt the company and said he would resign if the plan was voted in.[3]


And true to his word, Marc resigned at that board meeting in protest. I resigned right after, following Marc. March 2016.[5]


What the hell happened in the six months after — what everyone wants to know


The question that haunts everyone was asked by Josh Stein at DFJ, the lead venture investor on Mode Media’s board through Heidi Roizen: how does a $100 million company with 60%+ gross margins and substantial accounts receivable collapse to under $40 million in revenue six months after the founder-operator CEO leaves?


A company at that scale doesn’t suddenly collapse for natural causes. Yes, it was difficult times. An industry under transition. Tight financials. But there were actual revenues from top brands. Committed revenues built over time. Even the Japan subsidiary was profitable. The IPO process with Goldman Sachs was substantively complete, waiting for the transition’s tail end. Something substantial has to have occurred in the six months between after Samir left and September 2016, given Burda had stated they would help support the plan with the management team they appointed after the takeover.


I had no direct access to the company financials or state from the day after I left. The account that follows is based on what was relayed to me at the time by the finance, legal and sales team including directly from John Small, the CFO Burda’s interim management selected after I left.


In February 2016, while we were still leading the defense of the company, the CFO Ernie Cicogna and his team had secured a critical $10 million financing commitment from Partners for Growth, the venture lending fund affiliated with Silicon Valley Bank. The financing was the bridge we needed as typically the highest revenue is in Q4, and payments come in Q2 the following year in the US.[5]


I learnt after I left that the Burda-controlled board did not approve the Partners for Growth financing.[3] In place of accepting the PFG financing, Burda’s installed their own management — choosing CEO Jack Rotolo over President & CRO Dan Lagani, and CTO Fernando Ruarte over the heads of Engineering and Products, and CFO John Small over Ernie Cicogna. Now in addition to Samir, the CFO and the top venture capitalists and independent Director exiting, the entire top operating team was also suddenly not running Mode. These choices shocked everyone — Jack had risen through Apple regional sales to Mode but had been passed as the head of revenue several times, and suddenly found himself at a level he had no operating experience — something we had seen at Apple and Silicon Valley companies. (“The Peter Principle”) What Jack should have said was, no the plan will not work, and I am not qualified to be CEO.


When a portfolio company’s installed management makes a working-capital decision that seems unusual, call the founder. Even an exited founder. An hour of their time can prevent a hundred-million-dollar mistake.


Per John’s account: In addition to saying no to the PFG financing, they moved the AR line relationship away from its reliable longstanding partner, to a completely new AR vendor. Though Gary Effren pulled the trigger at the BOD in voting with Burda, this is what created the cash crisis. When needed, the new vendor’s line did not show up and Mode simultaneously had to repay the existing line in September. The entire cash of the US company, including the new investment by Burda, suddenly just disappeared.


The AR vendor a company chooses is not a back-office finance decision. It is a load-bearing piece of the capital structure. Stable banking and debt relationships in Silicon Valley for top entrepreneurs and venture capital firms exist for a reason.

This is not the first time this has occurred — multiple Silicon Valley companies faced similar mechanics with this and other AR vendors that promised aggressively. The result in every case is the same: Silicon Valley ventures worked because it had a system of stable multi-tiered financial companies with incredible people working tirelessly to support the innovation that understand how startups work.


When Jack Rotolo delivered his revised plan to Burda in September — the annual revenue collapsed from over $100 million to $40 million for the year following Samir’s exit.[3] Typically Q4 was a significant (60%+) part of annual booked revenue, 2016 alone should have generated $60M+ to Mode. The machine that took over 10 years to build had been stopped in 6 months.


Days later, on Thursday afternoon, September 15, 2016, the Burda-controlled company and board approved the shutdown of Mode Media. All US employees were terminated. No severance was offered. No healthcare cobra. More than 12,000 creators were left unpaid for work already delivered. Only some in Germany were paid later.[6][7][8] With over $37 million in accounts receivable on its books at the time of shutdown. And Mode Japan was profitable with its own accounts receivables.[5]


This shutdown was not just a failure due to the industry changes or a struggling underlying business in transition. It was a colossal failure of managing the capital structure and operating decisions made in the six months that brought a unicorn to its knees. Though Burda was a majority and fully controlled the Board and appointed the team that ran Mode at that time, Burda’s spokesman went on the record with the New York Post stating: as an investor, Burda is not responsible for this.[5]

Looking back, before the takeover, a Burda director retained a forensic firm — to examine the company and Samir Arora’s conduct as CEO. The review took approximately six months. Their firm found nothing. Not one expense was tied to the flamboyant founder framing that ran in the initial press. What they found instead was the internal financial and legal rigor of a corporation getting ready for an IPO and a CEO that had every financial and legal detail that had already been rigorously audited. The only resolution the Board passed was to have finance add an additional level for approval for Industry events and awards. This was a seasoned company with a team ready for an IPO, with multiple rounds with the SEC, top auditors and law firms.[3]


The press


The first wave of coverage in September 2016 ran a version of the story without any of these facts — $225 million burned, mismanagement. Business Insider, then owned by the German publisher Axel Springer SE, ran the first and most-cited piece. The piece was never updated and never corrected as new facts emerged. Ten years later, it remains a result when someone searches for Mode Media.[9]


Many other journalists did the actual reporting and got the record substantially right at the time:


Blaise Zerega, VentureBeat, March 28, 2017.[3] Reconstructed the full sequence — the Burda Plan, the September 2015 written investor opposition, the board takeover, the Andreessen resignation, the blocked rescue financing, the revenue collapse under installed management.


Anthony Ha, TechCrunch, March 21, 2017.[13] The first major mainstream tech press correction to the founder-spending framing, found records of founder managing fiscally through crisis, and reporting on the Mode Japan rescue.


Keith Kelly, New York Post, September 22, 2016.[5] Surfaced the Burda spokesman quote, the blocked Partners for Growth financing, and the board mechanics of the takeover.


The first wave of coverage can’t always get it right. Easy to sling mud — It takes time to dig for facts. Adam Lashinsky did it back in my NetObjects days — “So, how could I have gotten this so wrong?”[10] Mike Arrington edited his post to add “Matt Marshall at Venture Beat writes a very long post coming to a completely different conclusion. I’m not sold, but judge for yourself. Different viewpoints are good,” and went on to even posting an origin story on Samir at Apple of the work that led to the first iPod.[11] TechCrunch went from strongly negative coverage of Glam in 2008[12] to a corrective piece on Mode Japan in 2017[13]. Sometimes the public record simply takes time.


The wire instruction


Back to where we started — a few days after the September 15 shutdown, Yusuke called me. He told me that Mode in the US was being shut down and was instructed to wire all the cash from the Mode Japan JV’s bank account back to the United States and stop all operations.


Everyone else in this situation did. Except him.


The reality did not match what he was looking at. Mode had a great brand. Signed deals. Revenue still to deliver. AR still to collect. Real obligations to real partners. He did not panic. He just did not know how to make it work. He picked up the phone and called me with one question:


This does not make sense. Can you help?


Yusuke is a deep operator looking at the fundamentals. Signed deals. Revenue to deliver. AR to collect. A working business with real obligations to real customers and partners. A JV with fiduciary responsibilities. Even under a shutdown, trying to preserve value for creditors, purchasers and investors, employees and partners. This doesn’t match that. He did not panic. It just didn’t make sense to him. He observed and reached out to me.


Most people in his position would have sent the wire and shut down, like the rest of Mode’s CEOs and team did. Yusuke had the reflex to look at the fundamentals first. That reflex is rare. It is what separates operators from administrators. I first met Yusuke Akiba back in 2013 when he was interviewing for the CEO role at Mode Media Japan. Born in Tokyo, an executive at an international brand company within Dentsu, decades inside the Japanese institutional brands and publishing world.

I told him: Wait. I dropped everything I was doing and flew to Tokyo the next day.

He waited.


The rescue


Mode Media investors and shareholders unanimously asked Samir to return to save what could be saved.[13] I created Montaro to try to help save Mode Media Japan. The US was already gone — but we were able to save Japan and Asia.[14]

On January 5, 2017, Montaro purchased all of Mode Media Japan’s business, and assets. The shareholders unanimously elected Samir Arora as Executive Chairman, Yusuke Akiba remained as CEO, and Minako Matsushita as CFO. Operations continued for millions of monthly active users. The Japan JV creators, partners, investors and shareholders were protected.[14]


Sherwood Partners handled the assignment for the benefit of creditors on the US side. BrideClick acquired the US Mode Media assets in June 2017 to relaunch as Glam.com. In 2022, Static Media acquired and relaunched the core Glam.com and Foodie.com domains, which continue to operate today.[15][16]


Though it was already too late for the US, with the support of Yusuke, the management team, the shareholders and the investors, we were able to help save Mode Media Japan.


It took us 6 weeks to stabilize Mode in Japan in 2017.


Ten years


Today. It’s been 10 years since the transition.


The result: served over 2,500 brands, satisfaction at 96.5% and delivered on institutional brand goals 97.3% of the time.

What happened next didn’t make any news. No one wrote about how a significant part of a company under one management did not survive. Another, in Japan did, and now has had a decade of running a business well, in a brutal market. That is what hard operators do. They look at hard facts in a business in transition. They ask for help. They listen. And no matter how hard it is, they show up. They keep at it. Non-stop. They deliver across the years results that do not generate headlines. And, almost ten years later, when the work feels complete, they make the call to say it worked.

It is the last stage of Mode I think about most.


My responsibility


As the founder and former CEO of Mode Media, though I was not involved in the direct decisions that led to its sad demise and the horrible way the shutdown was executed, it is important to take responsibility and look back to the lessons learnt.

The strategic differences over the strategy of the company, managing through a technology driven industry transition and the move to larger publishers were real. All decisions I made leading up to that are mine and I bear full responsibility for them. I did everything in my power to protect and defend the company, employees and its creators while I was the CEO to the very end. I, along with our US investors, lost that fight. I resigned because I no longer could change the future as Chairman.


The blocked rescue financing, the AR-vendor decision that destroyed working capital, the revenue collapse, the September 15 shutdown — all of it happened after Marc and I had resigned. The forensic financial firm Burda hired found nothing. The contemporaneous reporting got the record substantially right, the venture investors’ wisdom and opposition is in the formal record.


Two phone calls from that period still stay with me.


Two weeks after the board vote that handed control to Burda, Gary Effren called me. He had been a Mode director. He had cast the vote with Burda in the BOD meeting that allowed an investor that was not a majority to take over our company. On his call he said, Samir, I am sorry. I made a mistake. I feel misled. As an independent director, I should not have voted with them. My wife saw me troubled at home today and asked me, did someone die? He wanted me to know. Gary did not make the New Plan. He did not run the company into the ground in the six months that followed. But the vote in that room was the vote that changed everything, and Gary pulled the trigger. He has had to live with that decision.


Bill Campbell taught me to keep boards as small and effective as possible. To choose your independent board directors as carefully as you choose your investors, and choose them for the stage you are at. An independent director who is right at a Series A stage may not be right for an IPO. Get this wrong and the votes that could change everything could be cast by people who do not have the context to cast them.


A few months after the September 15 shutdown, an employee from Mode reached me. He said, Samir, I am having a really hard time. When they shut the business I lost my health insurance, and my wife may die if I do not get help. That conversation was one of many. Many of the people terminated and creators that day reached out in the weeks and months that followed. I read every message. I responded to help as many as I could. I have not forgotten any of them. I deeply felt their loss. I kept thinking– what else could I have done?


Results of actions matter. I have carried it for ten years. It is mine to carry.


To the creators and employees: I know deeply that an apology doesn’t cover a medical bill or pay a mortgage. Or losing the company that helped you do what you are best at — create great content. The survival of Mode in Japan remains the proof that your work had value. We proved that the model and the company weren’t broken; the US market changes and its stewardship was. If the US had called me earlier and used what we did in Japan — Tell everyone what’s going on and involve them in the process, I believe we would have found a way. A platform is only as strong as the people who build upon it.


In Japan, as in my past, I was direct with all employees, creators and Japanese investors — that mutual trust helped us save Mode Japan. I remember reading that Jensen Huang did the same with Sega in 1996, when NVIDIA couldn’t deliver the Dreamcast chip. He flew to Tokyo, told CEO Irimajiri the truth, and asked for the contract payment anyway. Irimajiri trusted him. NVIDIA survived. Being direct and humble makes the difference.


Burda — and what kind of failure this was


From the start, Dr. Hubert Burda and I had a good personal relationship, and we even called Jacob our adopted german-son when he came and stayed with us in Woodside. Over the years Dr. Paul-Bernhard Kallen also had built a professional relationship — especially given his challenging task to steer an incredible print company with a long term business that was significantly declining. These were real relationships between people based on respect, and I do not want this account to be misread as a story about bad people or personal failure. It was not.


What happened at Mode was a result of significant late stage investment in a tech industry that changes rapidly, with corporate strategic decisions made at the surface operating levels that did not fully grasp the ways Silicon Valley companies scale. The new Plan looked good from a distance. It did not survive contact with the deep operating realities of a very large scale creator driven platform, a 12,000+ partners ecosystem, a top brand-advertiser sales machine built across a decade, an industry in transition, a technology platform drive company with a very high gross-margin business that depended on every layer of that stack.


The pressure on Burda’s leadership during this period was also real. Burda was one of large investors in Mode Media at the growth stage, with significant capital committed across multiple rounds, and their own German publishing business was navigating the long industry transition from print to digital. They needed and wanted Mode to work. But the deeper issue was not the pressure. It was the speed. Print declines slowly. A legacy publisher gets years of warning, year-over-year declines they can plan around, multiple budget cycles to adjust. That is the operating tempo Burda’s executive team had been shaped by for decades. Silicon Valley does not move at that tempo. The programmatic shift hit Mode and the rest of the digital media category in quarters, not years. Decisions about how to position the company had to be made in months, executed in months, and revised daily. Nothing in the legacy publishing world prepares an executive team for that speed — or for the operational depth required to navigate it inside a US digital media company at scale. That is part of what happened here. It does not change the fact that the Plan was wrong. It does explain why the Plan was proposed.

This wasn’t a clash of people, but a clash of clocks. In legacy publishing, you have years to adjust to a 5% decline. In Silicon Valley, a technology shift like programmatic — or today’s Generative AI — hits in months. When an “administrator” tempo is applied to an “operator” crisis, the friction doesn’t just slow the company down; it incinerates the capital stack or even the company.


This is one of the most documented failure patterns in corporate strategy literature. Roger Martin’s The Execution Trap in the Harvard Business Review, July 2010, makes the point directly: drawing a line between strategy and execution almost guarantees failure, because a strategy that does not account for the operating choices required to realize it is not actually a strategy. That is what happened here. The Plan was a strategic thesis from one side of the table. The operating implications — for the creators, for the brand-advertiser relationships, for the gross-margin stack, for the team that would have to execute it — were on the other side of the table. The two sides did not meet. The vote went one way. The six months that followed were the operating reality reasserting itself.


The speed of what happened next is also documented. Recent HBR research from ghSMART, Leading After the Founder, January–February 2026, finds that founder-CEO transitions carry a risk of failure or performance downturn two to three times greater than transitions involving non-founder CEOs. The reason is deep operating knowledge. A founder-operator has spent a decade or more inside the specific operating fabric of the company — the customer relationships, the publisher relationships, the team’s institutional memory, the financial mechanics of how each revenue layer works in practice. A successor brought in from outside that fabric level, even one with strong general or functional experience, cannot replicate that operating knowledge in the time the business gives them.


There are moments for founder-operators to move to a strategic role, a deep technology shift is not one of those.

The most widely-known example of this pattern in technology is Apple between 1993 and 1997. After Michael Spindler, Gil Amelio was appointed CEO in 1996 — though he was a highly experienced executive, was not from inside Apple’s operating fabric. By the summer of 1997, after just eighteen months under Amelio, Apple had less than ninety days of cash remaining. The company was rescued by Bill Gates and Microsoft’s $150 million investment, announced by a returned Steve Jobs at Macworld Boston in August 1997. The rescue gave Apple the runway to reset under founder leadership. The point of the comparison is not the rescue. The point is the speed. A successful technology company with a strong brand, a global audience, and real operating capability can move from health to ninety days of cash in less than two years when the operator-CEO is gone and the leadership installed in their place lacks the specific operating fit the business requires. That is what the HBR succession research describes. It is also what occurred at Mode in six months.


Marc and Ben built Andreessen Horowitz on a powerful conviction: the founder–operator. The firm’s founding thesis, in Ben’s words, was “what if we build a firm that enables the founder to be a CEO?”


The Apple-Amelio pattern, the ghSMART research, and what happened at Mode are all variations on the same observation: the operating fabric of a fast-moving technology company lives in the founder-operator, and successor CEOs brought in from outside that fabric carry materially higher failure risk during industry transitions.


We are now in the largest technology transition since the internet itself — the AI-driven shift across every industry. The pattern matters more in 2026 than it did in 2016.


These are seasoned executives in another industry. The corporate decision was not correct for ours. Both of those things are true.


I am often asked what I have learned from the multiple Unicorns years. I learned that Control is Context. Focus on building governance, watch for the leadership fabric with deep domain ‘tradecraft’ required to navigate the technology shift without losing the deep operating learning.


The company that survived and grew in Japan remains a North Star for our work today. It proved that even in a total market transition, remaining focused on deterministic facts and human trust can save and grow a business. Ten years later, the lessons are clear: The operators who stay in the deep fabric of the business are the ones who can help lead it through the fire.


Yusuke and the team in Tokyo


Yusuke Akiba is one of the best operators I have worked with in any company, in any country. He has done it quietly and with dignity. That is also a key part of who he is. The Mode Media Japan team — Keizo Mishina, Minako Matsushita, and the leadership we built along with our partners and investors — are the reason this story exists. I remember John Doerr telling me “Remember, what does not kill you, makes you stronger.” And the bond formed between us became forged like steel in absolute trust.


They saved a company that the world had been told was dead. They built it into something profitable, durable, and trusted.

I am grateful to all of them. The work they did saved much of what could be saved.


Ten years later — the industry


Just before I left, in 2015, Mode had changed its revenue mix — display advertising fell from over 60% to 27%, while premium and video grew to 62% of revenue, driven by short videos, streaming, and premium content. Now, ten years later, by 2025, the revenue gap between video giants and traditional content-site display ads has widened as consumer attention migrated to short-form and creator-led video. Globally, YouTube became the largest video content company, TikTok the leading short-form video platform. While digital grew 1000%+, print magazine publishing declined 91% to approximately $3 billion. Outside of Japan, no other digital media ever became as large as Mode in lifestyle globally.


Closing


This is the story I have never told publicly. It took time to find the facts. I have told it now because it’s time. Mode Media Japan is in its tenth year of operations. It worked.


I am grateful to Yusuke Akiba and the team in Tokyo. I am grateful to the venture capital investors who stood with me and the company. I am grateful to Marc Andreessen for resigning from the board in protest when it mattered. I am grateful to Heidi Roizen for creating the September 2015 letter and the early support from Tim Draper at DFJ and Jim Breyer and Theresia Ranzetta at Accel that helped us build Mode. I am also grateful to Anthony Ha, Matt Marshall, Blaise Zerega, John Motavalli and Keith Kelly and all the other journalists for digging in and doing the actual reporting at the time.


To the creators and employees who carried the cost of what happened in 2016 — I am sorry. The work you did mattered. You all mattered. Many of the brightest moments of the Mode Media years belonged to you. I have not forgotten.


The company that survived in Tokyo remains proof that the original vision was right, that the original team was capable, and that the strategic concerns we raised in writing in 2015 were not wrong. The work that lived deserves a public record. This is that record.


The version that became the public memory of Mode Media is not the version that actually happened. This is what actually happened.


Samir Arora

Silicon Valley, 2026


This is the long-form companion to Hard Operators #2. The Hard Operators series examines specific moments where the difference between operator and administrator becomes visible. Hard Operators #1: The Fourth Kind of Apple CEO.


Postscript


Eight years later, another arc closes. Martin Weiss came from management consulting. Trained in a career that can provide an extraordinary outsider perspective but can also be Strategy from the surface. No operating experience running a technology platform business at scale. I had seen this first hand at Apple at the CEOs level before. In 2015 he joined Burda to run Burda Digital and BurdaPrincipal Investments. Gary Effren cast the swing vote at the BOD to support Burda. Martin was assigned by Burda to take over Mode. The team installed by them was not seasoned. Sales declined. The financial decisions made killed the company. In 2017 he took the Burda Management Board seat for international print and digital and in 2020, became Chairman of New Work SE — parent of XING. On January 1, 2022, he was named CEO of Hubert Burda Media. Two years later, on January 10, 2024, the Verwaltungsrat announced his departure: “his outlook on the direction and steps of the company’s further development is no longer aligned.”[17] Seven months later, on August 26, 2024, New Work SE was delisted from the Frankfurt Stock Exchange. The squeeze-out of minority shareholders followed in June 2025.[18]


References


[1] Forbes, April 28, 2015 — George Anders. “Larger Than Twitter Or LinkedIn, Mode Media Hollers For Respect.” https://www.forbes.com/sites/georgeanders/2015/04/28/larger-than-twitter-or-linkedin-mode-media-hollers-for-respect/

[2] BuzzMachine, November 12, 2007 — Jeff Jarvis. “Glam: The success of the network.” https://buzzmachine.com/2007/11/12/glam-the-success-of-the-network/

[3] VentureBeat, March 28, 2017 — Blaise Zerega. “Mode Media: An IPO that never happened and a company that won’t die.” https://venturebeat.com/technology/mode-media-an-ipo-that-never-happened-and-a-company-that-wont-die

[4] Hubert Burda Media Holding KG. Financial Statements 2016. Shareholdings disclosure (Sec. 296 HGB / Sec. 313 (2) no. 4 HGB), listing Mode Media Corporation, Brisbane/USA. https://d1epvft2eg9h7o.cloudfront.net/filer_public/83/9d/839d3c8c-e680-4b8c-b322-b26487dbac26/burda_financial_statements_2016.pdf

[5] New York Post, September 22, 2016 — Keith Kelly. “Mode Media enters next chapter after abrupt collapse.” https://nypost.com/2016/09/22/mode-media-enters-next-chapter-after-abrupt-collapse/

[6] Wall Street Journal, September 15, 2016 — Deborah Gage. Mode Media shutdown coverage.

[7] Recode, September 15, 2016 — Peter Kafka. “Mode Media/Glam shuts down.” https://www.recode.net/2016/9/15/12936470/mode-media-glam-shuts-down

[8] Vox, September 16, 2016. Mode Media shutdown coverage.

[9] Business Insider, September 16, 2016 — Nathan McAlone and Eugene Kim. Mode Media shutdown coverage.

[10] TheStreet, March 22, 2000 — Adam Lashinsky. “NetObjects Defies Prediction.” https://www.thestreet.com/opinion/netobjects-defies-prediction-904775

[11] TechCrunch, October 27, 2009 — Michael Arrington. “Exclusive Picture Of Unlaunched Apple Tablet (circa 1990).” https://techcrunch.com/2009/10/27/exclusive-picture-of-unlaunched-apple-tablet-circa-1990/

[12] TechCrunch, May 29, 2008. “How Beautiful Is Glam?” https://techcrunch.com/2008/05/29/how-beautiful-is-glam/

[13] TechCrunch, March 21, 2017 — Anthony Ha. “Mode Media Japan lives on, with founder Samir Arora as chairman.” https://techcrunch.com/2017/03/21/mode-media-japan-samir-arora/

[15] TechCrunch, June 18, 2017. “BrideClick acquires Mode Media.” https://techcrunch.com/2017/06/18/brideclick-acquires-mode-media/

[16] VentureBeat, June 19, 2017. “BrideClick buys shuttered Mode Media to relaunch as Glam.” https://venturebeat.com/technology/brideclick-buys-shuttered-mode-media-to-relaunch-as-glam

[17] HORIZONT, January 10, 2024. “Überraschender Abgang: Vorstandschef Martin Weiss verlässt Burda.”

[18] Freshfields Bruckhaus Deringer, June 3, 2024. “Freshfields advises New Work on the conclusion of a Delisting Agreement with Burda Digital.”


UPDATE —Silicon Valley, May 19, 2026. A former Mode colleague reminded me to add a reference to a May 20, 2016 email from CEO Jack Rotolo to the Mode board. Jack’s words to the board in the email are: “The switch from SVB to FastPay created a timing difference in cash flows” and “We were un-aware and miscalculated this timing difference…” The original email documents working capital reduction of approximately $20 million in AR-backed financing and average deal size collapsing from $150K to $45K — a 70% per-deal decline.


Press coverage by stage


Mode Media coverage by stage. Titles are hyperlinked in the published Medium version.


Stage 1 — Early press: The NetObjects, Glam, and Mode rise (1996–2015)


1996. Fortune (Janice Maloney). 25 Very Cool Companies (NetObjects).

April 7, 1997. The New Yorker (Malcolm Gladwell). Just Ask For It (NetObjects).

August 25, 1997. BusinessWeek (Amy Cortese). From bagging groceries to instant success that took 10 years.

March 22, 2000. TheStreet (Adam Lashinsky). NetObjects Defies Prediction.

September 17, 2007. Forbes (Claire Cain Miller). Samir Arora has assembled the largest audience of women on the Web.

November 12, 2007. BuzzMachine (Jeff Jarvis). Glam: The success of the network.

May 29, 2008. TechCrunch. How Beautiful Is Glam?.

2008. Wall Street Journal (Kevin Delaney). Glam Media to Raise Big Money.

2008. Wall Street Journal (Emily Steel). Glam Media.

February 2, 2010. TechCrunch. Glam Media On A Roll: Raises $50 Million In Private Equity At $750 Million Valuation.

September 24, 2010. TechCrunch (Leena Rao). Glam Media Continues Hiring Spree.

October 10, 2010. Crain’s New York Business. Glam Media: biggest Web outfit you’ve never heard of.

The Times London (Andrew Davidson). Is this the future for media on the internet?

Fast Company (Lizette Chapman). The Next Media Giant?

January 2012. Wired (Felix Salmon). How sharing disrupts media.

April 29, 2014. VentureBeat. Glam rebrands as Mode Media, dives into streaming video.

April 29, 2014. TechCrunch. Before Rebranding As Mode Media, Glam Raised An Additional $15M.

April 28, 2015. Forbes (George Anders). Larger Than Twitter Or LinkedIn, Mode Media Hollers For Respect.


Stage 2 — First-wave shutdown coverage (September 2016)


September 15, 2016. Wall Street Journal (Deborah Gage). Mode Media, Once a Unicorn, Shuts Down Operations.

September 16, 2016. Business Insider (Nathan McAlone, Eugene Kim). Inside Mode Media’s end — never updated, never corrected.

Stage 3 — Corrective record (September 2016 — March 2017)


September 21, 2016. MediaPost (John Motavalli). Strike a Pose, Mode Media.

September 22, 2016. New York Post (Keith Kelly). Mode Media enters next chapter after abrupt collapse.

March 21, 2017. TechCrunch (Anthony Ha). Mode Media Japan lives on, with founder Samir Arora as chairman.

March 28, 2017. VentureBeat (Blaise Zerega). Mode Media: An IPO that never happened and a company that won’t die.


Stage 4 — Post-shutdown trajectory (2017–2026)

June 18, 2017. TechCrunch. BrideClick acquires Mode Media.

December 7, 2017. TechCrunch (Anthony Ha). Samir Arora unveils Sage Digital.

July 12, 2019. TechCrunch (Anthony Ha). Sage Plus for Experts.

January 21, 2020. BuzzMachine (Jeff Jarvis). The Next Net: Expertise.

July 21, 2020. TechCrunch. Marcus Samuelsson and Samir Arora launch Project Bento.

August 26, 2020. Hollywood Reporter. WME Signs Samir Arora and His Sage Digital Tech Company.

Wikipedia. Samir Arora.

Wikipedia. Mode Media.

Hubert Burda Media Holding KG. Financial Statements 2016. Shareholdings disclosure (Sec. 296 HGB / Sec. 313 (2) no. 4 HGB), listing Mode Media Corporation, Brisbane/USA.

Mode Media was founded as Project Y, the changed its name to Glam Media, Then to Mode Media.


Killing the Golden Goose. The Mode IPO Frenzy.

For financial readers, investment bankers, growth stage investors and anyone that skips the management section and reads the financial first in a red herring IPO prospectus.


The IPO process


April 2011. Twenty top investment banks pitched us.

Bulge bracket. Goldman Sachs. Morgan Stanley. Bank of America Merrill Lynch. Citi. Credit Suisse. Deutsche Bank. Barclays Capital.

Major full-service and advisory. Allen & Company. Jefferies. RBC Capital Markets. Blackstone. Lazard. Guggenheim.

Mid-tier and tech-and-media boutiques. Cowen & Co. Oppenheimer. Stifel Nicolaus. Piper Jaffray. Pacific Crest Securities. Canaccord Genuity. Think Equity.

We selected five: Goldman Sachs as lead (Nick Giovanni, head of TMT). Allen & Company (Nancy Peretsman). Bank of America Merrill Lynch, Barclays, and Piper Jaffray.


Why I waited


A less seasoned team without IPO experience would have gone public into the 2011 frenzy. The fundamental technology-driven transition to programmatic had started. A public Mode would have been forced into quarterly earnings cycles managing through an industry structural shift. Margin compression and revenue mix instability would have been brutal in a quarterly environment. These transitions are best done as a private company.


Stripe today is the operator-discipline example: one of the most valuable private fintech companies in the world, public-market ready for years, still private. BuzzFeed is the counter-counter-example: went public via a SPAC at $1.5 billion in December 2021, collapsed below $100 million within three years. Vice declared bankruptcy. Yahoo and AOL never recovered. The digital media companies that were public during the transition rather than through it became the public-market evidence for why waiting was the correct discipline.


The S-1 financial record, 2005–2011


Revenue 2005–2011: $0.03M to $97M. CAGR: 62.2%. Twenty consecutive quarters of year-over-year growth Q1 2007 through Q4 2011. Through a recession. North America EBITDA breakeven Q4 2010 and full-year 2011. Japan breakeven Q4 2011 and full-business EBITDA profitable 2012.



Mode Media consolidated revenue, as submitted for the Confidential S-1, May 2012.


The most important fact: Quarterly gross margin Q1 2008 to Q4 2011: grew from 1% to 48%.


Most Ad–networks with TAC were at 10–15%. The 48% placed Mode Media at the YouTube premium platform band. It then grew to over 60%, at the highest Facebook platform band level by 2015.



Quarterly gross margin as submitted for the Confidential S-1, May 2012. YouTube premium platform band (~47%) shown for reference.


The financial architecture by 2015


By 2014, Mode Media’s annual revenue had exceeded $100 million. The harder number was the margin. When I joined as full-time CEO, gross margin net of COGS, content, and distribution was under 10%. The Q4 2011 endpoint shown in the S-1 was 48%. From there, gross margin continued the quarterly march — approximately one percentage point per quarter — reaching 62% blended by Q1 2015, with content-line margin at 70%. At $10M in revenue, $1M was available to fund engineering, sales, and operations. At $100M, $62M was. EBITDA positive.


For comparison: Facebook’s gross margin during its own growth through the $100M-$300M revenue stage (2007–2008) was in the 50–60% range. Mode at $100M+ revenue was operating at gross margin economics comparable to or above where Facebook was at the same revenue scale. This was the YouTube and Meta platforms model, and the reason tech is now a substantial part of digital media.


2015 revenue mix transition.



Comparable platforms at $100M scale


Facebook 2007: $153M revenue. 2017: $40.65B. 10-year CAGR: 74.8%. Source: Facebook 10-K filings, Macrotrends.

YouTube 2008: approximately $100M ad revenue. 2018: $11.16B. 10-year CAGR: 60%. Moffett Nathanson 2025 standalone valuation: $475–550B. Sources: Alphabet disclosures, Statista, MoffettNathanson.


Facebook and YouTube actual trajectories from $100M base, with Mode base case projection (35% CAGR).



Forward projections 2015–2026


Three scenarios from $100M 2015 base, all below comparable platform rates:


EBITDA margin at scale: 35–42%. Mode’s operating cost structure was materially lighter than YouTube’s (no video streaming infrastructure at YouTube scale) and Meta’s (no user-generated content moderation at Meta scale, no Reality Labs/AI capex), with revenue at high CPM premium vs. YouTube’s blend. Same infrastructure load, more revenue per dollar of cost.


Conservative 30% CAGR: $1.79B revenue, $682M EBITDA at 38%, $14.3B EV at 8x revenue.

Base case 35% CAGR: $2.71B revenue, $1.03B EBITDA at 38%, $21.7B EV at 8x revenue.

Upside 40% CAGR: $4.05B revenue, $1.54B EBITDA at 38%, $32.4B EV at 8x revenue.



Forward trajectory 2015–2026, three scenarios.


The $10 million decision. A Board takeover. An External Plan.


February 2016. $10 million bridge financing committed by Partners for Growth. Not approved by the Burda-controlled board.

Saved: $10 million.


Future enterprise value unrealized by 2026: Billions?


The Future. A Fundamental Technological Shift. Now.


Can we build something that will meaningfully impact the world at this scale or more again, with all the lessons learnt?

Three years ago, one of my top engineers came to me and said, “The idea Sage AI had on building intelligence based on proprietary data was good, this (then) small company called OpenAI has implemented a new generative intelligence. We should start a new company now.”


But, that’s the title of another post. If you want to get involved, reach out.



 
 

SamirArora.com 

bottom of page