This story begins with a string board, because that’s how it was pieced together and it's the best way to illustrate it for you.
I’d been covering promising agency tech startup Hudson MX for several years, when in February, I stumbled upon a release announcing a deal giving it “substantial new funding” from a mysterious new investor. The deal also provided long-time investor Ascential plc a path to control it.
The announcement said Ascential CEO Duncan Painter and Chairman Scott Forbes would serve as Hudson MX’s board, working with Stuart Johnston, a long-time Ascential executive who was named CEO.
Johnston succeeded JT Batson, who left the startup late last year to become CEO and Secretary General of the U.S. Soccer Federation.
advertisement
advertisement
I covered the news and didn’t think anything more of it.
End of story.
A couple of weeks later I received a document revealing there was much more to it.
Codenamed “Project Hawk,” the document was the notice sent to shareholders explaining the company had been restructured and was now owned by two preferred equity shareholders -- Ascential and new investor MT II Holdings -- controlling a combined 87.5% of its equity.
The remaining 12.5% was now owned by a small group of common shareholders represented by Alexander Kassan, the son of MediaLink founder and CEO Michael Kassan, who was also among a new class of common shareholders.
All other outstanding shares of common stock and options -- most of which were held by long-time Hudson employees and early-seed investors -- “were canceled in the merger for no consideration,” the document said.
That’s when I began compiling the string board with post-it notes, photos and documents to piece together what happened, and how the people involved were connected.
“Why are you so interested in this,” one of the principals asked me, adding: “It happens all the time. It’s a nothing burger.”
On the surface, he was right.
The restructuring of startups that fail after burning through investment capital happens all the time. And more often than not, common shareholders -- even the ones who earned sweat equity and exercised cash options -- often get wiped out, while the preferred-equity investors preserve their stakes.
That’s the nature of startup culture and typically part of the terms of investor agreements.
What makes Hudson MX’s restructuring an important story for the advertising industry is the people involved -- some of the highest-profile and most influential people in the business -- and the nature of the product involved.
The principals include powerful management consultant Michael Kassan, legendary agency media chief Irwin Gotlieb, and Ascential, which owns and operates the Cannes Lions International Festival of Creativity, as well as a variety of businesses important to the ad industry.
In terms of the product, Hudson MX was the latest in a long line of startups that have promised to transform the technology advertisers and agencies use to process media buys, pay their vendors and make sure they deliver the best returns on their investments.
It’s an industry that for half a century has been dominated in some form by old, legacy mainframe computing technology first created by entrepreneur Michael Donovan -- the founder of Donovan Data Systems -- back in the 1970s, some of which continues to be the core of Mediaocean, the company that was created when Donovan merged with another promising agency tech startup, MediaBank, more than a decade ago.
It’s also a story that reveals some dysfunction within the advertising industry itself, which doesn’t always seem to understand what it wants and needs when it comes to developing technology and enterprise software to transform how it does business -- even as it advises the world's biggest marketers on the “digital transformation” of their businesses.
This article is the first in a series that will drill into some of those ad industry issues, but the goal of this one is to explain explicitly what happened with “Project Hawk” -- what led up to it, who was involved, what their roles were, who came out on top, and who did not.
Hudson MX was founded in 2016 by Batson and Neil Hitzig, an entrepreneur who had spent years developing new media-buying software, including a prototype of what became Hudson MX’s first viable product, BuyerAssist.
The product innovated a type of media buys -- spot broadcast -- that had essentially been ignored by other agency technology providers. It streamlined the process and utilized cloud-based software that was nimble and more efficient than legacy systems.
It got Hudson MX on base with most -- if not all -- of the major agencies, which began using it to remove some of the friction associated with their spot broadcast buys.
It also got the attention of Alexander Kassan, whom Batson pitched to become an early investor.
While not an active player within the ad industry, Kassan was a savvy investor and brought the opportunity to his father Michael Kassan, and the two of them became seven-figure seed investors along with other well-known people in the media business, including Irwin Gotlieb -- who was then GroupM global chairman -- and others.
Kassan -- who sold MediaLink to Ascential in late 2016, then introduced Hudson MX as an investment opportunity for Ascential and the company -- soon became its largest source of startup capital, investing as much as $100 million or more in a series of equity deals and loans going back to at least 2018, according to sources inside Hudson MX, as well as other investors.
It was during this period that Hudson MX’s ambitions grew from stand-alone solutions like BuyerAssist to developing a plan to build a complete, end-to-end holding company-level enterprise system that could replace Mediaocean.
That’s also when Hudson MX began burning lots of capital, hiring teams of developers, analysts, sales and marketing people to begin working with the major holding companies to build “the agency operating system of the future.”
For some on Madison Avenue, it was a repeat play of what happened a decade earlier, when then venture-backed startup MediaBank began working with Publicis to build an alternative to Donovan’s legacy systems.
Following years of contentious arbitration, threats of litigation and lots of trade press headlines, MediaBank and Donovan ultimately merged into a new company -- Mediaocean -- with the promise of combining the best of both systems and finally enabling Madison Avenue to achieve the kind of “digital transformation” it was pitching its own clients to undergo.
In retrospect, it was that all-in, Mediaocean replacement strategy that many believe was Hudson MX’s greatest folly.
It also exposed the fact that parts of big agency holding companies may not actually want to transition completely off legacy mainframe systems.
More than a decade after Mediaocean was founded, few if any of the holding companies have fully transitioned from Donovan’s legacy mainframe processing systems. Among other things, holding company financial departments have been loath to transition for fear of losing access to invoices and data that might be needed for financial compliance reasons years later, if a client requires an audit.
Fear of losing access to years of data -- or being cut off completely -- is one of the chief reasons that big agencies never moved from Donovan and/or Mediaocean in the past.
While BuyerAssist gave Hudson MX a relationship with the holding companies, getting them to transition whole hog was another matter.
Among other things, Mediaocean had long-term, staggered contracts with the holding companies -- making it difficult, if not impossible, to switch multiple companies to Hudson MX en masse.
To date, Hudson MX has only two clients using its system for all end-to-end media-buying processing: Verizon Media’s in-house media-buying agency; and independent barter media agency Evergreen Trading.
Hudson MX’s holding-company strategy focused on the ones whose Mediaocean contracts were coming up for renewal, and that put Dentsu first in line.
Dentsu declined to comment on its relationship with Hudson MX, but according to executives familiar with its deal, Dentsu effectively became Hudson MX’s holding company beta -- and as a result, began shaping Hudson MX’s development based on its unique needs.
That in turn created more complexity, demanding greater resources and burning more capital than Hudson MX originally planned for.
While the startup continued to court other holding companies, the team’s focus was on getting its system and software up to speed to meet a January 1, 2021 deadline to transition Dentsu from Mediaocean.
But nothing happened. Dentsu stayed with Mediaocean. And no one explained why.
Two months later, Michael Kassan’s MediaLink announced it had created a new consulting practice, Hudson Bridge, to advise advertisers and agencies on transitioning their technology to Hudson MX, and hired Dentsu Media’s U.S. COO Lucas Cridland to run it.
Nine months later Ascential sold MediaLink to Hollywood talent agency UTA for $125 million.
It was far from the end of their relationship.
Michael Kassan retains close personal ties to Ascential, including an influential presence at its annual Cannes Lions festival.
He also continues to count Ascential among MediaLink’s consulting clients.
And he remains an investment partner with Ascential in Hudson MX.
As for Hudson Bridge, MediaLink now says it never really developed as a practice and that Cridland only worked for the company a short while.
Cridland did not return a call to MediaPost, but he joined Shelton, Connecticut-based performance marketing agency BMG360 as CEO in July.
In retrospect, executives on all sides of this story say Dentsu may not have been well suited to become Hudson MX’s beta holding company client, but it was the first to have its Mediaocean contract come up for renewal.
Dentsu has been going through a series of high-profile executive departures and various restructurings of its media services organization.
While it's unclear exactly why Dentsu never transitioned to Hudson MX, there are a number of theories, including that an 11th hour assessment of Hudson MX’s technical capabilities gave it cold feet, and the holding company exercised a clause in its contract letting it out.
One knowledgeable executive says Mediaocean chief Bill Wise leaned on relationships with big advertisers, including at least one of Dentsu’s biggest, to put the kibosh on its transition to Hudson MX.
In May, long-time General Motors CMO Deborah Wahl joined Mediaocean’s board of directors, after officially retiring from GM.
Others say Wise had an almost opposite strategy of letting Dentsu transition fully to Hudson MX in order to demonstrate it wasn’t ready for prime-time and that the problems ensuing from that would be a powerful lesson for other holding companies.
Despite -- or perhaps because of -- the demise of the Dentsu deal, Mediaocean quietly made an offer to acquire Hudson MX last year.
During a series of discussions with Hudson MX management, as well as Ascential management, Mediaocean made a cash offer that “would have made all the investors whole.”
While Hudson MX CEO Batson and the Kassans were said to be in favor of accepting the Mediaocean offer, Ascential’s Painter passed.
Mediaocean executives say they were puzzled, because they knew how much Hudson MX had burned through, what its investors' losses were adding up to and how low the prospects were that they'd be able to turn it around.
“It felt like they were kind of delusional,” an executive involved in the discussions said of Hudson MX's rejection, adding: “It seemed like they felt like they were really in the running for some of these things.”
In September 2022, not long after Hudson MX passed on Mediaocean’s offer, Batson -- a lifelong soccer fan -- announced he was leaving to take his “dream job” as CEO and Secretary General of the U.S. Soccer Federation.
“Michael Kassan and Alex Kassan, two long time investors in HMX and advisors to me and the company, will be stepping in to take an active role during the transition,” Batson wrote in an email sent to the Hudson MX team, as well as its investors, adding: “Otherwise, the leadership team and the strategy of HMX remains in place and on track. My personal role will transition to being an active board member focused on customers and product strategy -- so you'll still be hearing plenty from me.”
Less than five months later, Ascential announced Hudson MX’s restructuring, and the Project Hawk document was sent to its shareholders informing most of them that their shares had been canceled.
There still are many unanswered questions associated with the document, including why long-term primary investor Ascential essentially sold a controlling stake in Hudson MX to new investor -- MT II Holdings -- for $30 million.
And why in all of its filings as a publicly traded company in the U.K., Ascential has maintained it doesn’t control Hudson MX, even though its two top executives are now on the startup’s board, and many current and/or former Ascential executives have been playing operational roles, including new CEO Stuart Johnston.
Others familiar with the deal say MT II Holdings, which is represented by John Garcia, executive chairman and a partner in private equity fund manager AEA Investors, is akin to a "bridge loan" enabling Ascential to buy time in order to execute a bigger strategic plan to spin off certain “ecommerce business” assets – including Hudson MX – as part of a new U.S. public offering on the New York Stock Exchange.
According to a Feb. 7, 2022 disclosure by Ascential, the restructuring deal gives MT II Holdings an option to sell 79% of Hudson MX to Ascential for $52 million that can be exercised beginning April 1, 2024.
The core asset in Ascential's spin-off plan, ecommerce company Flywheel, was acquired by Ascential in 2018 in a deal valued at upwards of $400 million, and by all accounts it has been one of Ascential’s most successful investments to date.
MediaPost was unable to find a preliminary prospectus of the IPO plan, and the only Ascential filing with the U.S. Securities and Exchange Commission to date, was a January 2020 application to issue 50 million American Depository Shares in the U.S. valued at $5 each.
According to people familiar with the plan, Ascential’s Painter would become CEO of the new U.S. public company, and the U.K.-based Ascential would continue to operate its other operations, including the Cannes Lions, though there is speculation the festival might also be on the block.
Ascential’s Painter did not take calls from MediaPost, and an Ascential spokeswoman said “the company does not comment on speculation,” declining to comment on any specific details, including some of the facts laid out in the restructuring document.
Meanwhile, dissatisfaction with Mediaocean’s legacy systems remains acute on Madison Avenue, although the company has been investing heavily in upgrading its systems and software and diversifying its portfolio since restructuring its own private equity owners from Vista Equity Partners to CVC Capital Partners in August 2021 with a focus on growth.
Since then it has acquired ad server and digital ad-optimization technology provider Flashtalking, among others.
Earlier this year, Publicis announced a seven-year renewal of its contract with Mediaocean and Omnicom and Interpublic are said to have renewed, as well, although those renewals are believed to have far more flexibility for the holding companies than previous contracts have had in the past.
And while WPP is believed to be coming due soon, it’s unclear whether it would seriously consider a switch to Hudson MX, although its top media executive at the time -- Irwin Gotlieb -- was one of its biggest supporters, as well as a personal seed investor when Hudson MX launched.
Gotlieb officially retired as chairman of WPP’s GroupM operations in April 2018, but not before embracing Hudson MX as an opportunity to finally transform agency media-buying technology in a way he had always dreamed of.
Gotlieb, long a proponent of leveraging technology to improve media-buying, built one of the earliest homegrown agency systems when he was at Benton & Bowles, and he was an early adopter of Hudson MX’s BuyerAssist technology.
As an early user, he actually wrote some code to refine it, which later became incorporated into its ongoing technology.
Gotlieb disagreed with Hudson MX’s strategy of developing customized versions of systems for each of the holding companies for a variety of reasons, including the fact that while agencies are masters of advertising media-buying, they are not necessarily the best designers of enterprise software systems.
Gotlieb also believed that without standardized operating systems, the ad industry would risk tremendous disruption due to the lack of interoperability between agencies, especially when talent moves from one shop to another and has no framework for the systems they use to plan, buy and post the ROI of media.
But Gotlieb, who became an early-round investor in Hudson MX at the same time the Kassans did, inexplicably was not treated the same when the Project Hawk restructuring took place.
While the Kassans, Batson, and Hudson MX co-founder Hitzig are said to be among the shareholders in the new common equity pool, Gotlieb and other seed investors still don’t know how the restructuring affected their equity, whether they own continuing shares, or what kind of settlement they got for the cash they originally invested.
Following the restructuring, many former and current employees and investors say it is difficult to see a path to success for Hudson MX due to capital it has already burned through.
According to Ascential’s Feb. 7 restructuring statement, the net result of Ascential's and MT II Holding’s investments “provide fresh funding of up to, in aggregate, $51.5 million” to help fund Hudson MX’s next phase of development.
Since launching in 2016, current and former executives and investors estimate Hudson MX has burned through more than $200 million to get where it is today. Others say the capital expenditures weren't nearly that high, and that some of the money was spent on acquiring shares back from early-round investors, not on Hudson MX's product development or operations.
According to the Project Hawk document, Hudson MX generated about $2.2 million in revenue -- primarily from its BuyerAssist software -- during the 11-month period ended November 2022.
Over the same period, Hudson MX reported $29 million in net losses.
“I think it’s accurate to say it’s harder and more complicated than everyone thought. And everyone thought it was going to be complicated,” Batson told MediaPost in retrospect, adding that Hudson MX underestimated how long the timelines were for an agency to transition from one system provider to another.
“They’re really long because they’ve already got buys in-flight, so you need to stagger the transition over a year or two vs. doing a hard cutover,” he said.
Meanwhile, at least one of the major agency holding companies -- Interpublic’s IPG Mediabrands unit -- and the biggest independent media-services company, Horizon Media, are believed to still be actively considering a switch from Mediaocean to Hudson MX, although they remain concerned about its long-term financial viability.
“I have confidence that the existing investors are committed to the continued financing of the business,” Michael Kassan told MediaPost in response to a query about Hudson MX’s financial commitment.
Importantly, Kassan is not only a common equity investor, but a consultant to Ascential, as well as an influential ad industry deal-maker, especially where the holding companies are concerned.
As for the hundreds of current and former Hudson MX staffers and contractors who exercised options only to see their common shares canceled in the Project Hawk restructuring, they’re still scratching their heads and weighing their options on what to do next.
Excellent research & reporting, Joe. Not sure why this wasn't the top story on my MP email today, it sure deserves it.
I know you've been around long enough that the moment somebody says “Why are you so interested in this? It’s a nothing burger,” the alarms go off. Chum in the water. It's like "These are not the droids you seek," LoL.
Looking forward to read what comes next.
Great story. I had some dealings with HMX during a test phase in 2020. We tried, but couldn't make all the moving parts work.
Very good article. Irwin was right, but technical complexity, agencies' limited IT expertise and the need for look-back records access aren't the only impediments to modern operating systems. Maybe the biggest one - and the most profound change since the Donovan age - is agency compensation: when you get paid by the head, and most of the heads are doing the grunt work, no one is in a hurry to automate.
Thank you for the first half of the story, Joe! Good to see some transparency here.
This is a great piece of investigative journalism! The industry could use more like this in the trades. Thanks a lot!