Why The Trade Desk should be building a TV Operating System (OS)

Let’s be clear. Building a TV OS and adding advertising to it is tough. The world needs another TV OS gatekeeping your access to the wonderful apps with the incredible content, as much as your pair of Crocs needs another hole.

From working experience in adtech and in the ads business of TV OS companies, I am confident that The Trade Desk is likely to make a successful pivot. I will share the one opportunity enabling them to make an aggressive gamble from B2B into B2C tech. I must explain why it is actually a defensive move. I will explore their unique position for the gamble. Finally, in this article, I identify a path to significant market share within five years.

Reasoning - For an aggressive defensive move?

The Trade Desk has been very successful since its $18 billion IPO eight years ago. Now valued in the $100 a share range, the company holds a $50 billion value. For many ad buyers, it is the primary Demand Side Platform (DSP) to execute Connected TV video ad spots. From Tubi to Netflix , Disney+ Hotstar to Vevo . The TV revolution is being programmativised!

Yet, their Open Internet goal has sizeable pressures elsewhere in the ad-funded internet. This includes, but is not limited to, the loss of identity for display ads in Safari and Chrome. As the most successful independent DSP business, Internet giants – such as Google and Amazon – are now its competitors. Their peer set are companies that have a direct relationship with the consumer.

The Trade Desk (TTD) cannot afford to be left behind. With every passing day, competitors owning a TV OS expand device distribution reach. This enables them to increase the volume of insights from the TV, the biggest screen at the heart of the home. To a lesser degree, current partners of TTD such as Samsung Electronics , LG Electronics and Roku also compete as they have their own managed service DSPs via their ads business Samsung Ads , LG Ad Solutions , Roku Advertising , Google Ads and Amazon Ads . The fear is being responsive to Google's whims once more. All of adtech knows this from monetizing users of the Chrome browser.

The TTD needs to use its market value to break out of the adtech niche to be able to survive and thrive. Ironically, it is a Google tool that enables the company to pivot to TV OS without building a system from scratch. The ability to fork the Android OS solves the make-or-break problem in developing a new TV OS. You avoid having to integrate hundreds of broadcaster apps. Each apps has limited teams who frequently unify in seeing the new market entrant as an untested, unattractive, unmonetizable, non-scaled platform. This includes both technical and commercial effort. It is a tried and tested approach. Amazon's current Fire OS is an Android fork, although internet article and rumours suggest that now they have built scale over 10 years, Amazon will move to a Linux-based OS for TV and Echo devices. It is called ‘Vega’ OS.

In my opinion, use of an Android fork makes a pivot to become a TV OS possible for TTD. I would never encourage coding a new TV OS launch from scratch. If you do not have all the local, regional, and global apps, your devices will remain on the shop floor. Without a decent volume active and in consumers' homes, streamers will not build an app for your OS. For that reason, advertisers will not be able to target your platform. Your OS becomes 'Other' in the chart that follows below. It is a vicious circle.

An Android fork prevents the need to pitch, contract and integrate the key apps that consumers demand, but does need ongoing support. This is something the TTD will need to resource internally to manage; their share price will depend on it.

Background

For television devices, hardware companies are also known as Original Equipment Manufacturers (OEMs). OEMs design and produce components to sell models to retailers. Or, they respond to a retailer's contract opportunity. An example contract opportunity is:

An electronic retailer in Egypt requests 20,000 devices of various sizes. Needs Amazon Prime , Netflix and local apps. Aiming to retail the range at $300 or less.

Most household devices, such as kitchen appliances, ship as hardware only. Yet, smart televisions need software and hardware. A retail expert whom I trust, the manager of my local electrical superstore, once told me that consumers buy a device based upon six factors. For consumers, importance weighs towards the screen size and value for money. TV OS needs to deliver the consumer their favourite apps (i.e., DAZN , Apple TV + , Rakuten etc.). To a lesser extent, trust in the OEM brand, the TV OS itself, and device technical specifications also play a role.

For OEMs and retailers, margins are low. Smart TVs start at under $100. Following Samsung Electronics 's lead, even factories and retailers know that ad monetization is the way forward. The addition of ads to a TV can create an annual margin of up to 10 times the value of the hardware sale itself. So, a market where OEM factories/RFI holding retailers paid a few dollars per device for an OS is quickly shifting. Now, the OS pays the OEM for getting a device installed in homes. Rev shares of ad spend and flat fees per device have been cited by sources close to such deals.

External factors

Compared with mobile OS (iOS vs Android), TV OS fragmentation is real and very complex. There are approximately thirty (30) companies with an OS product jostling for position. The best available records contradict each other. This shows the market's complexity that TTD would enter. These are a crowdsourced Wikipedia page and a chart created by the Connected TV Marketing Association (CTVMA).

It is correct that Samsung, LG, VIZIO and Roku use their own OS on the devices printed with their OEM label. Yet the CTVMA calculation assumed that Hisense Group (In-house OS: VIDAA ) and Philips TV & Sound (In-house OS: Titan ) have always used their in-house OS. The Wikipedia highlights existing gaps from past model ranges, where the OEM installed another OS. Thus, the CTVMA chart inflates their respective global install bases and potentially undervalues “Others”. Another trend is that many legacy TV OEM brands do not manufacture TV devices now. Instead, they choose to license their brand name to another factory. Examples include JVC , Panasonic and Hitachi High-Tech Corporation . Click on each company to see an example of the brand licensing: JVC, Panasonic and Hitachi.

Older consumers seek out legacy TV brands. Many value the biggest screen available, badged in the retailer's reliable white-label brand. There is a very large addressable market of devices for TV OS operators to pair with, if you can monetize it. The Trade Desk can certainly do that. They would sell ads direct to their agency and buy-side customers from their console. They would directly integrate into publisher placements in the OS. Less deductions from the post will result in a larger dollar value, for the OEM/retailer.

The TTD TV OS opportunity is global. As global as their customer base of advertisers, agencies and retail media partners. In fact, if they wanted to open a new geo, or retail media partnership, they could offer upfront fee per active device. The US market is saturated – accounting for 90% of present CTV ad revenues. Large market share in the Global chart belongs to companies which only operate in the US. These include Roku, a market-leading public company with 64% of TV OS revenue in the US in 2022. Since then, Roku has rolled back attempts to expand to UK and Germany.

Vizio, an OEM created with the ad-enabled value exchange explained to consumers in return for a lower purchase price. They have approx. 10% US market share. In February 2024, Vizio was bought by grocery giant Walmart for $2.3billion. The most recent entrant, Telly TV , has a very compelling offer for the consumer. Telly provides select US households with a high-spec $1,000 device for free. The value exchange is that ads show on a tablet below the screen as you view programming.

Decisions made today can take up to two years to result in a scaled and targetable number of devices in homes. There a few options available to TTD now. They are:

1.      Entering the TV OS space, using an Android fork. Gain market share in unfashionable, hard to monetise markets in EU and LATAM, where TTD can use its leading DSP position to ad monetize

a.      Focus on the white label offering. This will provide the smallest scale, operating in a bootstrapped fashion to grow and learn. Distribute less than 1 million devices in 18 months.

b.      White label + retail media partnerships with electronic and grocery store. Partners can provide increased OS distribution, have a local partner delivering a co-ordinated sales approach, which adds new customers to the DSP + who could drive local brand TV advertiser demand with unique datasets. Distribute 1-3 millions devices within 18 months

c.      1B+ Offer a substantial minimum guarantee to key factories (i.e. Vestel ). This will enable the TV OS to ship in multiple brands, in multiple geos, develop expertise with a mutually invested partner. Distribute 2-3 million devices. Key downside is competing with establihsed competitiors including Google, Roku, VIDAA, et al.

d.      Buy one or more of the ‘Others’.

Whale TV have published a number of 41 million active monthly users. Would obtain a team who know about TV OS. Ad formats may not be backward compatible and older devices might not be able to support a new OS, so TTD end up running two OS platforms.

2.      Go Global. Pitch biggest or fastest growing manufacturers.

Offer an unbeatable minimum guarantee per active distributed device. Could add 6-10million devices within 18 months distributed everywhere. The fastest growing include Hisense (no2. by global shipments since 2023) and TCL (no proprietary OS)

3.      Focus on the US only

a.      Buy Roku. Down 37% on their 52 week high stock price. Also down 91% a Covid pandemic induced all-time high. With 120 million distributed devices, is it worth a $9-11 billion bid?

b.      Compete with Telly. Offer a free, with commitments, device to a number limited by cost or imagination.

But given their background and internal culture, what could TTD do with the OS?

What innovation could they bring to it and what could it bring to their business?

Internal factors

• Software as a Service (SaaS) excellence

Within The Trade Desk Edge, The Trade Desk’s certification and education program there is a name check for Iponweb. Iponweb is the company that built most of the adtech ecosystem of SSP, DSP, DMPs, graph, and context vendors. It states that they built their DSP differently from the rest of the market. It explains how their logic works and why it is superior. We also know that they marketed their DSP in an alternative method to MediaMath and Turn. The result was that TTD grew from an outsider to a leader in the space. Congrats to Jeff Green and team for the win, and the explanation in the Edge program.

• In-House CTV Leadership expertise

Since 2021, the VP Product is Bill Simmons. Bill is a superbly commercial-minded technology leader. The founder of the OpenRTB Consortium, Bill held major roles at IAB Tech Lab and Prebid. He was also the CTO and co-founder at DataXu, a DSP that sold to Roku in 2019.

• Access to a new format: CTV Native/CTV Display

At present, TTD does not offer display ads on TVs to its customers. TV OS sells these directly to agencies or advertisers as a managed service via a few specialist ad sales houses or creative firms. Globally, there are more display ads than FAST video ads available each month. CTV Display placements are usually on the homescreen of the TVOS. They are also found in the electronic program guide, or as a pause ad when a consumer watches a channel. Some TV OS are able to offer targeting based upon data, but these units are not available for programmatic OpenRTB. Like CTV video compared to online video. The CPM is at a 5-15x premium vs their desktop and mobile counterparts. I would expect TTD to create OpenRTB standards that open CTV Display scale from other TVOS. In return for programmatic access, Samsung, LG, Roku and Vizio will be asked to support UID2.

• Shorten demand path (OpenPath) means more money for OEMs

Using OpenPath, CTV publishers will integrating directly with TTD. TTD will monetize their supply with priority. This will give TTD's TV OS will have an advantage over other TV OS where they have ad sales agents. Monies from advertisers will only see TTD and the agency take a cut, compared to the agency, DSP, SSP, Ad sales agent on other OS. This all takes place before the TV OS splits it with the OEM! The TTD can operate on a par with Google and Amazon.

• Opportunity to Speak Directly to the customer

The TV OS will likely offer additional value and features in return for a login. TTD TV OS can satisfy content recommendations, consent obligations, and ad personalization by establishing a first-party relationship with the consumer. The Trade Desk will be able to ask consumers and develop extended profiles for these TV viewers.

• Extend the UID2

First party relationships would allow TTD to match other devices in the home to the TV and the profile/s used to login in on the TV. To currently deliver these matches in the US, the IP address or probabilistic algorithms and graphs are used. TTD will invite advertiser and agency customers to maintain, improve and increase their TV targeting with the TV OS strategic move.

• Build a panel or brand-specific panels for key advertiser customers

Ask the right questions from the outset and you have a sizeable and cross-nation distributed panel. Even at a small scale the panel is larger than linear TV panels.

For the TTD to compete in retail media, owning their own panel is an unmatchable development. The US market will soon see the value of targeting brand awareness (video and display CTV campaigns) to customer loyalty data. This will be a direct result of the Walmart/Vizio tie up. A US-owned and operated OS would let The Trade Desk introduce a product. It would quietly drift into the slipstream of Vizio's offering. It would split the spend once Vizio/Walmart lead the market education.

• Create the largest Automatic Content Recognition (ACR) solution

A popular misconception of ad monetisation folks is that an OS gives you ACR. This is not true. TV OS owners still need to build it from licensing many partnerships. It can get expensive quickly and in some markets vendors do not exist so ACR cannot be created. Since April 2024, TTD has invested in VIDAA and other platforms' data from Nexxen and Roku. Samsung relies upon their 'Walled Garden' insight and will be unlikely to license it to TTD.

TTS would want to build the OS with the perspective to enabling data segmentation and targeting. This would make in-house ACR the most flexible for activation. They will want all users to be contributing to the ACR data from launch. I estimate that through partnership TTD has ACR from ~30% of US market and ~40% of other key markets. It is likely that over time there will be less need to pay competitors for their data sets.

• Disrupt to solve for measurement and attribution

This is certainly one for the distant timeline. Yet, the goal to make performance campaigns work at scale will be a race between Amazon, Google, and TTD. Linear and digital measurement companies will need to watch this race with great attention

Conclusion:

TTD can drive a comparative Average Revenue Per User (ARPU) to existing reported TV OS Vendors. ARPU is the benchmark used to measure TV OS financial performance. Roku made $40.68 ARPU in Q2 2024 from 83.6 million users. In Q1 2024, Vizio made $34.20 ARPU from 18.8 million users. There are 200 million Amazon Fire TV devices and over 150 million Android TV devices active. Google and Amazon do not share ARPU figures for their TVOS.

A pivot of this degree is bold. A TV OS pivot is a once-in-a-lifetime opportunity for adtech's most successful independent. There are a lot of good reasons—reasons unique to TTD—to proceed.

I forecast that Jeff Green and team take the opinion I outlined as 1C to combine retail media and drive CTV adoption in Europe or LATAM. This is a excellent proof of concept and will unsettle the developed US market. Within three years I believe that demands will surface from Roku shareholders to sell their US business to TTD. Any deal value is likely to be lower than $9.3 billion valuation based on $65 a share today. At that time any Roku deal would add a scaled, market leading volume of US consumers to the global offering of a TV OS and DSP by the end of 2030.

END - Thanks for reading to the end. What do you think? Please add a comment, share with someone who will enjoy the read or give a like to increase awareness.

During September 2024, I intend to be at ATS London, IBC in Amsterdam, Dmexco, and the Videoweek event in Cologne. To discuss this topic, FAST monetisation, CTV go-to-market, increasing viewership numbers, or DSPs and SSPs, please get in touch to arrange a meet.