The Trade Desk Stock Breakdown: Dying Business or Generational Opportunity?

Get smarter on investing, business, and personal finance in 5 minutes.

Stock Breakdown

This newsletter is adapted from a recent video Drew Cohen did on The Trade Desk.

The Trade Desk was one of the best growth stories in public markets.

From IPO in 2016 to peak the stock was up almost 50x.

For more than a decade, the company compounded revenue at extraordinary rates, and generated ~40% adjusted EBITDA margins, and became the dominant independent platform for buying advertising across the open internet. It was also remarkably consistent: for 32 consecutive quarters outside of Covid, The Trade Desk met or exceeded its own guidance.

Then everything started to unravel.

The company missed guidance for the first time in Q4 2024. Revenue growth subsequently slowed from 26% in 2024 to 18% in 2025, before collapsing to just 3% in Q2 2026.

Now Management is now guiding for revenue to potentially decline 12% year-over-year in Q3.

Along the way, The Trade Desk went through the largest reorganization in its history, struggled through a prolonged migration to its new Kokai platform, cycled through multiple CFOs, and became embroiled in a very public dispute with one of the largest advertising agencies in the world.

The stock has responded accordingly, falling roughly 90% from its all time highs.

Management’s explanation for their poor business performance is a mix of macro factors and poor execution.

However, when a business is struggling, even the managers themselves can misidentify the culprit.

The question for investors is simple: are the company's recent problems temporary, or is something structural changing in the advertising market?

The Trade Desk is the largest independent programmatic advertising “DSP”, which means they help advertisers buy ads on various platforms.

Their independence means they can be objective in deciding where to purchase an ad slot on behalf of advertisers, and this independence was their greatest advantage.

However, that independence also comes with a meaningful disadvantage: The Trade Desk does not own the same kind of proprietary consumer data as companies like Amazon and Google have.

While Google has been a long-term competitor, Amazon’s has been increasingly competitive in the space, leveraging their retail platform to provide advertisers a “closed loop”, which can help them actually know which ads drive purchases with high confidence.

At the same time as Amazon has made this ad buying market a focus, The Trade Desk has stumbled.

How much of this can be improved with execution versus a new reality of having to compete against Amazon?

And what exactly is a DSP? And what does The Trade Desk really do? We cover all of this and more in this week’s Five Minute Money Newsletter!

Business.

In short, The Trade Desk is a Demand-Side Platform, or DSP, that helps advertisers buy digital ad slots.

These digital ad slots can come from the open web (websites online), apps, audio (like podcast ads), or connected TV, which is internet connect televisions (think streaming Netflix).

An advertiser gives The Trade Desk a budget and parameters around things like who they want to reach, what type of media they want to advertise on, and what outcome they are targeting.

The Trade Desk then uses its software to decide which individual ad impressions to bid on and how much each one is worth.

So if Netflix has an ad slot available, The Trade Desk might decide that particular viewer is highly valuable to one of its advertisers, bid on the impression, and show the ad if they win.

In 2025 they generated $2.9 billion in revenue on roughly $13.4 billion total advertiser spend.

They primarily make money by taking a percentage of that spend, historically around 20%, as a platform fee along with fees for certain data and other services.

Unlike a SaaS company, though, advertisers aren't signing multi-year contracts committing to spend a set amount. They can move advertising dollars between DSPs relatively easily. And they are always looking for the best and highest use of their ad dollars.

The Trade Desk doesn't report product-level revenue, but there are six major parts of the platform worth understanding:

1. Kokai / Zuma
2. Data & Audiences
3. Identity
4. OpenPath
5. Deal Desk
6. Measurement

Let’s take them one at a time.

1) Kokai / Zuma

Kokai is The Trade Desk’s core software platform. It is where agencies and advertisers build, manage, optimize, and measure their campaigns across different types of digital media.

Kokai replaced TTD’s prior platform, Solimar, and incorporates the company’s AI system, Koa, to automate more of the campaign-management and optimization work.

Zuma is the newest version of Kokai, focused on making the platform simpler to use while adding more agentic AI capabilities.

It improves the UI of the platform (common complaint) and ads new AI Agentic capabilities, as well as enhanced measurement.

2) Data & Audiences

TTD also operates a large data marketplace that lets advertisers bring additional information into their campaigns.

This includes an advertiser’s own customer data through Galileo, as well as outside datasets from retailers and other third-party providers.

More recently, TTD launched Audience Unlimited, which bundles access to a broad set of third-party audience data rather than charging separately for each dataset.

This is particularly important because TTD does not own the same proprietary consumer dataset as Amazon or Google.

Each data set purchased though has a separate fee charged on top of it and with more AI buying through Kokai/ Zuma, the cost of Trade Desks services has been growing. Some advertisers have pushed back on this though feeling that the system is more optimized to Trade Desk revenues then to purchasing the best ad. (This is probably not true, but an advertiser seeing their bill grow with the new AI tools does create this impression.)

3) Identity

UID2 is TTD’s identity system for the open internet.

It is designed to help advertisers recognize the same consumer across different publishers and devices without relying on traditional third-party cookies.

TTD also operates EUID, a European version of the system, and OpenPass, which helps publishers establish logged-in relationships with their users.

The broader ambition is to create shared identity infrastructure for an internet that otherwise lacks the logged-in consumer graphs of Amazon, Google, or Meta.

4) OpenPath

OpenPath connects publishers more directly to The Trade Desk.

The Trade Desk is a DSP, which is a demand side platform that servers the advertisers.

On the other side are SSPs, sell side platforms which serves publishers (anyone who has ad slots to sell).

Normally, publishers use various SSPs to make their ad inventory available to buyers. OpenPath removes some of those intermediaries and gives TTD a more direct connection to publisher inventory.

This is disruptive to both the SSPs and was a shot at Google because they simultaneously announced they would stop supporting Google’s Open Bidding solution.

The idea behind it is to reduce middleman costs by going direct to SSPs, it also represents existential risk to the SSPs since they are cut out of the transaction.

While CEO Jeff Green said that they are not trying to become an SSP since they are still focused on maximizing advertisers spend (driving prices down), whereas SSPs want to maximize publisher revenue (drive prices up) and publisher tools, it still is very threatening to the SSPs.

OpenSource also does not aggregate demand from other DSPs like Google’s DV360, Amazon, or Yahoo so it really isn’t a replacement for an SSP. The other SSPs are still in the same auction it is just that if Trade Desk wins and it is through opensource, no SSP fees are paid so the publisher’s take is higher.

Nevertheless, though many SSPs like PubMatic (with Activate) and Magnite (with ClearLine) have tried to go direct to buyers to cut out the buyers fees.

The problem is they lack complicated ad buying tools and then still require an ad buyer to manage more accounts, which they don’t want to do.

Still though they got some big names including Disney, Warner Bros., Discovery, and DirectTV Advertising that participate. (This just means that their inventory is buyable through these means, not that it is exclusive).

GroupM Premium Marketplace, launched by large advertising firm WPP, aggregates several SSPs including PubMatic and Magnite for direct buying.

This bypasses traditional DSPs saving their clients money.

The reason why it isn’t more disruptive though is because it lacks themachine learning decision engine that Trade Desk built and is primarily used for bulk ad space buying where performance is as much of a focus.

The Trade Desk in turn has omnichannel capabilities (see an ad on Hulu and can follow up with an ad on a blog you visit or Spotify podcast you listen to) and can theoretically “close the loop”, which means track when a sale actually happens through Retail Media Partnerships.

TTD has built several related tools around this, including OpenSincera for publisher and inventory data, PubDesk for publishers, and OpenAds for improving auction transparency.

5) Deal Desk

This is essentially a digital version of an old way of buying ad inventory.

It allows publishers to sell bulk ads of special inventory and with discounting pricing.

Deal Desk is TTD’s product for managing these private marketplace deals.

It helps buyers organize, discover, and evaluate deals that historically have been cumbersome to set up and difficult to scale.

This is particularly relevant in Connected TV, where private deals represent a meaningful portion of the market.

6) Measurement

Finally, TTD is building out a broader measurement framework to help advertisers understand whether their campaigns actually produced the desired business result.

A major part of this comes from retail media networks. Retailers like Walmart, Target, and others have purchase data that can help connect an ad exposure to an eventual sale.

Trade Desk effectively buys the relevant data set on behalf of their advertiser to help target ads and eventually “close the loop” to track if it resulted in a purchase.

So if someone sees an ad on a streaming service and later buys the product at a retailer, TTD can use retailer data to help measure whether that advertising actually drove the purchase.

TTD also uses advertiser first-party data and other measurement partners to connect campaigns with outcomes like website visits, store traffic, and sales.

This is increasingly important because vertically integrated competitors like Amazon have a natural advantage: they often own both the advertising data and the transaction itself.

So why is The Trade Desk down 90%?

Whenever a company is down significantly, there is usually something wrong with the company. It is the investors job to figure out if that is temporary or something structural.

Management teams will always have their own reasons as to what is happening with the business.

They can be right, bias, or hiding the true extent of the issues they see.

Sometimes the management teams themselves can earnestly believe their stories but still miss a bigger shift.

The hard part as an investor is you have to decide if you want to take management’s explanations at face value or dig deeper. A good investor will cautiously listen to what managers say (especially if they are as smart and have as much skin in the game as CEO and Founder Jeff Green has), but at the same time decide for themselves if it’s the full story.

Very rarely will a manager come out and say “our business is in decline and we are trying to fix it, but we aren’t sure if it will work”. Now we are not saying that is what is at play here with The Trade Desk, but just noting that investors have a reason to be cautious.

The reasons management attributed to their decline are a combination of macro weakness and execution issues.

First, management has pointed to weaker advertising demand in several large categories, particularly automotive and consumer packaged goods (CPG).

These are meaningful verticals for The Trade Desk, so softness from large advertisers in those areas can have an outsized impact on growth.

These categories (especially automotive) are “top of funnel”, which means that this ads are to drive general brand awareness rather than a direct sale today. We will pick up on why that is important in a bit.

They have also blamed a number of self-inflicted execution problems.

The biggest was a major reorganization that changed how TTD covered large brands and agencies, creating overlapping responsibilities and disrupting some customer relationships.

At the same time, the migration from Solimar to Kokai took much longer than expected.

The company has also cycled through multiple CFOs and seen turnover across several other senior executive roles, adding to the sense that the organization has been unsettled.

These are all certainly plausible reasons why they are struggling but, in my opinion, they do not explain the full extent of the issues at the Trade Desk.

Last quarter they noted that customer spend among existing customers fell and they guided next quarter to revenues being down as much as -12%. This is the first time outside of 1 quarter during Covid that revenues will have ever contracted.

Bad execution and macro doesn’t explain the full story in our opinion.

Now I want to emphasis that this a hypothesis and so it can be wrong, but I will share what I believe is happening that is structurally pressuring The Trade Desk (and help explains why they are so exposed to macro whereas all of their advertising peers like Meta, Google, and Amazon continue to post stellar revenue growth).

My hypothesis is that while The Trade Desk has always talked about being performance focused, what they means to advertisers has shifted from metrics like reach, frequency, viewability, and even CPA to proving actual business outcomes like sales and Return on Ad Spend.

This new shift the Trade Desk is poorly suited for currently.

Thus, it has gotten harder for The Trade Desk to keep up with the walled gardens to prove strong ROAS and the most performance sensitive advertisers are stepping up spend with Meta, Google, and Amazon at the cost of The Trade Desk.

On the other end, advertisers who are less ROAS sensitive are balking at the premium prices that Trade Desk charges, which optically increased with Kokai, and instead moving to buy “cheap reach”.

Let us first set the stage more though.

First, Amazon has been improving as a competitor in CTV with their DSP allowing advertisers to buy ads across Netflix, Disney, and many other streaming platforms.

Importantly, when someone does use Amazon, they can access Amazon’s ecommerce data to better target ads and Amazon’s purchase data will help them prove when they drive sales.

Furthermore, they often offer steep discounts on ad purchasing, undercutting The Trade Desk because they don’t need to make money off the ad buying.

Amazon also has clean data rooms so advertisers can bring their data and benefit from Amazon’s data and closed loop ecosystem without sharing that data with Amazon. (While some advertisers will not doubt be skeptical of Amazon, the pull of using Amazon’s data and getting closed loop attribution is a strong enough value prop for many).

Below we see that Amazon Audiences helps inform ad buys on Netflix.

The Trade Desk might take 20% of the ad spend and Amazon can take 5% or even 0% in some cases. Since Amazon owns inventory, the inventory sold is pure margin for them. Additionally, since they own the data, they don’t have to pay for 3^rd party data providers.

This is one pressure that hits from two sides: 1) Amazon can prove ROAS better and 2) it is cheaper is an advertiser just wants to buy a ton of impressions.

(On the last earnings call Jeff Green says the real math is 8% for Trade Desk vs 4% if you are just including the platform market up and not other value added services, but the fact that he has to make this argument proves the point: advertisers feel like they are far more expensive.)

The next thing that happened with the roll out of Kokai was that it used AI to help deliver better results for advertisers, but in the process, it usually purchased more data or used other services from The Trade Desk.

While The Trade Desk would argue this improved the ROI, from the advertiser’s perspective, their costs went up.

For the advertisers that are ROAS driven and can prove that return, this was okay. For those less sophisticated advertisers or those looking for other metrics, it felt like prices were increasing and instead they diversified with cheaper alternatives.

Now the next piece of the puzzle is how advertiser expectations have changed overtime.

The Trade Desk has always said they were focused on performance, and that is true.

If you go back to their 2017 Investor Day, they were already showing case studies about lowering Cost Per Acquisition. One insurance company went from spending $250 to acquire a customer to under $100. A bank lowered CPA by 44%.

Their CTO even said they had “basically built this whole company by winning campaigns on a CPA head-to-head every step of the way through.”

So performance was always part of the pitch.

What changed was how advertisers measured performance.

Historically, advertisers cared more about metrics like CPA, clicks, viewability, video completion rates, reach, and frequency.

On CTV, for example, one of TTD’s biggest selling points was incremental reach. If you could show an advertiser that 40% of the households reached through streaming were people they were not reaching through traditional TV, that was valuable.

The important thing about most of these metrics is that The Trade Desk could measure them itself.

Over time though, as advertisers got used to seeing actual ROAS from Meta, Google, and Amazon, the measurement standard moved much closer to actual sales.

We can see in commentary across the Trade Desk’s earnings calls and events that they have shifted how they talk about performance.

Claude counted how many times each of these words/ concepts showed up across over a decade of transcripts and we can see that viewability and video completion rate stopped being mentioned over time, replaced by return on ad spend and conversions.

(it counts words per 10k to account for the fact that they have had more events recently)

The shift really starts showing up around Solimar (prior advertising platform) in 2021.

Jeff Green talked about marketers needing to prove “the ROI of every advertising dollar” and described Solimar as a “closed-loop system.”

At the same time, retailer data started becoming much more available.

Walmart, Walgreens, Kroger, Albertsons and others actually know what consumers buy, which finally gave TTD a way to connect an ad shown somewhere else on the internet to an eventual purchase.

Green basically described this himself in 2023 when he said measuring all the way down to the actual purchase was “that last piece that has been missing.”

While this has been a focus for years, the general growth of connected tv through the pandemic (a lot of cord cutting) and not great alternatives meant that it wasn’t the biggest priority for connected TV buyers. There were not great solutions to the problem.

So it is not that The Trade Desk suddenly became bad at performance advertising, it is that the scorecard moved toward something the walled gardens are structurally better positioned to measure.

Now why did this show up just in the past year?

My hypothesis is that it was a mix of Amazon showing they can help advertisers with a ROAS through connected tv combined with weak macro.

When macro is weak, top of funnel advertisers pull budgets first.

If you are an advertiser and you can see how an ad is directly generating you revenues, you will never pull that because it will just hurt your top line.

But top of funnel is hard to measure and doesn’t have the same direct impact to sales.

Trade Desk is more exposed to this sort of advertiser.

Trade Desk said as much on the last call noting weakness in CPG and Auto, classic top of funnel advertisers.

There is one line in the last Trade Desk earnings call that helps support this thesis. Jeff Green says give them “insights that make it so they are truly proving incrementality”. Incrementality means driving incremental sales, i.e. proving the ROAS.

Next quarter they have guided 3Q26 revenues to be at least $650mn, which is about 12% less than $740mn they did in 3Q25. This to me is further evidence that they are not meeting the advertisings current expectations from true performance advertisers, despite the long stated focus on “performance”.

So while management has blamed their short fall to macro and execution, I see it more as them being structurally disadvantaged as performance advertisers who can prove ROAS, which is what the market is demand for premium advertisers.

There is certainly a lot of advertising that is not performance driven, but The Trade Desk is seen as too expensive for that market. These advertisers are not convinced by The Trade Desk’s optimized ad buying, which might actually be worth the premium pricing because of superior targeting, but since they can’t prove it drives sales, they instead opt for “cheap impressions”.

On the call, they note that some advertisers have become more focused on buying cheap media rather the best media, but I think these advertisers might simply be disillusioned by the high fees to buy ads that they can’t prove are working.

Below Jeff Green calls this deliberately short-sighted and I think he is right. Paying the premium is probably worth it for a better targeted ad, but that argument just isn’t landing with these advertisers.

So where does The Trade Desk go from here?

1) Measurement.

The biggest thing they need to do is close the loop with measurement.

TTD is rolling out a new measurement framework designed to help advertisers prove incrementality.

The Trade Desk has spent years proving that it can buy ads efficiently. What it increasingly needs to prove is that those ads actually created business value.

If they can do that consistently, the premium they charge becomes much easier to justify.

2) Data marketplace

The next piece is their retail data marketplace.

Amazon has a huge advantage because it knows what people buy on Amazon. But The Trade Desk has partnerships with retailers that management says represent more than 80% of sales from top U.S. retailers.

That means TTD can potentially measure purchases across Walmart, Target, Kroger, Walgreens, Albertsons, and others rather than only within a single retailer.

For a large CPG advertiser, that could actually be a better measurement surface than Amazon, if TTD can stitch together the data.

3) Audience Unlimited.

Then there is Audience Unlimited.

One of the problems with Kokai was that better AI decisioning often meant using more paid third-party data, which could make the platform feel increasingly expensive.

Audience Unlimited changes that by bundling access to a large amount of third-party audience data for a flatter fee.

The idea is simple: let advertisers use more data without feeling like the meter is constantly running.

The beta test showed 38% lower data cost, 30% lower CPMs on Media buys, and a 2.7x increase in conversions.

4) Joint Business Plans.

Another positive is the growth in Joint Business Plans, or JBPs.

These are essentially deeper strategic relationships with large advertisers where TTD works directly with the brand on how to expand and improve its use of the platform.

Management says JBPs now cover more than 200 clients and are 6x the overall faster than the overall revenue growth rate.

5) Deal Desk.

Deal Desk is another important response that was launched in 2025.

Rather than fighting the shift toward private deals and Programmatic Guaranteed buying to get “cheap reach”, TTD is now building tools to help advertisers manage those transactions inside its own platform.

That matters because even if advertisers want simpler, lower-cost buying, TTD does not necessarily have to lose the spend altogether.

There are also several areas of the business that are still growing strongly.

6) Other.

Audio has become one of TTD's fastest-growing channels, helped by relationships with platforms like Spotify.

International is also finally becoming more meaningful, with EMEA and Asia Pacific growing much faster than the overall company (+30%) and international revenue is becoming a larger percentage of the business from 13% in 2023 to ~17% today.

And finally, Google appears to be putting less emphasis on the open web.

Much of DV360's growth has increasingly come from YouTube rather than buying ads across the broader open internet.

If Google continues to pull back from that market, it leaves The Trade Desk in an even stronger position as the largest independent DSP.

So the bull case is not that nothing is wrong, clearly they are having material issues.

It is that TTD can improve their technology to close the loop, utilize 3^rd party data, and help advertisers spend how they want to.

Now it should be noted that closing the loop has been mentioned since 2021, so it is no easy feat, but it now it seems to be a real focus.

Valuation.

At a $13.40 stock price, they have a $6.3bn market cap.

However, they have no debt beyond $350mn in leases and with $1.4bn in cash, they have a $5.2bn enterprise value.

LTM they have $585mn in operating profits, putting them at an 11x EV/ NOPAT multiple.

You don’t need many great things to happen in order to get a return from hear as an investor.

However, the fear is with shrinking revenues and operating deleveraging, earnings could continue to fall. Where is the bottom?

We won’t know until we see these initiatives turn around or macro start to work in their favor.

If they could stabilize revenues return to just a high single digit grower a 20x multiple could be fair.

If they can really capitalize on the opportunity though, the sky really is the limit with the advertising TAM currently around $1 trillion versus their LTM revenues of $3bn.

Valuation really isn’t the question here, execution is. This is a turn around which inherently means you need things to work that currently aren’t working for them.

Many investors will find this to be a risky proposition. Others may see it to be a great opportunity.

One thing is clear though: Jeff Green certainly believes in them.

He purchased $150mn of stock at a price higher than today—around $25. This puts his total ownership around 8% of the company.

Ultimately, it is up to you to decide if you want to join him.

For more on The Trade Desk, check out this video below.

YouTube FAQ

Nothing in this newsletter is investment advice nor should be construed as such. Contributors to the newsletter may own securities discussed. Furthermore, accounts contributors advise on may also have positions in companies discussed. Please see our full disclaimers here.