Some thoughts on recent adtech IPO The Trade Desk


The below is only the author’s opinion – it is the author’s best attempt to portray things accurately, but it may contain errors.

The below is not a recommendation, and it is also not investment advice. Please do your own work. The author is currently long shares of TTD.


The Trade Desk (“TTD”) is a company that sells software to ad agencies that helps them optimize their bidding for programmatic ad inventory. What they are trying to do is to allow agencies (on behalf of advertisers) to sort through billions of potential impressions in order to choose which to bid on and buy. TTD charges the users of their technology a percentage of what they spend (slightly below 20%).

They were founded in 2009, are based in California, and recently completed their IPO.

One important characteristic of TTD is who they serve (and who they don’t serve): their customers are ad agencies (e.g. WPP LN, OMC, etc.), as well as other entities that buy advertising as a service for others (e.g. TTD has mentioned serving Rakuten, the Japanese e-commerce company, when Rakuten buys ads on behalf of its merchants leveraging the data from its platform). TTD recently said that less than ½ of its business is from ad agencies, as opposed these other entities.

Unlike some of its close competitors, TTD (1) serves only the buy side (i.e. it does not serve publishers), and (2) does not directly serve brands as customers. As discussed below, these are points of competitive differentiation, since they do not have conflicts of interest and are not competing with agencies to serve the advertiser.

As stated above, TTD is paid a percentage of what is spent through their platform. One way to think about the composition of their business is to split gross spend by the type of advertising product — today TTD’s business is roughly:

  • Slightly below 50% display — at inception, this was 100% of the business but has since (especially recently) been shrinking as newer products grow very quickly
  • Roughly 30% mobile, though much faster growth (205% YoY in Q3 16)
  • Roughly 25% video, though much faster growth (185% YTD in Q3 16)

In terms of geography, TTD is currently only around 10% international, though on their most recent call they reported international is growing at 4x the rate of their domestic business, and they are in the early stages of building out their international footprint.

Industry and Competitive Position

The ad tech industry is a very complicated landscape — to get a sense of this, you can look at these LUMAscapes, which are visual overviews of the industry put together by LUMA Partners:

For purposes of my writing here, I’d summarize the relevant parts of the ad tech landscape as such:

  • Brands: end buyer of ad inventory, either spending through an agency or spending directly themselves, as part of an advertising campaign
  • Agencies: service provider for brands who (among other things) buys the ad inventory needed by the brand
  • DSP’s, or demand side platforms (which is what TTD is): they are the ones who provide the tools needed to actually buy the ad inventory — they can serve brands, agencies, or both
  • Ad networks and ad exchanges: pools of ad inventory where supply and demand are collected
  • SSP’s, or sell side platforms: they provide tools to publishers to help them sell their inventory
  • Publishers: creators of content against which ads are served

In terms of TTD’s most direct competitors, there are four that seem most relevant:

  • GOOG, with its DoubleClick Bid Manager product
  • MediaMath (private)
  • AppNexus (private, set to have IPO soon)
  • DataXu (private)

I say most relevant because (as can be seen on the LUMAscape) there is a long tail of DSP’s — the bulk of these are not that relevant to TTD because they are some combination of small and / or focused only on type of ad inventory (i.e. only video, or only mobile, etc.). The four competitors mentioned above are most similar to TTD in terms of being large players that offer a relatively comprehensive solution.

There are several reasons why I like TTD’s competitive position.

First, the industry itself inherently has nice qualities, meaning TTD (along with GOOG / MediaMath / AppNexus / DataXu) all benefit from certain advantages. These nice qualities stem from these being software businesses, so there are economies of scale and switching costs. For some detail on economies of scale, a few recent comments from TTD’s CEO are helpful

  • One point is that it costs money to look at each impression — therefore, as a DSP you are better off if you can leverage the cost of looking at that impression over a wide range of potential people who can buy it; if you are instead a small DSP or a brand trying to do it yourself, you have the same costs to look at each impression, but you’re not going to buy nearly as many of the impressions (since you’re only one source of demand)
  • Another point, for what it’s worth, is that the CEO said if you gave him $500M and asked him to start today, he couldn’t build a competitive business

Second, TTD specifically seems like a good operator. Some data points here are

  • Very strong reviews on Glassdoor:
  • Strong emphasis on customer service — TTD talks about this, but also other sources mention it as well (e.g. trade publications like Ad Exchanger, service provider ranking firms like Forrester)
  • For what it’s worth, TTD claims their technology allows for the most “expressiveness,” which is their term for how many ways you can slice the impressions you’re looking at in order to find the group that you buy; side note: unlike many of the other points here, this is not something I’ve seen third-party evidence of, but it’s something that TTD emphasizes as an advantage they have

Third, TTD’s strategic positioning (as compared to GOOG / MediaMath / AppNexus / DataXu / others) is helpful. TTD is fully aligned with the buy side (and specifically agencies and others who buy on behalf of advertisers) because of two characteristics

  1. They only serve the buy side, as opposed to also serving the sell side (like GOOG, AppNexus, etc.)
  2. They only serve agencies and others who buy for advertisers, as opposed to serving advertisers directly (like DataXu, now struggling but formerly strong competitor Turn, etc.) — see this article for more detail:


I think TTD management, especially the CEO Jeff Green, is very good — the most important points to me are

  • Green is well-aligned with shareholders — he owns roughly 25% of the company
  • As you will see if you listen to their earnings calls / conference presentations, Green is explicitly focused on the long-term — some examples of this are
    • He is investing today both geographically (various international markets) and by product (video including TV, mobile, audio, etc.) even though it will depress short-term margins
    • On the Q3 16 call, TTD talked about the “long-term” twelve times (excluding non-relevant uses like long-term tax rates, etc.)
  • He believes in ensuring that other stakeholders (e.g. customers, suppliers, etc.) derive value from their relationship with TTD — some examples of this are
    • When asked about working capital needs for TTD at a conference, he talked about how he hasn’t used the (growing) leverage he has with both customers and suppliers to improve his payment terms, because he doesn’t want to risk upsetting TTD’s long-term growth opportunity
    • He also mentioned that some of his employees have asked him why TTD doesn’t raise prices, and he said he wants to make sure customers derive a lot of value from using TTD
    • As a side note, this type of mindset to me is something I like, because you know as a shareholder that there is something being left on the table (like COST with their philosophy about pricing)
  • He is very focused on expanding the moat around his business — e.g. it is not often that you hear a CEO excited to report (as Green did on the Q3 16 earnings call) that he can invest today to build a moat in Indonesia (because TTD is so profitable) while competitors cannot

A point on insider transactions: in 11/16, recently-appointed (8/16) director Kate Falberg spent about $2.5M to buy 100K shares. The purchases were at prices a little under $26 per share.

Historical Ad-Tech Problems

One important point is the history of publicly-traded ad-tech companies. A number of ad-tech companies have IPO’d and then had significant declines in value — e.g. FUEL, TRMR, RUBI, etc.

One problem common to many of these failures (and an issue that has also caused problems for private competitors as well) is that many ad-tech companies essentially relied on an arbitrage model. They would look to source ad inventory at a low cost and then sell that inventory at very large mark-up’s. This business model subsequently has fallen apart for a couple of reasons

  • The first and most important is that as the industry has become more transparent and knowledgeable, it’s become harder to find pockets of inventory where you can earn a large arbitrage spread; in very broad terms, the ad tech world has moved from having many fragmented and opaque pools of supply to one that is increasingly operating as a with consolidated and transparent pools of supply
  • The second issue is that if your business model is premised on finding high margin arbitrage opportunities (like the companies mentioned above), it’s inherently harder as you grow (similar to Buffett’s idea that, to paraphrase, your 20th position is likely to be materially worse than your favorite); unlike TTD where growth leads to higher margins, these arbitrage-focused competitors had negative impacts on returns as they grew

As a side note, I’d guess that one reason such an attractive (in my opinion) opportunity exists in terms of TTD’s stock is that past public company ad-tech failures are coloring investor perceptions of TTD, even though TTD’s business model is very different.


Market cap calculation works like this

  • 5.4M A shares (1 vote per share — these were the IPO shares) and 33.2M B shares (10 votes per share)
  • 5.2M options at $2.96 strike
  • So we have 38.6M shares at $30.77 is $1,188M, and the options are an additional $145M — total market cap is $1.33B

Other things to consider are

  • Cash of $124M
  • Debt of $51M

Then EV is $1.26B (small side note: remember given that they haven’t used the growing leverage they have on W/C, future W/C trends will probably lead to more excess cash and a lower EV).

In terms of profits, there are a few things to consider.

We can start with 2016, using their Q4 16 guidance. This implies

  • $193M of revenue
  • $55M of (their definition) EBITDA
  • $50M of EBIT (assuming $3.6M of D&A and $2M of stock comp)

On the above numbers, TTD trades at

  • 6.6x 2016 sales
  • 25x EV / 2016 EBIT multiple

However, there are a number of things to consider in the above.

  1. I’d guess that their Q4 16 guidance is conservative — some reasons are
    1. Just my opinion, but the CEO seems concerned with under-promising and over-delivering
    2. Guidance implies 45% YoY growth in Q4 16, compared to 84% in Q3 16 and 83% YTD Q3 16 — while Q4 16 will only have a partial tailwind from political spend (given the timing of the election), this seems like too big of a slowdown, especially given the that higher-growing categories (i.e. video, mobile, audio) are becoming a higher portion of the comparison base
  2. EBITDA margin (28% at guidance), while impressively high given their current growth rates, is likely well below steady-state levels — several reasons to believe this are
    1. They talk about being able to do 40%+ EBITDA margins once they get beyond growth investment headwinds
    2. Business did 34% EBITDA in 2015 when there was less headwind from international / product investments
    3. Business is inherently very scalable (CEO emphasizes this a lot), and is experiencing rapid growth
  3. They are growing very rapidly, so forward multiples (depending on growth assumptions you make) are much lower than 2016 multiples
  4. To make the above examples more applicable, we could assume
    1. $200M in 2016 revenue (if they are being conservative)
    2. 50% growth in 2017 and 40% growth in 2018
    3. 40% run-rate (i.e. they may not be there in 2018 but in recognition of what they would do without the big growth spending) EBITDA margins in 2018, and similar D&A / sales and stock comp / sales ratios as today
    4. With the above numbers, that $156M in EBIT on $420M of sales, making the multiples 8x EBIT and 3x sales
  5. Also note that being headquartered in CA, tax rates at current corporate rates are very high (probably in the low 40’s)
    1. However, there is significant upside to any cut in the corporate tax rate
    2. I don’t explicitly put in a P/E, but you can just use the EBIT number and multiply it by whatever you think the long-term tax rate will be


There are a number of ways to look at growth.

The first thing to do is to back out the gross spend going through TTD — since their revenue is slightly less than 20% of what is spent through their platform, we can say that 2016 guidance revenue of $193M implies about $1B of gross spend.

Before getting into the various ways of slicing their addressable market (and therefore growth opportunity), note that data here is only approximate. Some problems with data include

  • There are various definitions of what exactly counts as programmatic ad spending
  • A portion of what people call programmatic ad spending goes to publishers / sites where TTD does not transact, i.e. some of the walled gardens like Facebook, Youtube
  • Some spending is by advertisers directly (i.e. not through agencies), so TTD wouldn’t compete for that portion

With the above caveat, we can get a rough sense of their opportunity. A Magna Global report from 6/16 ( projects a global market for programmatic banner display and video (this subset is most relevant to TTD since it excludes search and social, which genearlly aren’t relevant to TTD) of $19.5B in 2016, growing to $36.8B in 2019. These numbers would imply that TTD captures about 5% of the spend in its current addressable market, meaning it has a long runway just from capturing more spend, not mentioning the fact that its market is itself projected to double in the next three years.

Also note that TTD’s recent very high growth rates reflect being in the early stages of such a large opportunity, and TTD’s mix is shifting so that the (much) higher growth product categories (video, mobile, audio) and geographies (international) are becoming a larger portion of the business.

While the above growth opportunity seems very attractive, there is an additional very large upside option from the potential for programmatic TV. TV has been slow to shift to programmatic because (among other reasons):

  • Suppliers have greater power
  • The historical way of transacting is much more entrenched (due to the long history, amounts of money involved, etc.), meaning there are stronger vested interests in the status quo
  • TV’s (especially historically) did not have the same type of data-gathering ability as computers and phones do

With that said, there is an expectation that TV will eventually go programmatic, though the pace of change to this point has been slow. From my perspective, as show in the above numbers, even without TV going programmatic, TTD has a very large growth opportunity. TV then adds a very large option on top of that — i.e. US TV spending is around $70B per year, and TTD can also compete for international TV spending over time as well.

Said another way, as the CEO likes to say, their focus is on the opportunity to capture as much of the global $640B of advertising spending as they can. From our perspectives as investors, there are certain pieces of that pie that will probably never be open to TTD (e.g. agencies control only roughly 80% of spending; the portion of advertising spent on search is not likely to ever be accessible to TTD; etc.). However, against TTD’s current roughly $1B of gross spend, the opportunity is still very large.


There are a number of risks listed below, as well as how I think about them

One issue is that TTD does not have access to (or in some cases chooses not to link up with) a few important sources of supply, e.g. Youtube, Facebook, Snapchat. In some cases TTD is not allowed (e.g. Youtube), while in others they have chosen not to participate given the limited amounts of data suppliers give them access to (e.g. Facebook, Snapchat).

To the extent that you think Facebook and Google are going to take all or most of online ad spending, this is negative for TTD. At this point I am not very concerned about this issue because

  • The growth opportunity outside of these certain suppliers still seems very large (relative to TTD current business size) and growing
  • Facebook, etc. may face increasing pressure over time to make their data more accessible to advertisers, given recent measurement issues as well as potential pressure from large advertisers; this is a potential source of upside for TTD, since there is a possibility that over time currently off-limits inventory comes onto their system

The second issue to consider is that display advertising is a mature market. This is not a large concern for me given that

  • Display today is less than ½ of their business, and the other pieces are growing extremely fast
  • Display itself is still not close to 100% programmatic, so there is still a tailwind in display as the share of display that is programmatic moves towards 100%

The third issue to consider is potential pushback on the percentage of spend that TTD keeps. The reasons I am not concerned by this include

  • CEO has talked about this percentage actually increasing over the history of the company (it started around 15%)
  • As TTD scales and the DSP industry consolidates, they increasingly have more power over customers — their biggest customers in 2015 were OMC (12% of gross spend from all OMC agencies since TTD has a contract with the OMC holding company) and Mindshare (12% of gross spend, though there is more from other WPP agencies where TTD has separate contracts)
  • CEO has talked about employee suggestions of raising prices, which implies they may have pricing power

A fourth issue is the potential for advertisers to no longer use agencies over time. This is not much of a concern to me because

  • As CEO says, programmatic advertising is much more complicated than historical ways of buying advertising — if advertisers needed agencies in the past, it’s unlikely that they will be able to do things themselves now
  • Just empirically, agencies have continued to occupy an important place in the ecosystem — in fact, many competitors (most prominently Turn) suffered by trying to go directly to brands and cut out the agencies

One nice aspect of TTD’s business is that I’d expect it to be fairly acyclical. Its revenues are a percentage of what is spent through its platform — since the overall ad spending pie itself does not drop much in a recession and since TTD’s platform continues to greatly expand its share of that pie, TTD’s revenue should not be much affected by a recession. And then given its high margins, small changes in revenue have little impact on profits.

Resources for Further Learning

Here are a variety of resources for learning more about TTD and the industry

Transcripts and conferences

Industry publications

  • AdExchanger
  • TheDrum
  • Beet TV
  • Exchange Wire
  • Digiday

Competitor viewpoints


The above is only the author’s opinion – it is the author’s best attempt to portray things accurately, but it may contain errors.

The above is not a recommendation, and it is also not investment advice. Please do your own work. The author is currently long shares of TTD.

5 thoughts on “TTD”

  1. Thanks for the write-up. What do you think of the current valuation?

    Also, do you have the source for this: “TTD recently said that less than ½ of its business is from ad agencies, as opposed these other entities.” I can’t find that anywhere, it seems to be mostly agencies to me? Except some direct selling to brands, in cooperation with agencies.

    1. Sure, glad you liked it.

      In terms of valuation, here’s what I see. Results to date make me more confident that the revenue and profit guesstimates in the original post for 2018 are reasonable (but of course they are still just guesstimates). At $50 share price, EV is roughly $2.1B, so if it turns out that they can do $420M in 2018 revenue and $156M in 2018 potential EBIT, that’s 5x revenue and 13.5x potential EBIT.

      In general, I think of any stock in relation to my opportunity costs, i.e. what else I would put the money in. Not many things that I can find look better to me than TTD, even after the price run-up. If you have things that look more attractive, please let me know!

      Regarding the source for the point about agencies, see the Needham webcast linked to in the original post (the point in question is discussed starting around the 21:00 mark). The other roughly 1/2 is not brand-direct, but rather through entities that aren’t agencies in the traditional sense (traditional agencies being Omnicom, WPP, etc.), but others are providing agency-like services to some group of customers.

      1. Thank you. I agree that it looks interesting. I only get “this video has expired” when trying to see the Needham presentation, unfortunately.

        Do you know much much of TTD’s business runs/used to run through Facebooks FBX? They are (have?) shutting it down to only sell ads through “walled gardens” (correct me if I’m wrong), and I read somewhere that the FBX exchange was one of TTD’s biggest inventory sources – it doesn’t seem to have affected them much so far, if it has already been shut down, though. It’s a big worry for me that the best inventory sources online is not available through TTD (Google Search, YouTube, Facebook).

        1. As I understand it, FBX has been closed for some time (e.g. The loss of access to FBX would only have limited access to inventory for (i.e. desktop ads, not mobile) — I don’t think this was ever a big source of inventory. And whatever I think, the numbers are the numbers and they reflect no FBX (and have for some time).

          As I mentioned in my original post, the power of walled gardens increasing is a risk, and one that I think is important to monitor over time. Personally, it’s not a big deal to me at this point given

          1. There is still a huge amount of programmatic inventory outside of the walled gardens, growing fast, and I don’t think that will change.
          2. The risk also has a positive side — as TTD’s CEO mentioned on the most recent call, there is the possibility that over time the walled gardens open up to DSP’s like TTD. As advertisers are able to put more pressure on walled gardens (e.g. recent YouTube problem with brand safety; recent Facebook measurement problems; etc.), I think this increases the chance that they open up over time.

          1. Thank you, I agree with most of what you said. It’s hard to judge how the walled gardens might impact TTD, though. But it’s both a threat and an opportunity of course.

Leave a Reply

Your email address will not be published. Required fields are marked *