Happy Friday 18.01.19

Dinosaurs & Custom Algorithms: 2019 Is Already Awesome!

Happy Friday! This week is an opinion packed opus:

First off this week we look at whether PMPs, which started to fall out of fashion with some buyers last year, will go extinct in 2019 & what would happen in the market if they did? Next up we cover my favorite article this week, written by @cadielisejones from Beeswax, on whether brands should build their own algorithms, it’s a must read.

Then I hit pause on my buyside naval gazing & we turn our attention to the news that Double Verify have acquired contextual specialist Leiki & discuss whether building out their verification tools is enough to stop some buyers using the free options in their DSPs.

Finally something a little different: I postulate whether mounting governmental & consumer pressure will create a market where ALL browsers (including Chrome) block the very cookies that the programmatic ecosystem is predicated on… It’s a long newsletter this week, but a good one, Enjoy!

Will PMPs Go The Way Of The Dinosaur?

There was a great article this week in AdExchanger by one of my favorite reporters @SarahSluis:

It asks whether in a world where the open marketplace is safer than ever thanks to ads.txt & fraud refunds, if PMPs are going could go the way of the dinosaur? This is a fair question & one that many traders have been discussing for a few years now.

It’s fair to assume basic private auctions, that offer the same inventory as the open marketplace are going to go away. They carry with them operational overheads for both the buy & sell side & also oftentimes higher clearing prices as publishers tend to floor more aggressively.

For a couple of years many publishers have been offering custom formats, or have been bundling data with their inventory, as a way to incentivize buyers to use their private deals, how defensible the latter is long term vs the OMP & the data offered by the giants’ buying platforms remains to be seen.

My personal view is that long term the market will settle on the open marketplace for the majority of what advertisers want to buy, with programmatic & audience guaranteed buys used for inventory with high sell through rates. PG will grow in importance as demand around inventory such as video increases dramatically & this will play into the hands of those with buying scale & power.

I think what becomes of independent/specialist advertising exchanges once the above happens is the really interesting question though:

Why? Well Google & a couple of other advertising exchanges dominate the PG space, I don’t see this changing & this means independent advertising exchanges may well be left fighting it out for open marketplace spend & having to do so at ever decreasing take rates. That’s a hard place to be, & I think this (& other factors) will drive consolidation amongst exchanges in 2019 & beyond.

Should Brands Build Their Own Algorithms?

Cadi Jones the commercial director in Europe for DSP @Beeswax had a fantastic article in Campaign this week that explores whether brands should build their own algorithms:

If you’ve not read this article yet it is a must (must) read & whilst I do disagree with the sentiment that algorithms within most DSPs currently are “one-size-fits-all” (most learn over time, so are custom to an each strategy) I can definitely see the appeal of customization ahead of activating campaigns, especially for advertisers/agencies with deep data science expertise. Outside of customization I think Beeswax holds two interesting advantages over some of their competition:

As programmatic becomes the execution layer for more expensive inventory like OTT & Audio, paying on a SaaS basis rather than a % of media fee could be a big advantage, as in some circumstances the latter (may) mean comparatively less spend goes to working media, if the incumbents DSPs take rates remain the same. Longer term I think the second benefit referenced in that Tweet (transparency) could start a revolution.

Why? It’s true that 2018 was the year of transparency in programmatic, but only in a fairly narrow sense, yes there is more transparency in the supply chain then ever before (which was incredibly positive) but there’s still not enough transparency when it comes to the decision making processes used by DSPs about the impressions they do & dont buy.

Whilst I feel the concerns in the article around data sharing between brands are largely unfounded, we should definitely pay attention to the levels of transparency around internal DSP auctions, more clarity on what a DSP does & doesn’t buy would be great & here I think here Beeswax could nudge the incumbents in the DSP market it is looking to disrupt, in the right direction… Just as Amazon have recently done with viewable attribution. Again proving that competition is a good thing.

DoubleVerify Acquires Leiki

As the market has matured the buy side has become increasingly sophisticated in their use of the brand safety tools that sit within a DSP, to keep their clients ads away from the bad stuff out there. Equally the tools in these DSPs have consistently got better over the past half a decade & because of this competition, third party verification vendors have to work harder than ever before to gain an edge over the native tools within a DSP which are often offered at no extra cost.

This week DoubleVerify gave their verification stack a big upgrade by acquiring contextual specialist Leiki for an undisclosed fee:

Leiki uses sophisticated AI to understand the context of a page, this allows advertisers using the product to avoid non brand safe content in brand safe domains, such as hard news. This could be used for example to help a travel brand avoid articles about a plane crash but remain present in news outside of this.

Will this improve DV’s product? Almost certainly & it’ll be super interesting to see how they integrate Leiki with Pinnacle their platform which gives buyers a real-time view of their media buys from a qualitative perspective. But will that be enough to help them grow their market share?

In my eyes that remains to be seen, as I mentioned DSP tools are improving all the time & competition amongst third parties in this space is stiff: Just last year we saw Oracle combining the keyword specialist GrapeShot that bought, with (DV competitor) MOAT which it acquired in 2017. Again competition is good & it should spur the market on, meaning innovation & tangible benefits for advertisers.

How Long Before All Web Browsers Block Ad Tracking?

The past few years have seen mounting governmental & consumer pressure for improved online privacy. We’ve seen that with laws in Europe (GDPR) we’re seeing this in the US & on Monday @atbwebb from Bloomberg had the scoop that German lawmakers plan to stop Facebook from collecting user data on websites that have a Facebook Like button:

Facebook aside, calls for increased online privacy have reached fever pitch recently & not just in Europe, yesterday in the US Apple CEO Tim Cook called for the FTC to establish tools that would enable users to delete data on demand:

Because of these concerns, browsers are increasingly blocking cookie based tracking upon which programmatic targeting & attribution is predicated: Apple’s Safari has the tracker hostile ITP-2, Samsung & now Firefox’s new browsers offer similar features. These web browsers are used by ~ 35% of internet users & one wonders how long it’ll be before Chrome developers begin to feel the rising pressure & add in similar features too?

If they did it’d be the biggest story of the year, it’d mean 85% of browsers would block online trackers & that could be a real challenge for the industry. Given Google is an advertising business you might think this is all a little far fetched, but consider this: Only last week Chrome rolled out the blocking of “bad” ads globally:

Consumers are becoming increasingly aware of privacy & Google may feel they need to offer something or lose users to other browsers.

Whilst this would likely cause issues for the advertising side of the business, it’s not a stretch to assume they would be less challenged than others in the industry to deal with this as they own a browser with lots of signed in users & have an excellent privacy center that would no doubt allay consumer concerns over the browser itself tracking them (if they chose to go down that route). Worth noting their privacy site already lets you opt out of Google ads personalized by cookies.

I’m not saying all of this will definitely happen & I obviously know what Chrome’s developers are going to build, but given the way things are going it could happen. Let me know what you think on Twitter, @SimonJHarris

Okay that’s the second newsletter of 2019 done.

It was a long one this week, thank you for making it this far!

If you want to get in touch you I’m on:

Twitter here: @SimonJHarris Or LinkedIn here: linkedin.com/in/simonjulianharris/

Interested to hear if you like the thought pieces, or if you think I should stick to news!

If you know someone who might like to receive next weeks newsletter send them here:


Thanks & have a lovely weekend