Time to Re-Examine Google’s Ad Server?

After more than two years of saying very little about its preparations for GDPR, Google has now made several changes that reveal how things will change for the rest of the ecosystem. During a call with the IAB Europe GDPR Transparency and Consent Steering Committee, Google disclosed that it has a new tool in beta with some DFP and AdSense customers called Funding Choices.  Funding Choices limits the partners a publisher can share consent with to a dozen.

This is a similar consent tool to other Consent Management Platforms like Admiral and Sourcepoint. A full list of IAB-registered CMPs is here.

The Google consent interface greets site visitors with a request to use data to tailor advertising, with equally prominent “no” and “yes” buttons. If a reader declines to be tracked, he or she sees a notice saying the ads will be less relevant and asking to “agree” or go back to the previous page. According to a source, one research study on this type of opt-out mechanism led to opt-out rates of more than 70%.

Google’s and other consent-gathering solutions are basically a series of pop-up notifications that provide a mechanism for publishers to provide clear disclosure and consent in accordance with data regulations.

As a company that began its life as an ad server, we have been struggling to find out whether we play at all in this full-employment scheme for lawyers, since we do not hold data or sell it. The situation is made more fluid because publishers do not have to accept the Google solution, and large publishers like Axel Springer have developed their own CMP technology.

Another announcement made by Google last week seems to have made  multi-touch attribution attribution much more difficult, because as of May 25 google will no longer provide DoubleClick IDs for data from its had server and DSP, or Cookie IDs and IP addresses from its exchange.

According to Martin Kihn, Research Vice President for Gartner,

Without these IDs, exported DCM log files can’t be used to determine true reach and frequency or to build MTA models, which are by definition user-level. MTA is not the only way to measure the true impact of ads but is theoretically the most accurate and provides by far the most detailed results.

Of course marketers are scrambling. But didn’t everyone in the industry expect this? This, after all, is the objective of GDPR, to preserve the consumer’s privacy.  The consumer does not care about the accuracy of Multi Touch Attribution customer campaigns, and for Google especially there is no other alternative. Google doesn’t really have to care about its ad-serving business (DFP), which it acquired over a decade ago and which is responsible for a very small part of Google’s revenue.

And it’s not as though a company the size of Google can slip under the radar of the GDPR, because it has already been fined and I’m guessing that other than Facebook, Google’s going to be under the greatest scrutiny by GDPR enforcers.

Remember, not everybody has to use Google as an ad server.

 

 

Hulu to Allow Ad-Supported Downloads

Last week’s post, “Advertising isn’t Going Anywhere,”  was written a couple of days before Hulu revealed at its NewFront presentation that it was going to allow downloads, and they’d be ad-supported for all of the Hulu users who are on the free plan. My main point was that people can only afford a limited number of subscriptions — far fewer than they’d want. So it’s great for the consumer in several ways that an aggregator like Hulu will have downloading with the ads remaining in the downloads. Hulu has 20 million subscribers, most of whom are on their ad supported plan.

Hulu had a lot of updates at its NewFront/Upfront presentation this morning, but among the most interesting for me was that Hulu will offer downloading of its content, but with ads included. Since the vast majority of Hulu’s 20 million+ subscribers are on the ad-supported plan, this means Hulu is going to be breaking some new ground in downloading, relative to its ad-free SVOD peers Netflix and Amazon, both of which have been offering downloading for a while.

Hulu has lagged its peers by not enabling downloading, but its new feature, whose specific launch date was only identified as “during the 2018-2019 upfront season,” will differentiate itself immediately by having ads included. As we all know, Netflix and Amazon have sucked an enormous amount of ad-supported viewing hours out of the ecosystem, with downloading on these services becoming a bigger contributor. By including ads, Hulu directly addresses this, allowing advertisers to at last follow viewers into their offline viewing experiences. Hulu’s light load of ads, which are unskippable, only enhances the value proposition to advertisers….

Given the rise of ad-free SVOD, more advertisers will struggle to reach their millennial audience targets with 100% brand safety and viewability, making Hulu a prime option. Hulu has also been very smart about scooping up rights to popular library content to augment its still limited (though rapidly expanding) slate of originals, led of course by “The Handmaid’s Tale” which was renewed for season 3.

So if you want to download and view Hulu content, the only way you can do it without ads is to pay for the highest level of subscription, which is Hulu with No Commercials. This is going to create a problem for millennials, who are big Hulu watchers and do not like to sit through ads or pay for subscriptions.  I think it’s a safe bet that, unless the ads become terribly onerous, the convenience of being able to watch such a large variety of popular shows on demand will outweigh the indignity of sitting through an ad. This is a recommendation for the ad industry: don’t kill the goose that lays the golden eggs. If this works, we may see other sites introducing ad supported downloads as well.

Advertising isn’t Going Anywhere

Lately almost every quality publisher is experimenting with paywalls and subscription services.  They are now viewed as a panacea against the need to handle consumer data, and a hedge against the increasing use of ad blockers.

Of course subscriptions to newspapers and magazines were available back in the days of print, too. So why did print publications also run advertising?

One simple reason: subscriptions cannot support most publishers, unless they are in a niche category without a competitor. The Information, a technology business publication aimed at wealthy investors, falls into this category. But most publications do not. Even the Wall Street Journal, the leading financial publication and one that many people WILL pay for, cannot get by  on subscription revenue alone. Nor can the leading news sites like the New York Times and the Washington Post  support themselves solely by instituting paywalls.

Why not? Because the average middle class consumer, who has been reached in the past through advertising, cannot afford to subscribe to everything she needs to read, see, and hear. We live in the Information Age, and very few of us can afford to ignore the need to stay abreast of a very rapidly changing world.

She’d like to get the national and world news from a reliable source, so she visits the New York Times. But finding it too biased, she must also often visit the Washington Post.  Neither of them covers news from her hometown, so she reads her local news site. And since she’s a homeowner, she reads a lifestyle publication. Or a fashion magazine. Or Consumer Reports.

To save money, she cut the cord on cable TV a year ago, so for entertainment she relies on Amazon Prime video and Netflix. But some shows are only on Hulu. Her husband subscribes to ESPN. He’s also a reader of Bloomberg and Wired, and they just instituted paywalls, too.

And then there’s Spotify.

At the end of the year, she and her husband look at their iTunes bill and their credit card bills and realize that they have spent over $2000 on subscriptions to both video and text sites. They blew right through their budget. So they decide to cancel their subscriptions to save money.

Eventually they trade off the need for privacy and gravitate to sites where they can get news and entertainment free, just by tolerating a few ads. In the mean time, publications that have instituted paywalls find that they are experiencing a great deal of churn among their subscribers, leading to unpredictable revenue projections. Despite their paywalls, they must also run some, although perhaps not quite as many, ads.

And that’s how the great paywall experiment plays itself out. Advertising isn’t going anywhere.

 

 

 

Can Ads Work Without Targeting?

We’ve grown fond of thinking that we need data, data, and more data to have ads that work. That data bias has driven the creativity right out of the industry, and with it the consumer’s tolerance for advertising in general. And yet, many ads used to succeed and even still do, without targeting customers through data mining. These are the ads consumers didn’t mind, and they’re the ones that will survive GDPR, Facebook, and all the other horror stories about digital advertising.

“Old school” ads worked by simply raising awareness. Many pharmaceutical ads still work this way. They are usually bought in mass media, which pretty much targets everyone. And the message is something like “if you suffer from X, ask your doctor about Y.” Some of them, like one prize-winning ad for Vytorin, a cholesterol-lowering drug, gave even more detailed information to tell viewers that high cholesterol was due partly to genetics and partly to what you eat. But it was still information.

Some travel ads also simply give information. “Vacation on the beach for only $599 a week.” An ad like this makes almost anyone aware that this advertiser is offering a sale or a good value on a beach vacation. Airlines and hotels often advertise this way during slow seasons.

Yet other ads try to persuade. Occasionally, the persuasion is pretty overt. But ads can also signal things about products and services without being overtly persuasive. For example, SuperBowl ads very seldom try to persuade game-watchers of anything. However, they signal that the company that buys an ad is financially able to afford it, and thus is more trustworthy than some upstart brand. Or, if the brand is buying an ad for the first time, the signal is that there’s a new disruptive company in town, taking over from the incumbents.

Another way ads can work without targeting is by making promises. “If you buy this product, you will look like/feel like/live like the person seen in the ad.” Some of the promises are explicit, like automobile ads that promise a 100,000 warranty, and some are implicit, like the promises made by anti-aging products or skin creams.

A good brand knows that when it makes a promise, it has to keep that promise, or the brand will lose its good reputation. When an airline says its flights are 99% on time, it had better be able to support that promise, or we will soon see the end of the brand’s primacy in the consumer’s mind.

The last way ads can work without data-driven targeting is through context: in other words, by being placed around content that aligns with the brand’s philosophy, the consumer’s interest, or the type of publication. The best example of this is New Yorker ads, which contain a preponderance of luxury items and items that appeal to intelligent, educated people. These are audience-based ads, and they are not designed for a specific customer, but rather for a specific audience.

We have long attempted to persuade the industry that all this data can cloud, rather than clarify, the goals of an advertising campaign, and we are happy to see the industry coming around to this. We can have a very healthy advertising industry without over-using consumer data and violating consumer privacy rights.

 

Header Bidding Wins

Just as publishers were beginning to see increased revenues from the 2016 hack called “header bidding,” Google decided to get into the game. For a while, header bidding was a way around Google’s almost total control of real time bidding auctions. But now Google has released its own solution called Exchange Bidding to all publishers who use Doubleclick for Publishers. AdExchanger writes,

Exchange Bidding relies on server-to-server connections, which are faster than the page tags some header bidding solutions rely on, and lower latency improves ad viewability, and thus yield.

It’s also faster to install, with Exchange Bidding essentially a plug-and-play product compared to heavy development and collaboration required of publishers in header-bidding integrations. 

Meredith, an Exchange Bidding beta partner, said it joined because its pages loaded faster and because it consolidated exchange reporting within Google, which reduced measurement discrepancies with advertisers, Chip Schenck, the media conglomerate’s VP of programmatic sales and strategy, told AdExchanger last year.

 

Exchange Bidding, however, can’t be used by buyers who have preferred or guaranteed rates. The new product is only available right now for display ads, although it is useful for mobile web and in-app ads. If you are a publisher who relies on video inventory, this is not for you.

We were one of the first to offer header bidding, and that was about two years ago. Before that, Google  had a near monopoly on the auction market. According to an article in last summer’s Ad Age,  header bidding really shook Google up, and it began offering its own product as a test.  But it immediately ran into resistance, because it was trying to charge a 5% cut for the service. That’s been reconsidered, but in the meantime both Amazon and Facebook have begun to offer their own form of header bidding, forcing Google to play fair.

We think this is good for the industry as a whole, because Google’s monopoly turned out to be costly for publishers. What’s also good for the industry is the convergence among all the once high-flying ad tech companies, who charged publishers for services that did not add value.

This call comes just in time, because eMarketer predicts that 83% of digital spend will transact programmatically by 2019, and by 2020 digital spend will exceed all other forms of advertising spend combined.

The next thing we have to do on behalf of publishers is step up the filtering of non-human traffic and traffic to fake sites, and that is increasingly becoming a priority, especially through the widespread industry adoption of ads.txt.

We’re getting increasingly skillful at weeding out the negatives in our industry and focusing on providing a clean marketplace for digital publishers and demand side players to interact.

 

 

 

 

 

Publishers with Strong Brands Increase Revenue

People will watch advertising in your app if you promise them something worthwhile. If the promise is important enough, they will “endure” ads to keep you in business, which is how it worked in the good old days of print and TV. However, if you break your promise, and lose trust, you have lost something from which it is nearly impossible to recover. That is how publishers should look at their sites — as brands that deliver on a specific promise strong enough that users will be willing to look at a few ads in exchange for what you deliver.

How do we know this? It’s been proven, at least on a small sample. This morning I was listening to a podcast celebrating the 20th anniversary of Kottke.org, a blogging site hosted and curated by Jason Kottke. It’s just a bunch of short written pieces, but one of them struck a chord with me. It’s about how  positioning changes the user experience.

There are three major mapping apps. Apple Maps, Google Maps, and Waze.  Of the three,  Waze promises fastest route times, although it often under delivers. Ironically, Apple Maps is most accurate in predicting its route times, and Google Maps actually gets you places the quickest. This was tested by a man named Artur Grabowski  in 120 trips over the course of 2017.

At the other extreme, Waze (Alphabet) makes money through ads when you use their app. What better way to get people to use your navigation app than by over-promising short trip times when no one takes the time to record data and realize that you under-deliver? If an unsuspecting user opens Apple Maps and sees a 34-minute route and compares that to 30-minutes in Waze, the deed is done. Now Waze has a life-long customer who doesn’t realize they’ve been hoodwinked and Waze can throw at them stupidly annoying ads.

Now I maintain that you can only position a product that under delivers for so long before users get the picture. I also maintain that you can only bargain with users to accept a few ads, not a page full of them.

It took us a long time, but the pendulum is swinging back to a focus on what the reader/viewer/visitor needs rather than what the advertiser needs. Advertisers will have to cooperate and stop demanding interstitials, popups, and other garbage that turns away visitors. They will also have to stop demanding data that they can use for retargeting. We are in different times now; we’ve tried a bunch of digital advertising techniques, and we know which ones fail.

As a publisher, your brand and your visitors are paramount. If your value proposition is strong enough to have lifelong relationships with customers, like the New Yorker, or Vogue, or the New York Times, you will eventually sort the financial issues out. But you can never do that if you sacrifice your brand to the needs of advertisers.

Whose Digital Identity Consortium will Win?

Companies in programmatic rely on their own cookies or device advertising IDs to anonymously identify audiences. Thus a consumer can receive many identifiers, resulting in data loss through fragmentation. This is how we got to the problem of a flood of data, yet decreased understanding of consumers, less effective targeting, frequency management, optimization, attribution, and a poor consumer experience.
As the digital advertising industry has matured and the criticisms have increased the ad tech portion of the ecosystem has been looking for a single identity solution. However, with the coming of GDPR and all the movement toward transparency and brand safety, this year ups the ante.
Digitrust was the first effort of the industry at a unified identity for consumers. In November, Rubicon wrote:

Cookies have ruled the ad industry since the early days of online browsing and shopping — helping media companies understand who their audiences were and what they were interested in, while making sure advertisers could deliver the most relevant messages. But the growth of programmatic and the mobile explosion have changed the industry dramatically, leading to a growing sense that cookies are no longer enough.

Across the industry, technology companies, advertisers, and publishers are all seeking an alternative to cookies — and there are a variety of solutions vying to be the go-to way we identify users and analyze their behavior moving forward. DigiTrust, the non-profit, independent ID consortium supported by companies like Rubicon Project, Dataxu and OpenX is one such option…

At the time, Rubicon decided to back Digitrust, but in the interim two other options have emerged.

The Trade Desk Cookie Sync

The Trade Desk has opened their cookie sync database to other organizations. The database is stored in their proprietary domain, and is in line with our existing implementation. However, this solution is only helpful to ZEDO if it is adopted by other players, and as yet there is no information on that.

Advertising Id Consortium
The third of these options is the Advertising ID Consortium, which takes a collaborative approach to solving these issues by providing: 
  • An open and standardized pool for cookie and device IDs.
  • The availability of people-based identifiers.
  • An omnichannel identity framework that adheres to privacy, security, and compliance requirements, industry standards, and best practices.

This is a solution similar to The Trade Desk’s, although headed by Liveramp and Appnexus. The problem with it is that everyone is trying to build Cross Domain Targeting Technology, and if this consortium takes off, Liveramp seizes the competitive advantage.

Competition among the major players may prevent any single solution from gaining enough traction to “win.” That would be painful for those of us who would then have to implement three different solutions.

 

Does Advertising Really Need so Much Tracking?

Our old friend Doc Searls is taking another step toward putting readers in charge of their data and taking on the use of too much consumer tracking. In a prototype edition of the reborn Linux Journal, he is going ask readers to indicate what they want in their advertising:

We believe the only cure is code that gives publishers ways to do exactly what readers want, which is not to bare their necks to adtech’s fangs every time they visit a website.

We’re doing that by reversing the way terms of use work. Instead of readers always agreeing to publishers’ terms, publishers will agree to readers’ terms. The first of these will say something like this:

That appeared on a whiteboard one day when we were talking about terms readers proffer to publishers. Let’s call it #DoNotByte. Like others of its kind, #DoNotByte will live at Customer Commons, which will do for personal terms what Creative Commons does for personal copyright.

Publishers and advertisers can both accept that term, because it’s exactly what advertising has always been in the offline world, as well as in the too-few parts of the online world where advertising sponsors publishers without getting too personal with readers.

Notice he is not anti-advertising as a business model. He is for restoring advertising to what it used to be — brand advertising. He refers to data-driven advertising as “junk mail.”

He theorizes that we’ve (publishers) lost a lot here by putting data collection in the driver’s seat:

By now you’re probably wondering how adtech has come to displace real advertising online. As I put it in “Separating Advertising’s Wheat and Chaff”, “Madison Avenue fell asleep, direct response marketing ate its brain, and it woke up as an alien replica of itself.” That happened because Madison Avenue, like the rest of big business, developed a big appetite for “big data”, starting in the late 2000s. (I unpack this history in my EOF column in the November 2015 issue of Linux Journal.)

Madison Avenue also forgot what brands are and how they actually work. After a decade-long trial by a jury that included approximately everybody on Earth with an internet connection, the verdict is in: after a $trillion or more has been spent on adtech, no new brand has been created by adtech; nor has the reputation of an existing brand been enhanced by adtech. Instead adtech does damage to a brand every time it places that brand’s ad next to fake news or on a crappy publisher’s website.

Yes, Doc is a friend of ZEDO.  But he is also a terrific writer, and you owe it to yourself as a publisher or a member of our ecosystem to read what he has to say in Linux Journal.

Building a Brand: For Publishers

All publishers will not survive the latest onslaught of Facebook changes and GDPR compliance. At least not with an advertising model. But should they? The combination of an almost limitless content  supply of sometimes questionable quality, the “Amazon effect” on brands, and the intolerance of consumers for slow-loading pages and interruptive ads will cause a Darwinian contraction among publishers.

The internet saw the rise of almost countless niche publications, each one fighting for ad dollars. That led to a proliferation of targeted ads, which in turn led to the need to collect personally identifiable information. The fact that 65% of companies weren’t ready to comply with GDPR shows how complicated the ecosystem has become.

That number includes brands and publishers. The brands have already taken steps. P&G cut its ad spend last year by $200 million, and said its reach was 10% greater. It will take more steps this year. In related news, WPP lost 15% since last year.

What will happen? Only the fittest will survive.

We think a combination of things will move the ecosystem forward, including the decline of publications who bet the farm on Facebook to get traffic, as Little Things did, the introduction of more transparency in the media buying system, and a diversified revenue stream will keep the best publications in business.

Mixed revenue streams have already begun to keep the largest publications, such as the New York Times and the Washington Post afloat. Some sites (The Information and Stratechery) do well by subscription alone, although the subscription model won’t scale across the entire industry because there are a limited number of sites to which any one human can subscribe.

That’s why after all is said and done, advertising will remain the best way for keeping content free, and those sites who design for the new ad formats recommended by the Coalition for Better Ads and IAB will see less competition for ad dollars and probably higher CPMS. The new site designs will be cleaner, pages will be faster loading, and desired content will be easier to read — all of which should have already happened.

The internet presented a temptation to put too many ads on too many sites, resulting in the digital equivalent of a swap meet and what we are seeing is a natural fallback of the market from excess to normalization. A market with too much merchandise is just as difficult to shop in as one with too little.

This is not very different from what happened to the cryptocurrency space lately as Bitcoin’s price fell from $19,000 to $11,000.

Bitcoin is a good analogy here, both because the regulators are coming both for digital media and for cryptocurrency, and because its price did not fall to nothing, just like advertising won’t go away. What we are seeing in advertising, as in blockchain, is the adoption of new technology inducing a hype cycle, and the market coming off the hype cycle into something more normal.

The quality publishers will still succeed, most supported by advertising. Publishers, just try to keep the flight to quality in mind. We’re here to help.

Lost Trust is Hard to Rebuild

It was the mother of all ad frauds. A group of Russians working with the Kremlin and desperate to have anyone elected but Hillary Clinton set up a pseudo- ad agency with a budget of $1.25 million a month and bought ads on social media platforms like Facebook and Instagram, measuring viewability, comments, and engagement. The ads were paid for with fake Paypal accounts. In addition to their advertising campaigns, the group ran a cross-channel marketing campaign for Trump, staging rallies and counter-rallies and paying people to participate.

And although Deputy Attorney General Rod Rosenstein made a point of saying the results of the election weren’t swayed by all the bots and ads and phony grass roots efforts, how can we be sure?

Indeed, there are more ways to sway an election than just stuffing a ballot box. Thinking the election wasn’t influenced ignores the power of advertising to create brands. These Russians could sway the election by organizing massive rallies and making Trump appear more popular than he really was. That would create a bandwagon effect — something ad campaigns try to achieve all time.

As a result of this highly successful campaign, over a hundred “unwitting” Americans  participated in the Russian government’s effort to interfere in our elections, and our major social platforms were besmirched as they were ringing up the ad revenues. Until recently, Mark Zuckerberg did not grok the extend of Facebook’s complicity.

There are lessons for the online advertising industry in all this. We have all been busy reaching for scale. But we haven’t put sufficient controls on the messages we are sending, and we haven’t devoted enough time or effort to combatting ad fraud. News outlets desperate for ad revenues in a changing market were willing to run ads that annoyed consumers. Brands bought ads in places they should never have appeared.

What will be the result of all this, beyond indicting 13 Russians who will probably never be extradited? We suspect it will be a massive loss of trust in digital platforms on the part of consumers. And with it, an equally massive pullback from social media advertising on the part of brands.

We were long overdue for this correction. It’s another consequence of mishandling customer trust, as the use of ad blockers is. Our industry could use a giant dose of quality.

In the consumer advertising business, just as in the public opinion business, lost trust is difficult to regain. We’d have been better off sacrificing scale to quality and using better targeting techniques, just as the social platforms would have been better off scrutinizing their advertisers more carefully..