Skip to main content
Deep Dive 28 Jun 2021 - 6 min read

Google's reversal risks 'nasty return of cookie bombing and garbage attribution' if marketers shelve first party plans; what now and what next for marketers and publishers?

By Paul McIntyre & Brendan Coyne
Cookies on an industrial scale

Feeding the machine a little longer: Google's backflip has consequences, though some may relish two more years of the cookie economy. Pic: iStock

Google’s decision to punt the cookie apocalypse two years down the line has major ramifications for the $330bn digital ad industry. Marketers must now return to boards and explain the sky is not yet falling in. Parts of the industry will at worst lapse back to bad habits, lazy strategies and “garbage” attribution models. At best there will likely be a deceleration of the first party data transition. But that would be a big mistake, say industry experts. Here’s what just happened, what to do about it, and what’s next.

What you need to know:

  • Google’s backflip will likely see a “nasty resurgence” of cookie bombing and extend the life of “garbage” digital attribution models, per consultant and former Westpac digital chief, Nick Barnett.
  • But he has a letter for marketers to copy and paste to their board.
  • Former Qantas Loyalty lead Vaughan Chandler says marketers taking their foot off the gas “would be a big mistake”. But he thinks some may.
  • News Corp’s Suzie Cardwell says the shift to first party data models has too much steam to be derailed – and brands that require control of their data, audiences and measurement will keep going.
  • Third party cookies’ life extension may have an impact on DMP/CDP strategies, says adviser Chris Brinkworth. But does everybody really need a CDP?
  • 10 Viacom CBS’ Josh Slighting thinks industry must focus on the bulk of digital advertisers if it wants to avoid being continually sidelined by big polatforms.
  • Frankfurt-based consultant Phil Eligio breaks down what really went wrong for Google’s grand plan – and what publishers should do next.

Digital attribution won’t break in 2022 or 2023. It’s already broken. The truth is that the models out there now are garbage and the process of preparing for third party cookies has shone a light on how misleading and inaccurate they are.

Nick Barnett, Independent Advisor

What just happened?

Google said its sudden change of heart, 18 months after announcing it would stop the $330bn global digital ad market from tracking online users via web browser code called "cookies", was because it needed to move at a “responsible pace”. Google-owned Chrome Privacy Engineering Director, Vinay Goel said it needed more time, among other things, to thwart fast-emerging “covert” tracking methods like fingerprinting. But the advertising giant has also faced widespread resistance to its proposed user tracking alternatives to cookies.

“While there’s considerable progress with this [Privacy Sandbox] initiative, it's become clear that more time is needed across the ecosystem to get this right,” he wrote, stating that Google also had a duty to protect publishers that provide free, ad-funded content.

Google has been beset by delays and challenges to its cookie alternatives. It has reportedly struggled to get approval for key browser changes from W3C, the international standards organisation for web browsers. Only four of its 30 so-called Privacy Sandbox initiatives were in live testing and even then at a tiny scale. The main runner, federated learning of cohorts, or FLoCs, has been blocked in Europe as incompatible with GDPR. Both the EU and the UK’s competition authorities had launched probes, with the Competition and Markets Authority asking Google to commit in writing how it would act – and aiming to work through those commitments in fine detail.

That put Google in a legal bind, but also offered it a way out. Perhaps useful, given the situation it had created. Without workable solutions in Europe – and for now the rest of the world – it risked losing a lot of money.

Marketers and publishers that were scrambling for post-cookie solutions have heaved a collective sigh of relief, alongside those that depend on third party tracking for their business models.

But there are now real risks that the shift towards a first party data economy stalls, and marketers and publishers revert to habits they had begun to kick.

Meanwhile, marketers that have broached boards for funds to re-engineer their approach to digital marketing, growth and customer data, must now return and tell management that the sky is not falling in after all.

But all is not lost.

 

Marketers: Copy this letter

Nick Barnett, Westpac’s former Director of Digital & Media Technology turned independent advisor, suggests marketers copy and paste the following and send it to their CMO or head of department:

Dear CMO and executives, thank you for your ongoing support of our initiatives to balance competitive marketing ROI with evolving privacy standards.

Over the past year, there have been countless examples in the industry that validated our decision to invest in tech, data and buying platforms, but none more than Apple's new marketing opt in rates, showing that as much as 90 per cent of consumers are rejecting companies tracking them with third party data.

Favourably, Google has recently announced that we, as part of the industry, now have extra time to continue these preparations.

As such, we would like to seek your endorsement for us to use this time to develop a marketing leading marketing measurement model that will ensure that we invest our shareholders precious capital into activities that have statistically quantifiable sales impacts.

 

Do you want to win at faking short-term artificial targets in this financial year or do you want a statistically credible measurement model that your board buys into?

Nick Barnett, Independent Advisor

Hold the line

In the meantime, Barnett urges marketers to hold the line – even if it is tempting to fall back into bad habits.

“You’ve just got to continue phasing out your use of third party cookies, particularly as we're going into the new financial year. Nothing has changed, we just have more time to prepare, so keep going,” says Barnett.

He says there are five key areas of focus for digital marketers:

“First is targeting and retargeting prospects [using first party data]. Second is personalisation. Third is suppression of your wasted spend – impressions that don't need to served, which might be pre-conversions, or existing customers. Fourth is analytics and fifth is risk and compliance,” says Barnett.

“Focus on those five key areas as your guiding principles. Apple’s IDFA opt-in rates have also shown us that anywhere between 70-90 per cent of your customers and prospects do not want you using third party data to follow them. So that should be a good enough reason to stop using it, regardless of what Google is doing. So stop now and don’t wait for 2023.”

Measurement cooked, cookie bombing returns

Barnett urges marketers to use the 18-month stay of execution to rebuild measurement.

If they do not, it heralds a return to the dark arts of cookie bombing and misattribution for the sake of arse covering.

“There’s one very clear reason that hyper-targeting and cookie bombing will probably make a nasty resurgence over the next year, and that’s because nobody is measuring properly,” says Barnett.

“The only way to do away with those shoddy tactics and the use of third party data is to upgrade measurement models away from garbage like last touch digital attribution.

“Digital attribution won’t break in 2022 or 2023. It’s already broken. The truth is that the models out there now are garbage and the process of preparing for third party cookies has shone a light on how misleading and inaccurate they are.”

Digital marketers, he says, must ask themselves one question:

“Do you want to win at faking short-term artificial targets in this financial year or do you want a statistically credible measurement model that your board buys into?

“As a marketer, I would be questioning all third party data sources and certainly any platforms using them unnecessarily.”

Most companies we've been talking to have really only mobilised out of necessity or fear rather than out of desire. So it’s quite likely some companies will take their foot off the pedal. But that would be a bad idea.

Vaughan Chandler, Co-founder, Crystal Box & former Qantas Loyalty Executive Manager

Two years lost?

But Vaughan Chandler, former Executive Manager at Qantas Loyalty and Qantas Red Planet and Co-founder of Crystal Box, thinks deadlines tend to concentrate minds – and Q4 2023 now seems a long way away.

Most companies we've been talking to have really only mobilised out of necessity or fear rather than out of desire,” he says. “So it’s quite likely some companies will take their foot off the pedal. But that would be a bad idea, particularly if it happens at scale.”

Chandler underlines the call to stay the course.

It takes time as a marketing organisation to understand how to build your own audience and then to engage your audience across a publisher network in a first party environment. Companies that are further down the learning curve will benefit from that,” he says. Meanwhile, they can take comfort in the “security blanket” of their existing approach – before it is taken away.

“More importantly, all the evidence suggests that working with first party audiences is a more valuable and certainly more robust way of marketing than relying on third party audiences.”

Retailers, he suggests, will likely continue to pursue first party data and consent strategies. Others that are further away from their customers, such as FMCG, “might find it harder to build up that understanding about how to proceed”.

To avoid losing two years’ worth of momentum, Chandler advises companies to try and reconfigure organisational structures so they can make faster decisions.

“This will be an area that continues to be dynamic for the next five years,” he warns, hence working towards agility and independence will pay off.

“Don’t rely on your consent [to use data] coming from third parties. Don’t rely on the measurement frameworks that the walled gardens provide for you – create your own measurement framework that works across them.”

The risk for marketers is being forced to use the exact same audiences that their competitors are targeting.

Suzie Cardwell, GM for Data and Ad Product, News Corp

Alternative IDs ‘not dead’

News Corp GM for Data and Ad Product, Suzie Cardwell, says Google’s reversal should not change brand and publisher strategies – and agrees now is the time to hone both first party data capability, matching and alternative identity approaches.

Otherwise, she warns, brands will face the same challenges when Google ultimately decides what it is going to do. FLoC is not dead yet, and if everyone is using the same data pools or segments, suggests Cardwell, they may struggle to cut through should a cohort approach eventually emerge.

“There are two major risks for marketers. The first one is in the difficulty in differentiating. FLoCs has been put on hold … But the concept was that Google was going to create cohorts – basically groups of people and devices that could be targeted based on the activity that that Google sees them undertaking around the web,” she says.

“Those cohorts were going to be standardised, effectively one-size-fits-all for marketers. There was going to be some ability to do some sort of micro-targeting within those cohorts. But effectively they were going to be pretty much the same for everyone,” adds Cardwell. “The risk for marketers is being forced to use the exact same audiences that their competitors are targeting.”

Any marketer that is breathing a sigh of relief probably does so at their peril.

Suzie Cardwell, GM for Data and Ad Product, News Corp

For the last 18 months, publishers have been evaluating and implementing a raft of alternative ID solutions – both email-based and device/browser-based – in a bid to ensure they can track and target audiences after the cookie cull, and match them with advertisers' own first party audiences.

While Google has repeatedly warned marketers and publishers away from cookie workarounds, Cardwell thinks that alternative ID providers are unlikely to see a crash in demand as a result of the extended deadline.

“We are still going to need multiple ID solutions, some probabilistic, some deterministic, some a mix of both. I don’t think that fundamentally changes,” she says.

Moreover, Cardwell thinks marketers that have invested in first party data will be reluctant to return to legacy measurement and attribution approaches “where the platforms mark their own homework”.

As such, she believes alternative IDs “will continue to evolve over the next 18 months … and will be a really important part in providing the solution to third party cookie alternatives".

Says Cardwell: “I encourage marketers who aren’t already looking at identity solutions to make a start. The move away from third party cookies has begun … The momentum behind those solutions is such that they are not suddenly going to go away – quite the opposite – and reliance on third party cookies will slowly decrease. So any marketer that is breathing a sigh of relief probably does so at their peril.”

 

The amount of enterprise businesses that I've spoken with in the past six months that have said ‘which CDP should we use?’… But they don't even need a CDP. Tech has evolved such that you can create something that looks like a CDP, without the investment.

Chris Brinkworth, Independent Adviser

What now for CDPs?

Since Google first said it would phase out third party cookies in Chrome, data management platforms (DMPs, which are largely third party cookie-based), have been on the wane. Providers have rushed en masse to become customer data platform (CDP) providers, which are largely first party data-based.

But Google’s reversal may have given DMPs a life extension. Nick Barnett thinks that may dampen demand for CDPs in the short term.

“Moving away from DMPs to a CDP is incredibly difficult, and if Google, the shining North Star, is making a two year delay to its plans, then everyone else is going to be struggling as well,” says Barnett.

Chris Brinkworth, an industry advisor on cross channel measurement who is now launching a data cleanroom specialist consultancy, thinks that might actually be a good thing.

The amount of enterprise businesses that I've spoken with in the past six months that have said ‘which CDP should we use?’… But they don't even need a CDP. The technology has evolved so much that you could easily create something that looks a bit like CDP, so you don't need to put that investment into it,” he suggests.

Before getting into any technology discussions, the fundamental question for marketing organisations is about consent, says Brinkworth. If the Attorney-General’s privacy review eventually ends up “looking like a watered down version of GDPR,” it may be that people have spent a lot of time and money collecting data, only to find they cannot use it, says Brinkworth.

He thinks marketers should use the current pause to nail gold standard data consent and governance.

“By really understanding what consent means and working from a privacy by design perspective, [brands] will reap the benefits in the future,” says Brinkworth.

Not every advertiser is going to be able to harness their own underlying assets and stitch them together and use them for targeting and measurement in publisher sites. We need to do a better job to help bridge that gap.

Josh Slighting, Head of Data & Digital Audience, 10 ViacomCBS

Solve for small business, win

For all the talk of tech stacks and shifting data sands, Josh Slighting, Head of Data & Digital Audience at 10 ViacomCBS, points out the likes of Google and Facebook have become whales by scooping up ad krill. He says the SMEs need help too – and expensive technology is probably not the solution.

“We’ve got to watch that we don’t fall in love with our own gospel. Some of these solutions are very good for enterprise clients with sophisticated tools, big teams, processes and large ad spend. They have the ability to chop and change, move from a DMP to a CDP, all of the fancy buzzwords.

"But the largest part of the ad spending market is not a sophisticated kind of enterprise. They are small to medium businesses, entrepreneurs – and that’s where the lion’s share of Google's revenue is coming from. So I don’t think the answer [to digital advertising’s future] lies in the technology,” he says. “It comes back to principles like moving away from some of the opaque practices, removing some of the barriers.

“If anything, the walls keep going up. But not every advertiser is going to be able to harness their own underlying assets and stitch them together and use them for targeting and measurement in publisher sites and we need to do a better job, I think, as an industry to help bridge that gap."

If the publishing industry cannot accommodate the bulk of advertisers by volume, then it may find itself continually on the back foot when the big platforms make unilateral decisions."We’re not necessarily making the solution-set accessible for the wider advertising market," says Slighting. "We really need to think about that.”

In the meantime, Slighting says 10 ViacomCBS is not particularly affected whatever Google decides to do.

“We're a premium video business for the most part, and the bulk of our audience doesn't live on a website.”

FLoC Origin Trial testing has really been a non-event ... The amount of audiences exposed to testing has been so low that test participants have suggested that the data they were getting back was practically meaningless.

Phillip Eligio, Independent Consultant

In the weeds: What happened, what next

Frankfurt-based consultant Phillip Eligio is deep in the weeds on Google’s moves and the countermeasures it faces. He tells Mi3 the cookie reversal comes as no surprise.

“Firstly, for those regular attendees of calls conducted under the auspices of W3C discussing key parts of the Privacy Sandbox such as FLoC and FLEDGE [another browser-level cookie replacement initiative], the progress was nowhere near where it would need to be to gain widespread industry adoption by the previously-announced 2022 timeline,” he says.

While Google has claimed FLoC could deliver 95 per cent performance versus third party cookie approaches, that appears something of a stretch. Early testing of FLoC “has really been a non-event,” says Eligio, particularly with the GDPR fail.

“What testing from some participants has shown us is that FLoC usage quickly trends towards just another form of user identification informed by cross-site activity.”

Another major problem is that the volume of audiences exposed to FLoC testing “has been so low that test participants have suggested that the data they were getting back was practically meaningless when thinking about how specific ad trading use cases would work at a larger scale,” adds Eligio.

While the CMA’s potential regulatory intervention may have played a role in Google’s decision to delay, “it is questionable that oversight from a UK body will be accepted as the global proxy for various key stakeholder requirements from the Privacy Sandbox”, says Eligio. “They won't be the last regulatory body to seek this oversight power.”

But that shouldn’t stop interested parties – i.e. everybody – from feeding into the CMA’s proposal, as the process will likely influence what Google’s Privacy Sandbox should and should not do – a discussion that will likely continue for the next 18-24 months, he adds.

Don't stop now

Eligio agrees that Google dragging out its timelines “does not mean that the industry should stop looking at ways to respect user choice and privacy in the way that it works. There are still data leaks being exploited by bad actors which should be addressed with urgency”, he says.

“For publishers, this is also the time to continue the momentum from the work conducted so far in this area. Publishers should continue with efforts to build a better future for digital publishing, work on becoming more sovereign from platforms and other unhealthy advertising industry dependencies, and move into a stronger position to provide a valuable service to audiences over the long term.”

What do you think?

Search Mi3 Articles