Guest blog by Perry Braun.
Adobe’s Digital Trends report revealed that 60% of senior executives say the loss of third-party cookies will disrupt their marketing. It’s worrying that the other 40% either have a solution in place or are not prepared for this game-changing shift. Who would have thought that in 1994, Netscape’s Lou Montulli (then 23 years old) would have facilitated billions of digital advertising dollars and shaped the world’s digital media tracking? The cookie is approaching 30 and like its tastier namesake, it’s crumbling before our very eyes. Frailing under the pressures of user privacy rights and data privacy rights and regulations. In the world of digital marketing and advertising, this is a time for concern and my observation is that too few brands are adequately preparing for this drastic change in targeting and reporting.
Some search engines such as Firefox, Safari and Edge have already been blocking or restricting third-party cookies. But when Google stops using them in 2023 the market tremors will be felt by the industry, brands and even boardrooms. When you consider Google’s SOV across PPC, programmatic and other digital touchpoints, it’s a significant audience size. Think about your marketing investments, how much of it runs through Google, overall, and in comparison to other search engines? Maybe it’s time you considered what these changes mean for you in 2023 and beyond.
Why have cookies been so important for so long?
To understand why cookies are so important it’s worth understanding what they are and what information is possible to utilise for advertising. Norton has a great summary – cookies are small files used by web servers to save browsing information, allowing websites to remember your device, browser preferences, and associated online activity. There are 3 main types of cookies
- Persistent cookies: Persistent cookies can save data for an extended period. These are the cookies that allow websites to store username and password information for users.
- Third-party cookies: Third-party cookies seek out data regarding your online activity to send back to website owners looking to improve advertisements.
- Session cookies: Session cookies delete immediately after closing your browser. These are best known for allowing you to keep items in a shopping cart even after clicking on a different page.
But what does this mean for brands?
Shane Paladin wrote an article for Forbes and said “Today’s customers crave these personalized interactions”, I don’t believe this personally. Every time I receive an advert I don’t crave it to highlight my name or just show me products I have registered an interest in. How will I ever be made aware of new brands, similar products, etc? I don’t crave personalisation, I crave relevancy. I care about who holds my data, how it’s used and for me to be able to correct and remove my data when it suits me so that I will be targeted by ads that matter and by brands that matter to me.
Enter Google Topics API. Google’s most recent alternative to 3rd party cookies. Essentially, practitioners will be able to target people based on these categories. There are currently 351 proposed topics and this number will fluctuate, but it’s an interesting approach. The interesting part is how Google determine where people sit across these topics. It is proposed that over a 3-week window the categories an individual sits within are based on their browsing history. However, I would be keen to understand if people will be able to opt-in and out of certain topics as the idea evolves.
Whether you have an in-house team of practitioners or an agency, they will not have access to data that is traditionally connected to a 3rd party cookie. Both in-house and agency practitioners are going to be reliant on 1st party data either within a CRM and DMP, the success of Google Topics API and other data considerations. My advice is to invest time in finding a solution that can combine some or all of the below:
- First-party cookie data
- Google Topics API – being able to report and attribute values to specific Topics.
- Individual event data such as opening a newsletter and being able to correlate test versions and attributes (e.g. A/B0 and subject line insights)
- Media scheduling such as when TVCs or Radio are airing
- Semantic and natural language processing such as when a company is mentioned in a podcast or news (completely possible and you should check out Alembic for more info)
- click stream data. When someone engages with a digital touchpoint such as the website or advert.
To have all of this in one place feels like the holy grail so being able to achieve some connectivity across some of these areas is a good place to start.
Having a common location that can stitch together these events will be able to provide the best possible attribution model as it will include cross-device and location insights. However, you will have to rely on how values are attributed to each part of the journey. For reporting consistency this is essential. Interestingly, I believe that when a solution can deliver this type of connectivity it will then unlock the potential of predictive modelling. However, with most learning processing (agencies and suppliers will promote this as AI) it will take time, and investment, something that might be acceptable to larger clients but for SMEs, this is a luxury they might not be able to afford.
What does this mean for your agencies?
According to recent IAB research, publishers could lose up to $10 billion in ad revenue with the crumbling of cookies. Google reports that publishers face a loss of up to 50-70% of their revenue. One thing is for sure, their revenue is at risk. And that means agency ‘kickbacks’ are also at risk.
Rhys Chow-Seegoolam Head of Strategy and Delivery at Push Group highlights that “Whilst there are still a lot of unknowns about the exact impact that cookieless advertising will have, a lot of assumptions can be made especially when considering the impact we’ve already seen from the iOS 14 update”. He places a particular focus on the critical importance of testing. “Whether that (testing) is experimenting with different ad channels or ad content to name just a couple, the biggest success will be seen from advertisers and agencies that can be fast and agile in this ever-changing environment”.
Kushtrim Jeta Head of Digital Operations at Hearts & Science highlights that the risks for brands stretch beyond media and are preparing his clients for the impact on attribution. “Google, along with other media partners, is preparing advertisers for not only a cookieless future but a post-last-click future, where conversions are defined by a data-driven approach which looks at user interaction and predictive models to determine if a conversion can be attributed back. Advertisers will need to truly embrace data-marketing hybrid models, those who operate these two departments in silos would struggle to meet the marketing demands of the future.”
If your agency proposes its own programmatic solution then be sure to understand how this fits into the impact of a cookieless world. Importantly, ask them now how they are preparing for it.
If your agency proposes a 3rd party advertiser or provider be sure to understand how they are providing value and how their performance data is monitored (ad serving etc) and how their reporting fits into your new cookieless reporting.
Testing and adoption of new hybrid attribution models are certainly going to play a huge role. I predict that there will be an influx of ad agencies offering attribution modelling (the words AI will no doubt be in there somewhere). Mark my words.
An Advertisers perspective
After speaking with Simon Whitehead, Senior Sales Director for Africa at Teads, he said “There is still considerable value in cookies, mobile-based inventory from Teads will use a Device ID (MAID, GAID, IDFA). In the absence of either (and where integrations exist with publishers), Teads will transact on Logged In identity. Where none of these exists Teads will transact on a probabilistic ID generated from user agent information, like device type, browser, operating system or content consumption”. He went on to explain that “Teads also creates targetable audiences via its Teads Interest Graph. This is created via appending user content consumption information to a probabilistic ID. Users’ content consumption over time is scored inside a particular content topic area and those the over index is placed into a targetable audience corresponding to that area e.g. Fashion Interest”.
Time will tell how advertisers like Teads incorporate or correlate their consumption information style information with Google’s Topic API. Those advertisers that are not reliant on 3rd party cookies to target audiences and offer brands and agencies significant audience size and insights will prosper when the cookie crumbles. Simon continued, “Teads also develops cookie-less demographic segments based upon these probabilistic IDs. These audiences are created based on user-agent information and content consumption. They are passed through Nielsen’s DAR study for validation of both gender and age range and are found by Nielsen to be 30% more accurate than their study norms for On Target Audience accuracy.”
He went on to add:
Teads utilise semantic page analysis from over (circa) 2 billion users. Going far beyond keyword inclusion and focusing specifically on page categorisation, topics, and semantic analysis at the article and sentence level. This leads to significantly higher levels of accuracy in determining the true context of a page rather than just hitting on specific keywords.
My biggest takeaway from speaking with Simon is that agencies might not have access to these audiences and insights, making advertisers a more important part of investment decisions. Agency-owned DSPs might have to take a hit on revenues as they build for a world without 3rd party cookies.
What to do next?
Preparation is key and Gartner identifies three core considerations:
- Prepare for sustained disruption
- Rethink ad measurement practices
- Adapt to a walled garden world
More specifically, I would strongly suggest client-side budget holders and reporting teams focus on managing internal expectations and forecasts for future performance. Make leadership and key departments (sales, product, merchandisers etc.) aware that change is coming and YoY, MoM even WoW metrics will be impacted. I would also explore some of the tasks below to help prepare for change.
- Identify all the places you use third-party cookies.
- Update data privacy where applicable.
- If you haven’t cleaned your data and considered long-term value exchange I would get your thinking caps on and make your dev/ data teams your new best friends.
- Consider consumer transparency and demand for editing and deleting data.
- Understand what this means for cross-device tracking and alignment.
- Consumer trust and loyalty are going to be critical. Consider how to build trust and where possible loyalty programmes to enable you to stay relevant.
- Have a strategy and preference capabilities to offer opt-down options rather than the blanket opt-out.
- Keep up to date with the developments around Google Topics API and understand how this works within the way you currently buy digital media from Google.
- Step up your efforts to reassure consumers and target audiences that their data is safe.
Trends and projections
Digital advertisers will need to transform or introduce data brokerage-style solutions. Wealth of Geeks describes it as big companies scrambling to get their hands on a valuable asset (consumer data and preferences). However, the risk here is that brands will need to maintain where customer data has been acquired.
Brands will significantly ramp up their efforts to engage and offer content in exchange for up-to-date information (B2B) or promotions (B2C). Resulting in:
- A stronger focus and investment in content production and diversity. Production quality and content types (video, podcasts, downloadable materials etc) will have to improve.
- Content diversity will mean more brands providing hyperspecific content and personalised information. There are examples today where this is effective, for example, Go Compare and car insurance. That is a value exchange, that enables consumers to share their information for targeted offers for products. But how will this be applied for specific brands, and what can brands learn from brands such as Direct Line that don’t appear on comparison sites? How do they tackle these issues today, and how can this be rolled out at the brand level across all sectors and audience types?
- Social audiences will play a more significant role in creating awareness and attracting new audiences. Keep them close and treat them well! Their authority and volume will help fill some of the voids that cookies leave behind.
- Marketers will have to be more creative in how they reach new audiences. However, also be prepared for some more traditional advertising options such as traditional media buys and buying content and data from trusted third parties, resulting in a higher cost of CPMs.