L'apocalypse des cookies

The future of ecosystems and digital platforms according to Google

sgelinas@adviso.ca (Stéphanie Gélinas) — Thu, 22 Dec 2022 12:20:24 GMT

The guest of the aperitif: Jean-Philippe Gauthier. JP has been in the Quebec digital landscape for over 20 years. He is currently Head of Platforms & Digital Marketing Transformation at Google. He also ran the Mediative advertising network, worked at Bell (Sympatico and MSN), LaPresse and Streamtheworld. Let's say he has quite a few behind the tie!

Why the performance indicators you use will become unreliable

abailliache@adviso.ca (Aurélie Bailliache) — Thu, 22 Dec 2022 12:17:49 GMT

The challenge of the moment, with the apocalypse of cookies , is to fully understand the repercussions of such a transformation in the digital world. Today, we are going to take a closer look at its consequences on the measurement of data, what we call analytics . Get ready, you will have to question several performance indicators that you observe on a daily basis to measure your effectiveness!

We will therefore focus mainly on the repercussions of the cookie apocalypse on primary cookies. Unlike third-party cookies, which tend to disappear, primary cookies will remain relevant, but will experience major changes:

the lifetime of primary cookies will be limited to 7 days (or even 24 hours in some cases) on the Safari browser;
Cookies on Firefox will be deleted daily unless there has been user activity within 45 days.

CHANGES IN YOUR DAILY LIFE

What does this limited lifespan imply? In fact, this means that it will be difficult to recognize a user from one visit to another, which will have repercussions on two elements:

multi-channel attribution of your consumers' journey across all of your digital marketing efforts and e-commerce activities in general;
direct (or 1:1) measurement of overall performance on your Google or Adobe analytics platforms (i.e. post-click conversions).

Let's take an example: Isabelle arrives on the site www.monsite.com from a Facebook ad, browses a bit, then leaves the site. She returns there nine days later in organic search to finally make a purchase there.

That's what happens in real life, but what about data collection?

Day 1: A user 123 is identified, a visit originating from a Facebook campaign is recorded.
Day 9: A new user 456 is identified, a visit from organic search and a conversion is recorded.

The following data will therefore be counted in Google Analytics, depending on whether Isabelle uses the Google Chrome or Safari browser.

Safari having its limits regarding the lifetime of primary cookies, this has an influence on the following indicators:

Number of unique visitors
Number of new visitors
Number of Assisted Conversions
Distribution of attributed direct conversions by channel.
Acquisition cost

You get the idea? Now ask yourself these questions when looking at the data:

Is the volume of new visitors real? You may not have acquired as many new users as you think.
Are the conversions attributed by channel representative of reality? A declining conversion rate does not necessarily indicate poorer performance!
Are your CPA and your ROAS representative of reality if you only observe them from the Google Analytics or Google Ads platform? On paper, costs do increase when looking at direct ( last click ) conversions and assisted conversions, but direct last click conversions alone remain the same.

These are reflections to integrate into your daily life before sharing any data related to the performance of your marketing activities. The cookie apocalypse is also affecting all content optimization and personalization activities. Your thoughts on A/B test results should also take this into account. You can also learn more by reading the article written by my colleague Nicolas Duval.

THE FUTURE OF PERFORMANCE MEASUREMENT

What to do in the short term?

In your reports, consider the margin of error of your analyzes based on the proportion of your traffic that comes from Safari and Firefox.
Avoid cross-referencing or comparing primary clicks or conversions data in Google or Adobe Analytics with that of your Adtech platforms, which depend on a third party. The margin of error will only increase.
In Google or Adobe Analytics, consider the margin of error in customer journey analysis, based on the percentage of traffic coming from Safari and Firefox.
Note that if you add UTM parameters in your campaigns, the lifetime of primary cookies in Safari will be 24 hours instead of 7 days.
Exclude the Safari browser from your testing experiences (A/B, MVT).

In the longer term, a lot of things are changing and falling into place. The watchword for 2021: adapt! Whether in terms of platforms, processes or people, we will have to be flexible to deal with this upheaval.

PERFORMANCE ANALYSIS WITH GA4, BIGQUERY AND MMM

Google is developing a platform called GA4. With GA4, Google democratizes the analysis of large datasets with the integration of BigQuery. Companies will be able to export their GA4 data to BigQuery for free to perform advanced analyses, or even enrich their CRM data before pushing it to other platforms. To learn more about 4 good reasons to activate GA4 in 2021, it's here ! As far as we know, the new version will increasingly use machine learning to compensate for data gaps caused by the possible disappearance of cookies and IDFAs.

Another trend seen in the industry is the return to statistical inference models ( The Drum & Parker , 2020). More specifically, it is the return of one of the methods most used by marketers before the digital era: marketing mix modeling (MMM). Instead of analyzing the performance of digital media campaigns in silos, it's about analyzing the impact of all marketing activities on sales, both online and offline. As the organization matures from the point of view of data use, it is also possible to add other internal and external factors to have an even more complete view (weather, price indices competition, etc.).

DOWNLOAD OUR PREPARATORY GUIDE TO THE COOKIE APOCALYPSE

WHERE TO START TO PROPERLY MEASURE PERFORMANCE IN 2021?

To navigate this changing industry and enable reliable and robust modeling, it is necessary to properly integrate the tagging platforms to capture a maximum of data points that feed the models of the different platforms.

In order to limit the loss of data captured by your media or analytics platforms on the client side (on the web page), "new" (or old 😉 ) methods for integrating pixels and other tracers are coming back into fashion: we are talking here about server side tracking. Google Tag Manager now allows its server-side integration to enable new tracking methods that are more robust than those on the client side (Google, 2021). Please note that server-side tracing does not give you the right to override legal regulations in terms of managing user data. This type of integration will however prove necessary for a good number of organizations, both from the point of view of measurement and audience creation, because it will make it possible to play on the window of the limited lifespan. for primary cookies.

Finally, we can also maximize the collection of primary data by identifying users who have an online account. User data platforms (CDP) will be more and more interesting to exploit. It might therefore be wise to put efforts in place to build a solid user database.

TRAINING AT THE CENTER OF THE HUMAN

Who says new platforms, new technologies and new processes also says training. Indeed, it will be essential to invest in the training of the people who are at the heart of your marketing activities in order to be able to continue to move forward with confidence in your next initiatives.

These people (you!) will first have to:

fully understand the challenges of the cookie apocalypse to ensure that you control the use of current data;
adapt as of today the actions which are posed on the marketing activities;
think about the medium and long term strategies that could be put in place within the company – GA4, BigQuery, CDP, MMM and so on.

SO WHAT DO WE DO?

Perfect solutions do not exist, or at least not yet. On the other hand, it is important, above all, to understand the changes that affect measurement and to adapt our data analysis according to the new rules of the game. We must therefore constantly adapt and continue to show ingenuity to optimize our marketing tactics.

We are currently experiencing great upheavals. This is a golden opportunity to get ahead and create a competitive advantage. Like what, even in the storm, it is possible to stand out.

Visit our special section dedicated to the Cookie Apocalypse to learn more about it

Beyond cookies: Understanding the new marketing measurement paradigm

rkamena@adviso.ca (Roger Kamena) — Tue, 31 May 2022 03:31:00 GMT

Despite the fact that it was myself and my team at Adviso who were among the first to popularize the term “cookie apocalypse” as recently as four years ago, today I’d like to be among the first to say we need to stop using it.

Why? Because things have changed and this expression no longer reflects the extent of the changes emerging within the marketing ecosystem. Today we should instead refer to a paradigm shift in marketing measurement—a transformation that has been advancing very rapidly for the past two years, and whose final phase will result in the disappearance of third-party cookies in 2023.

Let me explain further.

Basically, the disappearance of third-party cookies in 2023 is just one piece of a much larger puzzle in the world of marketing and online advertising. As Meta (Facebook) sadly learned too late in 2021, when its stock fell by 45 percent, it’s the entire idea of the measurement and tracking of web users that is being disrupted, not just the use of cookies.

In an article I co-wrote with my thesis advisor, Daniel Lemire, in 2018 on The Conversation, I explained that the significant changes being made that affected the flow of data were not being done solely to protect user privacy. In the background, an all-out data war is going on within GAFAM, which is now witnessing the arrival of none other than Elon Musk, with his offer to buy Twitter.

Despite all of that, and to bring things back to the specific context of data related to online advertising, we should remember that cookies aren’t the main player when it comes to identifying users online. Since 2019, over half of advertising budgets have been devoted to mobile applications, with a dwindling portion given over to advertising using cookies in a web browser (Chrome, Safari, etc.).

As a result, if you’re only talking about the end of third-party cookies in 2023, then you’re only talking about 40 percent of the investment in digital advertising. In this view, you could say that the apocalypse in user tracking for ad targeting already happened after the launch of Apple’s ATT (App Tracking Transparency) protocol in iOS 14.5, which now requires mobile apps to ask for user authorization before tracking their activity on other applications and websites.

Furthermore, Google just killed its own mobile identifier on Android and the Google Play network, which was called GAID (Google Ad ID, the counterpart to Apple’s IDFA, which has been obsolete for a long time). The progressive deployment of the Google Play service affects all applications following the advent of Android 12 in 2021, and extends to all devices supporting Google Play as of April 1, 2022.

Application developers now have to use app set ID to retain their capacity for analytics measurement or even to detect fraud. But when it comes to advertising, the paradigm is changing completely. Google is launching four tools or APIs to replace cookies and mobile identifiers with an SDK, specifically:

Topics API, a new method of targeting that respects privacy and was proposed by Google for third-party cookies. This is an audience targeting system that attributes bundles of interests to users based on the websites they’ve visited in the past, except that all of this happens within a sandbox environment, even within the browser, so that no data is sent to advertising servers or third-party platforms.
Fledge API, which will become the new marketing standard. Fledge will make it impossible to track specific targeted users. The system will be based on an auction model even from within the browser in order to display the most relevant ad without sharing user data to a third-party server.
Attribution Reporting API from Chrome, which will facilitate the continued measurement of when a click or view on an ad leads to a conversion, such as a purchase on the advertiser’s site. However, it was created to not allow third-party platforms to use it to track users on a website. In short, this API is a way of continuing the generation of attribution reports without revealing all of a user’s details and their activity.
Runtime SDK is in some sense Google’s version of the ATT protocol, except it’s less restrictive. Based on the application sandbox idea, the aim of this system is to restrict access and sharing undisclosed data from a user’s application by third-party SDKs. It exerts control over access to data and to and to specific APIs.

Unsurprisingly, this paradigm shift corresponds to another change at Google, which will terminate the current version of Google Analytics by 2023 in favour of Google Analytics 4. The complete re-engineering of the platform seems to synchronize perfectly with the end of third-party cookies and GAID. Coincidence? Maybe.

But one thing is for sure: There’s no coincidence that GA4 is equipped with a significant capacity to reconcile user data through machine learning in order to make up for gaps in data caused by all these changes. Furthermore, it’s partly thanks to the unification of mobile and web data provided by GA4 that this is made possible.

With this in mind, if you haven’t yet started your migration to GA4, don’t lose even one more minute!

In concrete terms, how will this affect the everyday reality of marketers?

It means that the third-party cookie apocalypse is actually just one of the final stages of a host of paradigm changes in marketing measurement that have been happening for the past two years. In other words, the apocalypse won’t be starting in 2023, rather it will just be going through its terminal phase.

Unfortunately, many marketers still don’t realize that even today, on the back end, all those measurement systems related to marketing analytics are starting to be phased out. As of this year we need to start taking into account all the growing inconsistencies in our analytics tools, even with first-party cookies.

For example, since 2017, in the post-ITP (Intelligent Tracking Protocol) world of Safari, it’s becoming increasingly difficult to reconcile user data coming from an Apple browser. These inconsistent data will scramble the cleaner data from other sources of traffic coming to your website or application. More specifically, the effects are as follows:

Distorted increases in certain metrics:
- Number of unique users
- % of new users
Reductions in some others:
- Number of visits by user
- Conversion rate
- Visit frequency
User journey is harder to identify
Hard to recognize users if there’s no log-in aspect

Even without having a specific percentage at this time, it’s already possible for our Analytics and Data Science team to confirm that in every year since 2017, there’s been a significant deterioration in all the measurement activities listed above.

The other changes deployed by Apple and Google in 2021 and 2022 will only accentuate the disappearance of the former marketing measurement paradigm. In 2023, we’ll witness its final stage, or perhaps I should call it the first stage of an entirely new marketing paradigm.

What will be the new marketing measurement paradigm in 2023?

It can be encapsulated in just one sentence: Prioritize more seriously the collection and exploitation of primary data.

As Suzanne Vranica explains in the Wall Street Journal, shrewd advertisers have been taking the paradigm shift very seriously for the past two or three years. That’s why these companies are in the midst of launching a whole new range of initiatives involving the collection of primary data about their customers.

At a tactical level, the nerve centre of primary data collection involves obtaining email addresses by consent. As reported in eMarketer, nothing is out of bounds when it comes to convincing consumers to share this important piece of information. These companies are developing aggressive primary data acquisition strategies while putting the accent on value exchange as a motivator. According to a report by Fluent in June of 2021 called “Consumers, Data, and Control: Driving Loyalty and Trust Through a Strong Value Exchange,” here is a list of growth tactics for incentivizing the exchange of personal information:

A special discount in exchange for their email address
An exclusive offer
A contest
A free gift
Exclusive content

But this won’t be enough to collect a ton of email addresses. You first need a well-considered data strategy paired with solid infrastructure for managing these data while ensuring respect for privacy. In this sense, the major primary data management initiatives are:

The development of a data strategy and governance plan
The creation of a marketing data lake or warehouse to combine data from CRM, email, sales system, online analytics (Google or Adobe) and media (Google Ads, Facebook Ads, etc.) platforms
The development of models or the acquisition of tools to unify data and construct a holistic view of the customer
The integration of secondary data from reputable partners in the industry, such as Vividata, Nielsen and Comscore
The implementation of customer data platforms (CDP) and infrastructure
The integration of these platforms on the server side to exploit the one-day window and limit data loss
The development of mixed marketing models with aggregated primary data
The development of machine learning models specifically adapted to the company’s primary data

The question you should be asking yourself as a marketer is the following: Out of all these initiatives, how many of them are currently active or in development within your team or department?

If you’re part of a company that still hasn’t developed a road map for acquiring and managing primary data, what are you going to do when January 1, 2023, comes along?

Google delays the end of third-party cookies to 2023: What's the impact?

rkamena@adviso.ca (Roger Kamena) — Tue, 06 Jul 2021 11:23:00 GMT

GOOGLE CHROME PUSHES THE END OF THIRD-PARTY COOKIES TO 2023

The web "gods" at Google seem to have heard the cries of distraught marketers when they announced they were ending support for third-party cookies on Chrome from January 2022. The giant has announced that it will postpone the rollout of the Sandbox project to 2023 .

The reasons mentioned are generally technical and logistical, but it is also a show of compassion towards web publishers and ad tech companies who will be greatly affected by this enormous change. In his statement, Vinay Goel, Director of Privacy Engineering at Chrome, explains the decision:

“…we have to move forward at a responsible pace. This will allow ample time for public debate on solutions, continued engagement with regulators, and migration of publisher and advertiser services. This is important to avoid compromising the business models of many web publishers that support freely available content. »

Another factor is, most likely, the investigation launched by the Competition and Markets Authority (CMA) in the United Kingdom, which aims to determine whether the blocking of third-party cookies on Chrome could prove to be a lever of unfair competition. Under an agreement with Google, the CMA will oversee the upcoming transition to ensure ethical practices are in place. It should also be noted that the United States Department of Justice and the Directorate General for Competition of the European Commission are also closely monitoring the progress of the case. This means that Google will have no choice but to comply with the rules of the game.

Regardless of the main reason behind the postponement, this news is encouraging for marketers who clearly also rely heavily on third-party cookies for live delivery and measurement of their audience targeting tactics.

SHOULD MARKETERS SLOW DOWN THEIR 'POST THIRD-PARTY COOKIE' PLANS?

Ideally, no. On the contrary, having the opportunity to save some time should motivate you to stay the course towards your “post third-party cookie” marketing roadmap. If Google is offering an extended transition period, it is because the American company sees the importance of the changes brought about by the cookie apocalypse and understands the magnitude of the management challenge facing industry players, a storm through which it will not always be easy to navigate.

Indeed, the announced change will require a considerable effort of adaptation on the part of publishers, marketers and engineers of intermediate ad tech platforms. Therefore, an additional year of preparation to avoid going through an overly chaotic transition appears less like a luxury than a necessity.

As an example, consider Facebook's migration to its Conversions API. There are three important levels of change to consider, or the three big “Ps” of IT:

(Platform) Manage the technical implementation : You have to consider the time and the technical issues related to this transition from measurement by pixel to measurement by API, on the side of the developers and integrators on whom your team relies.
(Process) Manage operationalization : Next, consider the time needed to adapt to this new way of measuring once it has been properly implemented. It is also necessary to plan the issues of reconciling previous reports by pixels to those carried out by API.
(People) Manage the training of marketers : Once the API is implemented and the operationalization processes are well defined, we must train the people in our teams to understand and implement these processes and this new way of doing things as soon as possible . It will take time.

In the example above, we are only talking about the Facebook Conversions API . Now imagine going through the same process to adapt to Google's FLoC and FLEDGE model, while getting used to the new Google Analytics 4 interface . Add to that the countless meetings your team will have to hold with each of your vendors and digital marketing ad publishers. They will all want to meet with you to explain how they will adapt their technological solution to a world without third-party cookies.

It's starting to be a lot to manage, isn't it?

So instead of seeing Google's announced postponement to 2023 as a reason to slow down your plans to mitigate the effects of the announced apocalypse, see it instead as an opportunity to get some extra time to prepare for what's to come. is, in any case, inevitable.

So, take the opportunity to get ahead. Google is pushing back its deadline to help everyone adapt to this major change. Don't waste that precious time by spending the coming year procrastinating. There is a lot on the board, and this, for each of us!

Facebook Conversions API: Accelerate your server-side data collection

jcmauss@adviso.ca (Jean-Christophe Mauss) — Tue, 01 Jun 2021 00:00:00 GMT

For a few years now, consumers have become increasingly aware that they generate data when they’re online. In fact, according to a recent survey, close to 97% of the population is concerned with protecting their personal data. For this reason, many big digital players have established measures that enable users to better control the data they share online. Changes such as Apple’s creation of the App Tracking Transparency framework or the removal of third-party cookies announced by Google will have major effects on companies whose marketing activities currently rely on the analysis and activation of data. It goes without saying that we’re about to experience the biggest shakeup in our industry in years.

In order to be able to adapt to this evolving context, Facebook strongly recommends that advertisers integrate its conversions API (Conversions API or CAPI for those already familiar with it).

What’s so great about Conversions API (CAPI)?

Formerly known as Facebook Server-Side, Conversions API’s main purpose is to send information that’s impossible to get with a pixel (such as in-store sales or calls) in order to bring performance reports and optimization to a whole new level.

The basic pixel gathers information relating to activities on the site and sends them to Facebook through a third-party cookie. Just to remind you, a third-party cookie allows the sharing of information between two different domain names.

Given the coming disappearance of third-party cookies, CAPI will allow you to continue sending information to Facebook through server-side data transfer, thereby getting around client-side limitations.

The idea is to collect a maximum of primary data in order to increase the chances of matching them with a user. For example, the only primary cookie data limits the ability of the social media giant to recognize a user. On the other hand, if you configure your server integration by adding an IP address, an email address or external identifier, for example, Facebook can now adequately identify the user and use this data to ensure the proper functioning of its advertising activities.

In a recent webinar for its partners, Facebook predicted that the optimization of conversions with a web pixel would no longer be possible by the end of 2021. Advertisers that don’t migrate to CAPI will therefore be forced to optimize based on lower value actions, such as clicks. This type of optimization could prompt an average increase of 150% in your costs per acquisition. Trust us, you don’t want to go there! 😉

Other than avoiding the interruption of digital media activities, using Facebook’s conversions API enables advertisers to take advantage of many benefits, including:

better visibility on conversions through the capacity to send additional information such as CRM data, in-store sales and qualified leads (MQL/SQL);
better control of shared data given the possibility of adding quantitative and qualitative data such as product margins or customer lifetime value; and
more robust integration, since CAPI is less likely to be affected by site crashes or lost connections, for example.

Maximize the volume of primary data sent

Primary (first-party) data form the core of CAPI’s operations. Generally speaking, CAPI needs at least one user attribute for transmission to Facebook. To explain it in a simplified way, when CAPI sends anonymized user data to Facebook, the social media giant tries to find at least one occurrence of this information in its internal database. When such an occurrence is found, FB can assign an action, since it recognizes this user within its member base. You must therefore gather as much information as possible on this user in order to maximize the matching potential with Facebook.

When we’re talking about user data, we’re referring to the following attributes:

Application with purchase example

Now let’s put all this into practice! In the example below, we attempt to track a purchase without third-party cookies using CAPI. If we send information relating to the primary cookie, email address and personal identifier (external ID), the matching potential with a user is higher. Note that the primary cookie is set through a web pixel, which is why Facebook recommends using a pixel in tandem with CAPI. The duplication of the event will happen subsequently in the platform.

Facebook Example of information collected and sent using CAPI and the Facebook match:

Let’s go more in-depth with a concrete example.

Example 1: Sending an anonymized email

A user lands on your website and wants to make a purchase. To do so, one of the conditions required is that the user enters their email address into a confirmation form. Since the purchase is considered a conversion, CAPI records it as such and sends a request to the Facebook server. The request contains the anonymized email address for the user (in accordance with all the rules surrounding user consent).

Once Facebook receives the data, it tries to find this email address amongst its billions of profiles in its database. If it succeeds, the conversion is attributed to the identified user.

Example 2: Sending a primary cookie

When a Facebook pixel is installed on your site, it creates a primary cookie to identify visitors. This cookie is stored in an fbp variable under the domain name of your website. In general the content of this cookie takes the following form:

fb.1.1596403881668.1116446470

Why use a cookie with CAPI? This indicator is placed on every page of your site and can be sent as a result of every action. When additional information is sent on a user, such as their email address, Facebook can then recognize that the previous actions recorded by this primary cookie are attributable to the same user.

The information transmitted between your web server and Facebook resembles the text file below, with the fbp variable clearly identified in the “user data” section:

Example 3: Sending a personal identifier

Sometimes you need to send specific information that is related to your organization to an anonymized client number, the primary cookie for another platform or even a loyalty program identifier. Facebook gives you the option to indicate the desired type of attribute through a personalized variable called external ID. The recommendation is to send the information with other identifiers in order to more efficiently construct a profile of your users. For example, here is a message containing an anonymized external ID sent to Facebook by CAPI.

The limits of CAPI

While CAPI is an effective tool that we strongly recommend using, it’s not without its limitations, three of the main ones being:

It’s a lot more complicated to integrate than the web pixel and requires particular skills if you want to add it to your platform without the help of a development partner.
In many cases, CAPI comes with fees, as opposed to the web pixel. Its integration necessarily involves infrastructure costs since it’s set up on the server side (as its name indicates) and you have to pay to access a host server. If you go with Google’s App Engine, that will cost you around $50 per server on a monthly basis. If you follow the advice of the team at Google and get three in order to limit data losses in case of a server crash, the bill increases to around $150 a month. Otherwise, you could decide to go with services offered by an integration partner such as Zapier or Segment. WordPress also offers integrations with PixelYourSite.
It’s a technological solution for an industry problem. It does NOT give you the authorization to ignore consumers’ desire to not be tracked if that’s their preference.

Procedure for integrating Conversions API (CAPI)

Now that you have a better understanding of how CAPI works, here’s what you need to get started:

a Facebook pixel on your website (although it’s not a requirement, it will improve the quality of the events transmitted);
a Facebook Business Manager account (something every company should have!); and
an access token, which you can find through the FB event manager (recommended) or by using your own application.

Once the token is generated, the event manager will give you access to a new tab called “Implementation.” Click on this tab and follow the instructions on screen. Facebook will guide you through the steps for creating a CAPI application and a system user for the conversions API.

Ready, set… go!

Go to the event manager and select Add events using the conversions API.

You will be asked to select the events you would like to track.

You must then choose the advanced parameters (custom parameters) you would like to send to Facebook as well as the information parameters about the customer. The more information parameters you can send, the greater the likelihood of a match with a user. Before forwarding them to a developer, ensure that all the events you want to send are actually available on your website. This goes without saying, since obviously you can’t send an email address if the customer isn’t providing one.

Once you have had your configuration reviewed, you will be asked if you want to do the integration yourself or if you need help from a developer. If you’ve already generated your access token, you can skip to the next section, which covers event configuration with a payload (data sent through the http web protocol).

You can build your payload with all the information shared beforehand by a colleague (or with yourself!)

Once the payload is created, you need to host it server-side, which can be accomplished with Google Tag Manager (section Server-Side tagging), for example.

Note that the conversions API can be integrated directly on your server by a development team or even with the support of integration partners such as a tag manager (GTM Server-Side), a CDP (Segment, Tealium), an integration system (Zapier), etc.

In all of the above situations, the team at Adviso can guide you through the integration process for a next-generation tracking system.

The end of cookies: 3 specific solutions (video)

jfrenaud@adviso.ca (Jean-François Renaud) — Tue, 01 Jun 2021 00:00:00 GMT

Why Marketers Need to Return to Media Mix Models

rkamena@adviso.ca (Roger Kamena) — Mon, 05 Apr 2021 11:19:00 GMT

In a recent interview with AdExchanger , Chetna Bindra, Product Manager for User Trust, Privacy and Transparency at Google, said that as the cookie apocalypse of Chrome announced for 2023, marketers need to work harder. According to Bindra, "[Google] will not build an alternative solution to continue to collect data about individuals as they browse the Internet." Also according to Bindra, this also means not using Chrome users' login information to feed Google's advertising platforms.

If so, the trend towards a complete blocking of individual user data is set to continue – and it's likely to only intensify over time! In such a context, marketers will have fewer tools at their disposal to observe user behavior and link it directly to the performance of their campaigns.

For now, Google offers alternative audience building solutions through proposals like FLoC and FLEDGE, which promise a depersonalized approach to segmentation and targeting based on user cohorts. But these solutions will no doubt evolve over time, too. Another technique to mitigate the consequences of the data apocalypse is the use of server-side data collection tools, through tag management solutions. However, who knows how long such approaches will remain viable. In a few Safari or Chrome updates, they may be completely deprecated.

In this context, the wisest will avoid putting all their eggs in one basket. To tread carefully in this new environment, marketers have two options: increase their ability to collect direct data, or rethink how they think about performance measurement. In this regard, we are seeing a strong comeback of media mix models.

This approach has been around for decades and is used to measure the overall influence of channels such as TV, radio and outdoor display (OOH) on sales. Essentially, it's about using the predictive power of statistics to determine which channels to focus on to increase business results.

With the advent of digital marketing, companies have moved away from this approach to more direct tracking methods. For each impression or click, it is possible to observe the user's behavior to know exactly how many conversions can be attributed to the different media placements.

Naturally, this method is entirely based on cookies, more specifically third-party cookies. With their disappearance, this method of analysis specific to digital marketing will become less and less precise.

Further, the scope of this attribution system is limited by the limited share of digital conversions that are directly associated with specific clicks (and, sometimes, impressions). In this context, the majority of conversions will be attributed to placements at the smaller end of the funnel, such as remarketing or paid search.

By definition, such a method therefore ignores the bulk of your brand marketing efforts – the widest part of the funnel – such as videos or TV, podcasts and radio, or sponsored content. Yet these are the placements that often take up the largest share of your media budget. So how can marketers demonstrate the value of these less targeted initiatives?

Media Mix Models (MMMs) are a great way to demonstrate the value of your marketing strategy by reporting results that reflect all of your efforts. In short, rather than focusing on measuring the conversion rate for each click collected, you should broaden the scope of your analysis to measure the impact of your campaigns on the overall performance of your business.

Rather than limiting the analysis to direct attribution from specific conversions to certain channels in the campaign, MMMs look at all marketing efforts and analyze the overall impact in terms of sales, opportunities or transactions. 'business.

In doing so, MMMs help identify the channels or combinations of channels (hence the name media mix ) that will have the greatest influence on your overall performance.

In the field of statistics, we use to apply this technique models called multiple regression models. However, with the advancement of technologies in the data sphere, new approaches are emerging, including some advanced machine learning techniques . It should also not be forgotten that marketing experts today have a much greater amount of data to produce an effective media mix.

In an ideal world, creating a data warehouse or data center is the best way to collect data from a variety of sources that will build robust models. But it is not absolutely necessary to build a data warehouse to initiate an MMM approach. Good results can be obtained simply by collecting data from the reports of your advertising platforms, your CRM and your sales reports.

Additionally, advertising or digital marketing data may be combined with external data such as search engine trends, Comscore reports, or economic statistics.

If you want to continue to demonstrate the positive impact of your marketing initiatives despite the disappearance of cookies, media mix models are a proven scientific method.

Visit our special section dedicated to the Cookie Apocalypse to learn more about it

The death of the third-party cookie and its impact on content and customer experience

nduval@adviso.ca (Nicolas Duval) — Thu, 25 Feb 2021 12:22:00 GMT

Nothing is constant except change. – Buddha

The eventual removal of third-party cookies, scheduled for 2023, is expected to have significant implications for a host of digital marketing products and solutions. Targeting, measurement and attribution will be transformed by these changes to varying degrees. Some alternative solutions are being developed, but it is difficult to imagine a world where digital media will not be transformed by the pro-privacy initiatives claimed by Apple and Google. But what about content and customer experience platforms and tools? Reminder: The impact of the end of cookies on your paid media initiatives

EXPECTED IMPACTS ON THE CUSTOMER EXPERIENCE

The majority of platforms used as part of your content and customer experience strategy should not be directly affected by the possible removal of third-party cookies. The impact of these transformations will be felt more in other channels that could influence and increase the importance of your primary data acquisition strategies. Here we offer an overview of the changes to be expected on your web assets.

DOWNLOAD OUR PREPARATORY GUIDE TO THE COOKIE APOCALYPSE

CRM / EMAIL

Expected impacts: None to low Post-cookie importance: High

Customer communication tools have the advantage of being a primary source of data — since your customer databases belong to you, the removal of third-party cookies theoretically has no influence on your email strategy (other than value it). A CRM like HubSpot will not be affected. “ HubSpot's cookies are first-party cookies, so they're not impacted by the Chrome updates. Your HubSpot tracking code will continue to function, and your HubSpot reports will continue to contain data. Source On the other hand, some CRMs offer a component for enriching your audiences or your data, such as Salesforces Audience Studio.

The cookie apocalypse could therefore have significant repercussions on these types of solutions, but it should still be remembered that they are generally not 100% dependent on third-party cookies.

“ Audience Studio has prepared for this change and can collect many more types of IDs next to cookies: hashed emails from your customers and other customer IDs, MAID and Ad IDs that all mobile devices have and is used in apps […]. A key argument for why the impact on Audience Studio will be limited is that currently the share of digital ad spend using cookies is only around 25%. – Source _

CUSTOMER ACCOUNTS AND LOYALTY PROGRAMS

Expected impacts: None to low Post-cookie importance: Very high

It's a safe bet that an advantageous loyalty program will remain one of the best sources of primary data. It is still necessary to have an interesting offer in return for the information of your users. “ Publishers will need to create identity solutions. […] Users behaviour, preferences and activity can then be tracked as publishers will now be able to track the user journey and predict their likely actions.

The data tracked using the identity solution can then be stored into a Data Management Platform (DMP), which can then be leveraged as first-party data for campaign targeting or cross promotion. – IAB SEA+India Data & Analytics Committee

CONTENT STRATEGY

Expected impacts: Low Post-cookie importance: High

Content is an important gateway to most sites and a great opportunity to build or refine audiences based on your users' interests. A well-oiled content strategy can therefore offer many advantages if used in synergy with a relevant loyalty program.

“ The next best option to cookies based behavioral targeting is anything keyword or keyword contextual-based advertising. Years ago everyone discounted it and we moved further and further away from keyword targeting, but now we're going to have to go straight back to it. – Jon Kagan, VP of Search at Cogniscient Media

Note that creating interactive or downloadable content is a good way to build contextualized audiences.

AB TESTING AND PERSONALIZATION

COOKIES AND A/B TESTING

Expected impact: Low Post-cookie importance: Unchanged In the majority of cases, optimization and A/B testing platforms use first-party cookies and therefore will not experience the impact of removing third-party cookies. “ Optimize's JavaScript snippet uses cookies to store some data in visitors' browsers. In most contexts these are first-party cookies. – Optimizely Documentation

The most advanced features will be the ones that will feel the brunt of these changes the most. For example, a dynamic personalization tool connected to a third-party data targeting platform will no longer be able to function as is.

APPLE AND A/B TESTING

Expected repercussions: Medium to high Recent changes from Apple are hurting. Indeed, with ITP 2.3, Apple limits user tracking to 7 days.

“ So, with ITP, a visitor coming back more than seven days after they initially triggered an experiment runs the risk of seeing a different variation. A/B testing in these conditions is useless and does not produce reliable results. – Lauréline Saux A conversion test usually needs to be run over a long period of time to be meaningful, and all of a sudden targeting and measurement are no longer reliable on Apple devices. Adjustments with respect to targeting and interpretation of data are therefore necessary.

HEATMAP TOOL

Expected impact : Low Post-cookie importance: Unchanged Heatmap tools (eg Hotjar or Crazy Egg) also use primary cookies. If these should not experience repercussions following Google's changes, the measurement components could be transformed in the same way as the performance analysis tools.

IN SHORT

The impact of the cookie apocalypse on your owned channels should be relatively small, but the impact on your paid media and data could create significant upheavals. A healthy collection of data is therefore to be preferred to better position yourself against your competition. Do you think you are well positioned? Is it time to rebalance your efforts?

Visit our special section dedicated to the Cookie Apocalypse to learn more about it

The future of advertisers on Facebook

jfrenaud@adviso.ca (Jean-François Renaud) — Thu, 25 Feb 2021 12:21:00 GMT

For the aperitif this week, we welcome Jean-Hugues Roy, former journalist at Radio-Canada and professor at UQAM, to discuss the recent tremors that have hit the web giant Facebook and the potential repercussions on the decisions of advertisers.

4 Big Reasons to Activate Google Analytics 4 Before 2022

rkamena@adviso.ca (Roger Kamena) — Mon, 01 Feb 2021 05:08:00 GMT

As mentioned by Jean-Philippe Gauthier (Head of Platforms and Digital Marketing Transformation at Google) during his interview with our co-founder Jean-François Renaud, 2021 is a transitional year for Google Analytics. It’s the year users of the current standard version will be asked to transition to the Google Analytics 4 version.

This will be a major change and probably one of the most important in the history of Google Analytics. Given this, and as we explained in our preceding article on the subject, we recommend that you not uninstall your current version, but instead activate GA4 in parallel, the reason being that this new iteration is still under development and therefore not quite on equal terms with the old version.

In a sense, this parallel activation could be thought of as what DJs call a wordplay transition technique. In simple terms, for a successful wordplay transition, a DJ must take a word or line of a song and use it in a successful mix into the next track.

Source image link

In the context of Google Analytics, it’s about using the current configuration of your program as a reference point for the next implementation. This way it will be possible to compare the two instances and make adjustments while the old version is still online.

What are the incentives for implementing GA4 now?

For some, it’s tempting to say, “Why not wait until 2022 to implement GA4?” If that’s you, then here are four reasons to not be one of the procrastinators and instead get cracking as soon as possible.

#1 Better preparation for the cookie apocalypse

GA4 is not a solution that can directly mitigate the cookie apocalypse. However, the platform was created to make up for the measurement differences and disparities caused by the disappearance of third-party cookies and IDFA. According to Russell Ketchum, Group Product Manager at Google Analytics, the new version will use more and more machine learning to fill in the data gaps caused by the loss of cookies and IDFA. This is in part possible due to the unification of mobile and web data offered by GA4.

# 2 Enriched media audience creation

The unification of mobile and web data brings several advantages, including the possibility of creating audiences that are richer, more granular and more relevant for media targeting. Thanks to GA4, there are opportunities to develop an audience strategy that is not just more advanced, but multi-screen and multi-device. In an economic period when every marketing dollar counts and audience targeting is the lifeblood of media targeting, these functionalities shouldn’t be taken lightly. What’s more, based on all the audience creation options available with the new signal technology developed in GA4, the future looks very promising for marketers thirsty for more advanced segmentation.

#3 More in-depth marketing analytics and e-commerce reports

GA4’s approach in terms of performance reports has also be significantly overhauled. The new reports are more in-depth and real-time measurement is greatly improved. Also, the unification of webpage and mobile application data is a major change. This new paradigm will enable us to have an improved, unified understanding of the consumer within one single analytics platform. With the current version, you have to separate the analytics from the web and from your mobile application. This is an outdated approach in a world in which users interact with your brand on several devices and through various media (web + app). Furthermore, for those who haven’t read my article on the subject, remember that all raw data from GA4 can be exported free of charge to BigQuery. This option opens up immense opportunities for in-depth analytics on consumer behaviour, on both your website as well as your mobile app, and especially for the overlap of behaviour between the two.

# 4 A much more granular view of the customer funnel

One of the most interesting functionalities of GA4 is the way in which funnel analyses have been reconsidered and remade. With the current version, funnel analysis is very limited and allows for very few personalized reports. In version 4, these reports are more like those of GA 360, the paid version. This means it’s possible to tailor-make a customer journey funnel that’s a lot more specific to the realities of your business. For those who, like me, are familiar with huge licensing and implementation fees associated with platforms specializing in customer journey analytics (such as C3 Metrics), this new functionality is very generous on the part of Google Analytics. But of course, there is probably a reason behind it. It’s likely that Google’s objective in providing this gift is to evolve clients into the paid version, or even to encourage them to invest more in their advertising platforms, given the integration possibilities between the two. On the other hand, for digital marketers that are already invested in the Google Ads platform, this means you will have improved abilities in advertising messages, targeting, segmentation, personalization and customer experience.

Screenshot of the conversion funnel in GA4

One precaution to consider before going any further

GA4 is without doubt the near future of analytics at Google. Waiting for 2022 before making the transition would mean a major delay in your performance measurement. However, GA4 is not just in beta mode, with all the shortcomings that implies, but the platform was created based on a completely new model for implementation and tagging.

That means it’s important to familiarize yourself with the technical subtleties of this new paradigm before making GA4 your main analytics platform. Your first step should be to create a solid transition plan and integrate it into your analytic road map.

Start with an audit and diagnostic of your current implementation in order to create an implementation strategy for the new platform.

If you have any questions or need help with either your audit or the creation of a parallel transition strategy between the standard version of GA and GA4, please don’t hesitate to get in touch with our analytics and data science team.

Visit our special section dedicated to the Cookie Apocalypse to learn more about it