IAB Europe Guide To Contextual Advertising July 2021
IAB Europe Guide To Contextual Advertising July 2021
IAB Europe Guide To Contextual Advertising July 2021
Summary Page 31
Contributors Page 32
2
Section 1. Introduction
From the beginning, contextual advertising has been a key tool for brands
looking to reach relevant and engaged audiences. Indeed, print publishers were
the first to use “environment” to sell the promise of reaching audiences, who
would likely care about their message or product as they were reading relevant
content at the time. With the emergence of online publications, contextual
advertising via the web began to leverage search terms, with websites adding
keywords to their pages to open up more relevant placements. Eventually,
keyword engines integrated directly with publishers’ properties, enabling them
to sell pages that matched such keywords via their ad servers. Context offers a
strong proxy for, and also compliments, audience interest and the development
of contextual targeting enables advertisers to identify specific audiences at
scale.
A number of factors has made finding ‘The right person, in the right
environment’ increasingly challenging in recent times. And with this trend likely
to accelerate, there will be increased reliance on alternate solutions, such as
contextual, as a tool to drive user engagement and campaign performance.
3
The aim of this guide is to provide further education on the alternatives to third-
party cookies and to help planners and buyers of media navigate contextual
advertising. This guide offers a definitive round-up of what contextual
advertising is, and the opportunities it provides in Europe, in a post-third party
cookie world. It also offers some key considerations and best practices to ensure
it is used efficiently and optimised for success.
4
Section 2. Definitions
What is Contextual Targeting and Advertising?
In its most simplest form it’s about identifying the most appropriate and relevant
content against which to display an ad. For example, an ad for running shoes
could be placed on a news article about an upcoming marathon, or an ad for
laptops on a tech ecommerce site. The idea being that it improves the
advertising experience, as the user will only be delivered an ad that is relevant in
context of the content they are reading or viewing.
This concept of contextual targeting has been fundamental to the marketing mix
for well over 15 years but has seen a resurgence more recently due to a
tightening on data and privacy regulations. Contextual Advertising, as such, has
evolved and now extends beyond just content and keywords on a page. It can
now include and is not limited to text, image, audio, location and semantics of
content.
5
Section 3. Why Context is Important
The contextual adjacencies of digital advertising is now a hot topic among
marketers looking for a highly effective and privacy-friendly way to target their
campaigns.
As consumers increasingly spend more time and money on the internet, they
are becoming ever more privacy conscious and data aware, with data privacy
remaining a key concern to 94% of them according to the IAS Context Matters
study. Coupled with the enactment of stricter data privacy regulations such as
GDPR (General Data Protection Regulation) and CCPA (California Consumer
Privacy Act), and the decision to sunset the use of third-party cookies and to
restrict in-app trackers, there is a universal shift in the ecosystem to incorporate
solutions that leverage a privacy-first approach and enable a balance between
consumers’ privacy concerns and advertisers’ need for relevance and precision.
Advertisers, as such, are not only increasingly looking for targeting strategies
that they can rely on regardless of new privacy regulations and regardless of
updates by media companies around the use of user-IDs, but that also enhance
user experience. As such, advertisers are amplifying the use of contextual
targeting and shifting budgets towards this solution with $412B projected to be
spent on contextual advertising by 2025. Contextual solutions provide the scale
advertisers need without the use of hashed emails, user opt-in or any cookie-like
mechanism, making it even more valuable in the current landscape.
6
Further fuelling this growth is the evolution from simple keyword targeting to
the use of more sophisticated, machine-learning technologies that provide a
holistic understanding of the content. This in-depth analysis allows brands to
reach audiences in a more meaningful way to create more relevant consumer
experiences.
That is the promise and the power of contextual targeting. Rather than trying to
build the right audiences based solely on third-party data, contextual targeting
helps find audiences based on the content they are viewing at that moment, and
then matches brand-suitable advertising for that specific environment.
7
Section 4. The Contextual Opportunity in a Post Third-
Party Cookie World
2020 kicked off with one of the most fundamental changes to the online
advertising industry, when Google announced its goal of making third-party
cookies obsolete by 2022. Where marketers had once relied on cookies for
targeting, re-targeting, display advertising, and behavioural marketing in general,
they now needed to change tact. And even though this has now been postponed
until late 2023, the industry is still very much re-thinking its strategies towards a
cookie-less era.
Without the use of cookies, marketers need new ways to optimise campaign
performance, and to ensure the efficiency and effectiveness of their ad spend.
In today’s privacy-centric ecosystem, contextual targeting is gaining even more
popularity – it's privacy-safe and it works. A DV/Sapio study overwhelmingly
shows users are receptive to ads that are relevant to the content they’re viewing,
with a significant percentage stating that they’ve actually tried a new brand
product as a result of seeing a contextually relevant promotion.
8
Additionally, when we consider that context is defined by much more than just
content (i.e. mindset, moment, screen size, setting, whether someone is co-
viewing, etc.), we can identify many new opportunities to generate meaningful
impact with campaigns.
When we look at CTV as an example, we can already identify moments that will
have a big influence on the impact you will generate with your ads. For instance,
if you are watching a 90-minute Hollywood movie with your partner on the
couch at 9PM and you see an advert for a last-minute holiday to Ibiza, you might
just be in the right mindset to start discussing your next vacation to a sunny
island. However, if that same ad would be shown to you at 8AM in the morning
while you are tuned in to the latest business news on CTV, it will surely not have
the same impact and it might even frustrate you at that specific moment.
9
4.1 CTV and Contextual Advertising
With emerging solutions like CTV not being reliant on cookies, in this new post
third-party cookie era, contextual targeting will play a bigger role. In the context
of CTV planning, contextual targeting is even more important if you consider co-
viewing. Research by SpotX revealed that 76% of CTV viewers watch together
with at least one other person in the room. If we assume for a minute that
cookies exist, in this scenario, we might classify this user as a female between
35-44, interested in fashion, fitness/sports and beauty and in the ad breaks, ‘she’
would see ads for online shopping, sportswear, fragrances, etc. Now if we look
beyond the ‘cookie’ and understand the ads were shown on the latest
Pixar/Disney movie at 11AM on Saturday morning, we might have opted to
instead show ads for mid-size SUVs, family holidays, etc.
Targeting ‘users’ without taking context into account will, in many cases, be just
as likely to negatively impact your core KPIs and will harm the user experience at
the same time. When you can start to think about ways to connect with
consumers in the moment and look at the relevant contextual signals we do
have readily available, we will not only make those moments more relevant but
we will also be able to provide a more meaningful customer experience.
10
A publisher understands their readers, users, viewers, and listeners and the
content they are consuming on a daily, weekly, and monthly basis. Everything
publishers do within digital media ultimately has the end goal of influencing a
particular audience – contextual is no exception. By bringing a publisher's
insight into contextual targeting you can ensure the environment matches the
audience. Publishers can also help offer additional unique insight through
bespoke audience tools that identify user’s behaviour and what that audience is
currently interested in. Incorporating this real-time insight into a contextual
campaign ensures the perfect blend between audience and context is achieved.
Some measurement companies can also work with advertisers to convert their
first-party data into cookie-free contextual signals, to ensure they can continue
to target at scale in a privacy-centric manner. This is done using a crosswalk
between first-party audience data and privacy-forward contextual signals.
11
Section 5. A Step-By-Step Guide to Planning Contextual
Campaigns
For those looking to incorporate the contextual opportunity with success, follow
the crucial steps below:
To plan for contextual campaigns, advertisers first need to understand the goal
of the campaign and the target audience to define the plan and strategy they
will execute. Also, it is essential to comprehend what contextual targeting is, as
well as gaining more insight into the publisher’s inventory, the SSP providers,
and testing the scale when applying these targeting capabilities on the DSP’s
side. Then buyers can begin to apply contextual elements that allow reach into
their desired audience.
• Once the contextual provider has been selected, the first step will be to select
the contextual segments that most closely align to the campaign goal and
target by using contextual categories that align with the IAB. Also to go a step
further to target cookie-free audiences. Brands should look to include as
many positively targeted segments as possible to ensure scale and then refine
the list from there.
• Know when to layer data or not on the DSP side. Advertisers need to
understand that contextual targeting is not based on behavioural targeting.
However, advertisers must know when it is possible or not to layer data over
publishers’ first-party data and third-party data.
12
• Keep in mind the duration, flight dates, and budget campaign. When
advertisers plan they need to manage the reach, scale, and CPM expectations
when applying contextual targeting campaigns.
• For example, if advertisers want to target women between 25-45 years old,
interested in business and technology in Greater London, they have the
possibility to target this type of audience hitting ZIP codes from the Greater
London Area. Plus music stations popular between this genre and age, such
as Hit Radio genre, and business and tech formats, and applying age and
genre targeting.
• For instance, keeping the last example, if advertisers want to target women
between 25-45 years old, interested in business and technology in Greater
London. They have the possibility to target this type of audience, hitting ZIP
codes from the Greater London Area. Moreover, they want to target women
only using smartphones during the morning and evening when they are out of
the home. Therefore, they can also target the mobile device when the listener
uses the mobile phone data connection type or is in specific geo lat/longs
during certain time slots in the morning and evening.
13
Also, it’s good practice to adopt the ad creativity to the content language, the
time of day, or even to the content sentiment, to be relevant to the audience.
Advertisers must identify how the environment differs when users consume
different media types such as audio, display, OTT, video, etc.
• Next layer in brand safety, brand suitability, and negative keyword lists,
targeting based on the needs of the brand.
• Advertisers must identify what inventory sources can provide the interests
and audience segments that match the brief's requirements. In this stage, it’s
required to get granular information about publishers’ inventory related to
content (such as type of media, content genres, website and app domains,
sites, online radio, and podcast show names). Also, the first-party data
richness that publishers can provide.
4. Validate and test the reach before starting the contextual targeting
campaign
• Check the reach from the DSP side to ensure the expectations about the scale
match the inventory that advertisers can bid on.
14
• If the reach is low, make the necessary adjustments according to all interests
and segments related to the audience targeted. Also, advertisers must
double-check all the information related to the inventory with the inventory
providers.
15
Section 6. How to Buy and Apply Contextual Advertising
In a direct buy campaign, publishers will need to apply the contextual targeting
on their end to form a bundle of inventory and contextual segments. Publishers
who follow IAB’s contextual signalling Audience and Data Transparency Standard
can use self-defined audiences.
Publishers that have a Data Management Platform (DMP) and have cooperation
with contextual data providers might also be able to receive additional
contextual segments in the DMP, on top of the data they already have.
Publishers also often have bespoke formats that can be contextually targeted to
offer the advertiser a high impact and total page ownership solution. Often
these formats are not available in a biddable environment, meaning the only
execution option available is via programmatic guarantee.
16
Private Marketplaces (PMP)
Private marketplaces are a great opportunity once you have identified those key
partners that are determined as being important to your plan.
However, it is crucial that you don’t lose sight of your Brand Suitability goals and
ensure that the quality of the media being passed to you via these curated
marketplaces or PMPs.
First request that your supply partners pass fully transparent URLs within the
PMP/ Marketplace so your contextual partners are able to verify the brand
suitability of those URLs. Just like OMP buying, it’s recommended that you apply
your pre-bid contextual avoidance segments against your PMP to ensure you are
paying for quality impressions that meet your brand suitability requirements.
On the other hand, you can centralise seller relationships, generate, or ask for
customised deal IDs. Also, you can execute campaigns through audience
segmentation data, brand safety, audience privacy preservation, all while
improving flexibility and transparency in the media buying process. This process
creates a buying method incorporating and mixing publishers’ first-party data
PMPs in each curated marketplace. Then it allows reaching the desired audience
into a single package.
For example, you can ask or search for curated deal IDs from marketplaces. Also,
you can ask directly for information such as sites, apps, online radio stations, or
CTVs, to double-check brand safety, inventory with GDPR compliance, forecast
segments scales, and reporting.
It’s essential to identify what SSPs make the audience data transactions easier by
generating PMPs for a specific, repetitive, or tailor-made use to activate
campaigns based on contextual targeting.
17
Curated Marketplaces
In this type of marketplace, publishers with first-party data and data providers
can upload their type of content, semantic context, and segments. Then, the
inventory can be packaged or combined to make it available to you at scale. In
addition, this process creates a selling method to make it easier for you to
access a unique inventory via curated deal IDs.
For instance, publishers’ inventory is plugged into the SSP, the content is filtered
by IAB categories such as sports, healthcare, or beauty genres. Then, you can
apply publishers' first-party data to build vertical custom audiences. Lastly, this
inventory can be shared through PMPs.
18
Section 7. Campaign Best Practices
As the digital advertising ecosystem shifts, 2021 is a key moment for the
industry to embrace alternate solutions such as contextual advertising. Brand
safety and suitability, targeting and optimisation should be prioritised to ensure
success. Best practices to consider include:
Advertisers should:
• Look to first use brand safety categories and suitability thresholds rather than
relying on extensive negative keyword lists that need to be continuously
updated during a campaign based on the ever-evolving new cycle.
• Remove the keyword exclusion list from your DSP and from your preferred
vendor and rely on contextual segments in their place. It’s also imperative that
you apply this approach at a post- and pre-bid level in order to ensure brand
suitability whilst minimising wasted budget on non-working media.
19
Once your brand safety and suitability strategy is in place and you have
identified the appropriate contextual segments that meet your vertical topical
and brand specific needs it is crucial to consider its application in relation to the
type of programmatic buying you do. For example: applying segments in the
open marketplace or private marketplace (programmatic guaranteed, etc.)
You should look to test contextual solutions that offer cookie-free audience
segments powered by behavioural insights in preparation for the cookieless
future. These types of cookie-free audiences can offer a way to reach specific
audiences outside of endemic pages, providing more scale.
Cookie retargeting campaigns perform well, but they often have limited scale.
Contextual targeting is a complementary solution that can help add scale by
targeting audiences that are likely to convert due to high-relevance between the
content and your brand.
20
Optimise based on results
Pay attention to content where you are seeing conversions and work with your
contextual vendor to add similar types of content to your targeting strategy.
3. Creative Messaging
Identify the correct contextual strategy that aligns with your consumer profile. It
is also imperative that you work with those contextual providers that have
flexibility in their solutions and have pre-bid integrations with all the major DSPs
in order to maximise and scale the opportunity.
4. Campaign Measurement
21
Contextual advertising insights will also play an important role in helping you to
measure and improve outcomes by looking at the impact of different content on
user engagement and ad exposure, and then using these insights to optimise
campaign performance.
22
Section 8. Technical Standards for Contextual
Advertising
Contextual targeting exists through programmatic because, like any other
information, it’s part of OpenRTB. OpenRTB is a protocol that enables tech
platforms (SSPs and DSPs) to communicate in real time, when bidding on digital
media.
Below is the data available from the SSP to the DSP via the bid request in
OpenRTB: :
• Domain: In OpenRTB you have the domain (root domain) but in our case the
most interesting field is the page, which provides the full URL visited by the
end-user. Most of the time, the URL is enriched with keywords and can be
used to identify the content of the web page.
o Content: The supply can inform the buyer about the language of the
content in the bid request (language in OpenRTB).
o Creative targeting: Via an other field (WLAN in OpenRTB), the supply
can ask the buyer to send a creative fitting the languages requested.
• Deal ID: More and more tech platforms build Curated Marketplaces to allow
buyers to reach performance, address challenges and provide new revenue
outcomes for the publishers. The only way to pilot and access these
opportunities is through Deal ID.
• Category: The content category of the website or the app is described in this
field (cat in OpenRTB). It can contain one or many categories. In order to be
on the same page between the SSP and DSP when describing content, there is
a content taxonomy, which is useful for contextual targeting and even brand
safety.
23
• Regulations: The supply must determine if the bid request is subject to a
specific regulation (e.g. GDPR) by detecting where the user is located.
Regarding the contextual targeting, this field (regs in the bid request) can be
used in a different way by not targeting a user based on the geo but to
identify whether or not the content of the web page or the app is appropriate
to receive certain creatives (e.g. Children’s Online Privacy Protection Act
(“COPPA”) can be applied here and indicated to the buyer that the audience of
the supply is under 13).
• Content: While the previous definitions talk about identifying the content of a
website or an app overall, the field named content enables it to be more
accurate by providing information about the content in which the impression
will appear - the ad slot itself. With this field you can identify the title of the
video, provide the name of the series. In the case of audio, you can provide
the name of the album etc.
For example, within the bid request from the supply side, the buyer is notified
that the website toto.com is identified with the category “International News”
and the video ad slot is playing a video identified with the category “World
Soccer”. At the end, the buyer can refine the campaign.
A DSP can also send some data to a SSP (via the bid response in OpenRTB):
• Adomain: Enables the buyer to inform the supply side about the advertiser
(domain of the advertiser) behind the creative.
• Language: The buyer can share the language of the creative explicitly.
Even if the OpenRTB protocol provides fields and data in order to feed
contextual targeting, the programmatic supply chain needs to be aligned in
many ways:
• As we’re talking about billions of requests per day, you should be able to
ingest, analyse and leverage the data.
24
• Even if the SSP supports the data flow, the integration with the publishers
needs to support it.
• Each tech platform is different. For example a connection between SSP A and
DSP A is not identical to the same SSP A connected to another DSP B
(different OpenRTB version, compliant or not with data, etc.)
With all the available data, and signals coming from the publishers, the
platforms have never stopped innovating in order to create refined ad
placements to fulfil buyers needs. Below is an overview of contextual targeting:
• Text and media analysis: More and more platforms innovate to understand
the text with the pictures and videos included. Computer science is used to
get the right keywords and/or categories for the buyers behind.
OTT, CTV and audio inventories move in this direction to extract signals from
the media played (video or audio) in order to increase the efficiency of ads.
Today many contextual providers use additional signals that come from outside
of the content, meaning they are able to offer segments not entirely reliant on
keywords. These segments can look like behavioural audiences at first glance
but they are tied to an ad impression and not a user ID, ensuring they are more
privacy compliant.
25
As well as leveraging the power of AI, we are seeing a new wave of contextual
providers employ the following techniques to reach new audiences for media
buyers:
• Signals – These providers look at external signals on the bid request such as
the user’s location, the device they are using and the time of day to inform
targeting. These providers can do things such as recognise change in emotion
and sentiment based on the content a user is consuming and therefore
predict consumer behaviours to a high degree of accuracy.
• Feedback loop – These companies are finding out what works for an
advertiser’s campaign, potentially even from an established campaign, and
then discovering content that has similar attributes to build a real-time
feedback loop. In this way, a campaign is targeted to the most performant
inventory from the get-go, and machine learning means these segments can
be constantly updated. This is a good tactic for prospecting, performance
campaigns and is driving conversions for brands today.
26
Section 9. Key Considerations
There are several ways contextual targeting can be used in intelligently to deliver
effective results. Here are some of the key considerations for contextual
targeting:
Campaign KPIs: Think about your KPI’s and work back from there. Know
whether you want to prioritise scale or performance.
Brand suitability: Don’t think about keywords you want to avoid, but think
about the topics or type of content you want to avoid.
Creative messaging - Look at the creative and think about how the messaging
could inform the context you choose to target. Play with the creative message in
a campaign so it resonates with the context of the environment it is placed in.
For example, brands that have corporate goals around reducing their
environmental impact should be running their campaigns against content
promoting sustainability.
27
Video/CTV: Today contextual providers can classify video and CTV content by
leveraging a combination of the second-by-second audio and frame-by-frame
visual recognition technology to provide a true understanding of the video
content. This can be done for VOD, OLV and live-streamed content. Be sure to
keep this in mind when thinking about utilising video for your campaign.
“
Contextual advertising is making its way back into the
boardrooms of most advertisers today and will be one of
the driving forces for changes after the years of
Bart Sieswerda, predominance of behavioural and cookie-based targeting.
Director, Business
As an industry, we should embrace contextual advertising
for all of its obvious benefits and apply our learnings from
Development, the ‘old ways’. When we truly start putting our customers
Nordics & Benelux, first again and are mindful about privacy, brand suitability
and our messaging, we might actually get consumers to
Rakuten Advertising
start liking advertising again (beyond just being a necessary
evil to access premium content for free). This does mean
that the industry will need to make some drastic changes;
• Text and media analysis: Analysing and extracting signals from pictures,
audio or video would provide additional revenues but this is still a fresh
concept that needs more work.
29
The key to the future expansion of contextual targeting is industry collaboration.
Technology platforms will need to work with those specialising in contextual
targeting, allowing them to build on top of their technology in order to provide
advanced, comprehensive solutions for advertisers. In doing so, these platforms
will be able to offer brands a better understanding of high-performing content
that inform successful online and offline marketing strategies and build accurate
and extensive audience profiles based on more than assumptions about past
behaviour.
Looking to the future, the possibilities that contextual targeting will bring across
screens and formats is exciting. While likely that regulations and consumer
viewing habits will continue to shift and evolve, contextual targeting, capable of
delivering scale across a range of KPIs, will remain a safe bet for technology
platforms, advertisers, publishers, and ultimately consumers.
30
Section 11. Summary
Whilst the concept of contextual advertising isn’t new, it’s evolving and growing
and expected that $412B will be spent on contextual advertising by 2025. But as
with any opportunity it is important that we, as an industry, work together to
support the buying of this medium.
IAB Europe is working with some of its members to provide additional insight
into post third-party cookie solutions, such as contextual advertising, in a series
of webinars, guides, and working groups.
Now is the time to prepare for a post third-party cookie era and we’re excited to
see how contextual advertising will flourish.
31
Contributors
IAB Europe would like to thank the following contributors who helped to author
this Guide:
32
Garrett McGrath, Vice-President, Product
Management, Magnite
33