Trends in programmatic 2021


Last year, we took it upon us to predict some of the most exciting trends we believed would dominate the programmatic advertising space in 2020.

Little did we know that a few months into the new year, the world would be turned upside down, and by the time our Face-to-face meetings turned into Facetime meetings, most industry predictions were thrown out of the window.

Now, we’re turning our focus to the new year and look at some of the key trends we believe will define the programmatic advertising space in 2021.

Artificial intelligence was one of the key topics we included in last year’s forecast, and although 2020 certainly turned out to be a big year for AI and machine learning in programmatic, we can only see this grow further this year.

With customer journeys becoming more complex and target audiences increasingly specific, it’s becoming more and more difficult to display your ad in front of the right people at the right time. AI’s ability to identify and accurately predict buyer patterns at scale and in real-time is something that human power simply can’t live up to.

In the past, AI has been something of a buzzword, but after seeing the success early adopters have achieved after implementing AI into their programmatic strategy, more and more advertisers swear by the power it holds.

Programmatic Digital Out Of Home or DOOH refers to the automated media trading on digital out of home platforms such as billboards or signage. It’s the evolutionary offspring of one of the oldest and most traditional advertising channels; Out Of Home (OOH) advertising.

Marketeers have long struggled to bridge the customer journey gap between online/offline, and at home/outside of home. Combined with smartphone location data, Digital Out Of Home has the power to drive conversions even in the offline world.

Both DOOH and OOH digital advertising were hit hard by global lockdowns brought on by the pandemic, but the Advertising Association + Warc Expenditure report predicts that it will make a comeback in 2021 with an expected growth of 38.7%. 

Take a look at our programmatic DOOH here.

With more than 150m monthly listeners in the US alone, podcasts are becoming more and more popular.

eMarketer estimates that spend on podcast ads will double by 2022 and reach $106.5 million, and it’s not just in the US that podcasts are gaining momentum. According to WARC, 38% advertisers globally plan to boost their podcast ad spend in 2021.

We can reach users across apps such as Acast, Spotify,, Tunein, Deezer across both iOS, Android and across a range of devices including smartphones app & web as well as smart speakers and even in car.

Here at Advant Technology, we’re big fans of programmatic podcast advertising and have vast experience of running both local, national and international campaigns. Read more about our digital audio advertising solutions here.


A guide to contextual targeting


An ad for a stroller displayed on a site for expecting parents or a shampoo ad shown before a hair makeover video. Makes a lot of sense, right? 

The method that allows for ads to be displayed alongside highly relevant content is called contextual targeting and can be done through specific keywords or categories relating to the central theme of a website or page.


Before launching a campaign, we choose a set of specific keywords, topics and themes relating to your ads on the Display Network. Google then analyses the content of publishers’ websites and matches it against your chosen keywords, topics, language and location. 

An example would be if you are a restaurant chain looking to run an awareness campaign for your newly-opened branch. If a person is nearby and reading a food blog, your ad might get displayed.

Because the ad isn’t disruptive (the user is already interested in this type of content), the user is more likely to respond positively to your ad. 


The Google Keyword Planner is a great starting point to start building your keyword list. There are best practices for how to maximise the success from your keyword list. Typically, Google recommends that each ad group should contain anywhere between five to 50 keywords, and they should be shorter and more to the point than the keywords you’d typically use for organic search. 


For a successful contextual ad campaign, it is also important to consider the buying journey of your target group. Do you want to target users in the consideration phase? What type of content will users who are ready to make a purchase consume? By getting a good understanding of this journey, you’ll be able to draft up a more effective list of keywords. 


We recently worked with  Bompas & Parr who, together with the Design Museum, organised the competition Fountain of Hygiene to raise funds for the British Red Cross in light of Covid-19. Read more about the campaign. 

The aim of the campaign was to maximise the number of competition entries. To do so, we used contextual targeting to get the ad displayed alongside relevant content and in front of designers and creatives who would be likely to enter the competition. 

Below you’ll find two examples of ad placements for this campaign.


contextual targeting example

The Design Week placement shows an example of category contextual targeting where a page matches a specific pre-assigned category for which the ad matches. 

The Design Week has a large readership of designers and creatives and our ad would likely fit well alongside any page on the site.



The Evening Standard  shows a result of keyword contextual targeting where the ad is displayed alongside content that matches a specific keyword.

Although the site isn’t  necessarily specifically targeted towards our main target group, we were still able to display out content alongside highly relevant content.


Our team are experts at planning contextual ad campaigns, so if you’re keen to try out this method for your next ad campaign, please get in touch and we’d be more than happy to help you plan it. 


Open Auctions


I know what you’re thinking. A digital advertising phrase that is not an acronym? What is going on? 

Calm down there cowboy. You can always trust the digital advertising community to come up with TLSs (three-letter acronyms). For Open Auctions, they’ve settled on RTB (Real-Time Bidding).

What are open auctions?

An open auction is a real-time bidding process that allows multiple publishers and advertisers to bid on ad inventory. 

What are the pros?

  • Cost-effective: these types of auctions are great if you’re looking for a large reach and have a relatively small budget to play with
  • Straightforward: easier and faster to set up compared to other media-buying methods 
  • Targeting: access to multiple powerful targeting options including third-party data
  • Optimisations: the advertisers can be creative with formats and optimisations to engage with the user
  • Reach: you can reach target audiences irrespective of the sites they are visiting 

What are the cons?

  • Non-premium: inventory us not as premium as other buying-methods
  • Ad fraud: chances of ad fraud can be greater compared to closed auctions
  • Brand safety: a chance to appear on sites you don’t want to associate with
  • Lack of transparency: no set price structures 

When to choose open auctions:

  • When you want access to a large range of inventory at lower rates and want to have the fluidity of easy access 
  • When access to third-party data and other targeting technologies would be beneficial to overlay on top of your open auction

When to avoid open auctions:

  • When you want access to more premium inventory (mastheads, roadblocks etc.) 
  • When the publisher is offering premium inventory along with premium targeting only accessible buy some form of programmatic media 

If you want to learn more about other types of programmatic media-buying, check out our blog post on PMPs (private marketplace deals).


Private Marketplace Deals (PMPs)


A guide to private marketplace deals

If you’re new to the world of digital advertising, you may have noticed that the industry is full (and we mean FULL) of buzzwords and acronyms. To help navigate this notoriously complex landscape, we thought we’d break down some of the most commonly used industry jargon. 

We’re starting off with one of our favourite TLAs (three-letter acronyms that is): PMP 

What is PMP?

PMP stands for Private Marketplace and is a programmatic media-buying method.

Compared to open auctions, which you can read about here, PMP is an invitation-only bidding process where publishers invite selected buyers to take place in real-time bidding to purchase their inventory. 

The minimum price of the bids is referred to as ‘floor price’ and is negotiated before the auction starts. Private marketplace deals are more targeted and premium than open auction deals, which is reflected in the price. 

What are the pros of private marketplace deals?

  • Transparency: It gives both buyers and sellers transparency of the auction
  • Relationship building: Publishers and buyers negotiate and agree on terms and pricing, which allows for mutual understanding and strong relationships to be built
  • Premium inventory: Gives access to higher-quality inventory before it ends up on the open market
  • Targeting: Allows advertisers to target specific audiences 

What are the cons of private marketplace deals?

  • Only partly-automated, meaning the buying process can require more manual labour and time compared to open auctions 
  • Non-guaranteed inventory: there is no guarantee you’ll get the inventory you want
  • Pricing: Costs tend to be slightly higher than open auction deals

When to choose PMPs:

  • When you have a whitelist of publishers you want to display your ads across 
  • When you want access to higher-quality inventory before it ends up on the open market
  • When you have a specific target audience for your campaign

When to avoid private auctions:

  • When a campaign calls for more premium inventory (mastheads, roadblocks etc.)
  • When you want a fixed price for the inventory 

If you want to learn more about other types of programmatic media-buying, check out our blog post on open auctions here.


Trends in programmatic 2020


2020 is an exciting year with continued growing investment in the programmatic advertising space. Millions of apps, sites and endless formats from which to choose from. The fragmentation of media and digital touchpoints have arguably made marketers’ jobs increasingly harder. At Advant Technology we help brands and agencies navigate the complex world of programmatic through expert knowledge of ad tech platforms.

The trends we’re exploring this year include:

Artificial Intelligence

Undoubtedly one of the hottest topics in programmatic advertising right now. We’ve seen massive uplifts of up to 40% in performance campaigns optimising towards metrics including cost per acquisition, cost per view and effective cost per click. You can explore our dedicated post here.

AI has been helping analyse the thousands of data points available to improve optimisation beyond what a human can achieve.


This fascinating technology will enable a more connected world. Speeds that can surpass 100giagabits per second is 100 times faster than 4G technology. This will enable advertisers to deliver a richer advertising experience. High-speed 5g means less buffering time for rich media and heavy high-quality video files thereby improving the user experience and potentially encouraging more advertising revenue, clearly then a huge opportunity.

Digital Audio Ads

The world’s first radio station was KDKA launched in Pittsburgh in 1920. Radio’s come a long way since then playing a huge role in how we receive our news, listen to music and more, fast forward 100 years and the US will be spending more time listening to digital audio than radio according to emarketer.

Contextual Advertising

In a post-GDPR and CCPA apocalypse, world audience data activation has many challenges. Contextual advertising technology does not use cookies or customer data and is, therefore, having a resurgence. Contextual ads use keyword and semantic technology to understand the context of the page and thus allow advertisers to deliver ads that are more relevant.  


Basic targeting methodology using a DSP


A DSP aka demand side platform allows you to target users via multiple formats that generally include display banners, video and native advertising. Whatever the format, you need to make sure that your ad has the best possible chance of reaching the right audience. In this short post, we’ll walk you through the main ways a DSP allows you to target. Don’t forget to research and fully understand who your audience so that you select the right targeting parameters. Let’s dive in and take a look:

Geo-Targeting is probably the most basic targeting methodology and at the very least you should apply some kind of country level targeting. Clearly there’s no point in advertising to people in Timbuktu if you only sell to people based in the United Kingdom.

Options include Country > state/region > zip code / post code or LAT LONG co-ordinates.

Day Part Targeting refers to the hours of the day that your campaign is live. We often have requests from advertisers who only want to advertise between the hours of 7am and 10pm. This is normally associated with brand safety, however, there maybe other reasons for using day part targeting. Let’s say you have an app that helps people relax, well maybe it makes sense to advertise just before bed or first thing in the morning, where people are in a different frame of mind and are perhaps more open and could be more responsive to your messages.

Audience targeting, typically allows you to select audiences based on their interest or behaviour. Here’s a quick snapshot at the most effective routes.

  • Affinity Audiences (interest)
  • Customer match (using email address, phone number or other method to learn more visit Liveramp 
  • Demographic such as age and gender
  • In-Market audiences e.g. inmarket to buy a used car

Re-Targeting, we’ve all seen those annoying ads that seem to follow us around after we’ve visited a retailers site. These ads are known as retargeting ads and can have a product feed showcasing the exact products you viewed + other products that you might be interested in. These product feeds can be implemented using an adserver such as campaign manager from Google DV360. Where most businesses go wrong with re-targeting is not frequency capping and thus bombarding the user with too many ads too often. Make sure you control the number of times an ad is seen by a user respecting the balance between selling and privacy and applying a reasonable frequency cap such as 1 in 24 hours.

Device targeting, perhaps the product you’re selling is an app? Well it probably makes sense to only serve ads on mobile devices, where the path to purchase is shorter. Device targeting allows you to control whether you target mobile, desktop or tablets.

Browser Targeting – Maybe you need to target Apple users only. Well perhaps a good starting point would be targeting the Safari browser whose audience is typically Apple users. Browser targeting on the most widely used DV360 platform include:

  • Chrome
  • Firefox
  • Internet Explorer
  • Microsoft Edge
  • Opera
  • Safari

We’ll look to deep dive into some of the more advanced DSP targeting tactics later but for now we hope you find the above a welcome introduction.

Photo by Yucel Moran on Unsplash Photo by Nathan Dumlao on Unsplash Photo by Daniel Korpai on Unsplash


Benefits of Programmatic Advertising


Programmatic advertising has been for many years been a medium to buy advertising online, however, in the past couple of years we’ve seen the growth of programmatic for offline advertising like Digital Out of Home and this is only set to increase.

For the purposes of this article, we’ll be focusing on online programmatic advertising which is defined by DIGIDAY as

“the use of software to purchase digital advertising, as opposed to the traditional process that involves RFPs, human negotiations, and manual insertion orders. It’s using machines to buy ads, basically.

Now we know what it is, let’s explore the benefits:


Programmatic advertising does have it’s bad players and there are people out there who want to game the system, but on the whole programmatic advertising does allow advertisers to see exactly where their ads are being placed, on which sites/apps and the performance of each placement and whether those ads are viewable.

Real-Time Measurement

Understanding how your campaign is performing in real-time is a real joy, no more waiting around for weeks on end to see the effects of your advertising or waiting for your agency to provide laggard reports. Real-time measurement allows you to see key data 24 hours a day (if you’re that keen). Here is a full list of the metrics you can export from DV360 right now! As you can imagine a huge amount of data is generated so it’s key you focus on the metrics that matter to your campaign, shameless plug, we can help you with that!

  • Begin to Render Eligible Impressions
  • Begin to Render Impressions
  • Billable Impressions
  • Click Rate (CTR)
  • Clicks
  • CM Post-Click Revenue
  • CM Post-View Revenue
  • Engagement Rate
  • Engagements
  • First-Quartile Views (Video)
  • Midpoint Views (Video)
  • Third-Quartile Views (Video)
  • Complete Views (Video)
  • Impressions
  • Media Cost
  • Post-Click Conversions
  • Post-View Conversions
  • Revenue
  • Total Conversions
  • Total Media Cost
  • Tracked Ads
  • TrueView: Views
  • Active View: % Audible and Visible at Completion
  • Active View: % Audible and Visible at First Quartile
  • Active View: % Audible and Visible at Midpoint
  • Active View: % Audible and Visible at Start
  • Active View: % Audible and Visible at Third Quartile
  • Active View: % Audible Impressions
  • Active View: % Full-Screen
  • Active View: % Fully On-Screen 2 sec
  • Active View: % In Background
  • Active View: % Measurable Impressions
  • Active View: % of Ad Played
  • Active View: % of Completed Impressions Audible and Visible
  • Active View: % of Completed Impressions Visible
  • Active View: % of First Quartile Impressions Audible and Visible
  • Active View: % of First Quartile Impressions Visible
  • Active View: % of Midpoint Impressions Audible and Visible
  • Active View: % of Midpoint Impressions Visible
  • Active View: % of Third Quartile Impressions Audible and Visible
  • Active View: % of Third Quartile Impressions Visible
  • Active View: % Play Time Audible
  • Active View: % Play Time Audible and Visible
  • Active View: % Play Time Visible
  • Active View: % Viewable Impressions
  • Active View: % Visible 10 Seconds
  • Active View: % Visible at Completion
  • Active View: % Visible at First Quartile
  • Active View: % Visible at Midpoint
  • Active View: % Visible at Start
  • Active View: % Visible at Third Quartile
  • Active View: Audible & Fully On-Screen for Half of Duration (15 sec. cap) Impressions
  • Active View: Audible & Fully On-Screen for Half of Duration (15 sec. cap) Measurable Impressions
  • Active View: Audible & Fully On-Screen for Half of Duration (15 sec. cap) Rate
  • Active View: Audible & Fully On-Screen for Half of Duration (TrueView) Impressions
  • Active View: Audible & Fully On-Screen for Half of Duration (TrueView) Measurable Impressions
  • Active View: Audible & Fully On-Screen for Half of Duration (TrueView) Rate
  • Active View: Average Viewable Time (Seconds)
  • Active View: Custom Metric Measurable Impressions
  • Active View: Custom Metric Viewable Impressions
  • Active View: Custom Metric Viewable Rate
  • Active View: Eligible Impressions
  • Active View: Impression Distribution (Not Measurable)
  • Active View: Impression Distribution (Not Viewable)
  • Active View: Impression Distribution (Viewable)
  • Active View: Impressions Audible and Visible at Completion
  • Active View: Impressions Visible 10 Seconds
  • Active View: Measurable Impressions
  • Active View: Not Measurable Impressions
  • Active View: Not Viewable Impressions
  • Active View: Viewable Impressions
  • % Clicks Leading to Conversions
  • % Impressions Leading to Conversions
  • Audio Client Cost eCPCL
  • Audio Revenue eCPCL
  • Client Cost
  • Client Cost eCPA
  • Client Cost eCPA (PC)
  • Client Cost eCPA (PV)
  • Client Cost eCPC
  • Client Cost eCPM
  • Client Cost Viewable eCPM
  • CM Post-Click Revenue
  • CM Post-Click Revenue + Cross-Environment
  • CM Post-View Revenue
  • CM Post-View Revenue + Cross-Environment
  • Conversions per 1000 Impressions
  • Gmail Conversions
  • Gmail Post-Click Conversions
  • Gmail Post-View Conversions
  • Post-Click Conversions
  • Post-Click Conversions + Cross-Environment
  • Post-View Conversions
  • Post-View Conversions + Cross-Environment
  • Revenue
  • Revenue eCPA
  • Revenue eCPA (PC)
  • Revenue eCPA (PV)
  • Revenue eCPC
  • Revenue eCPM
  • Revenue Viewable eCPM
  • Total Conversions
  • Total Conversions + Cross-Environment
  • Video Client Cost eCPCV
  • Video Revenue eCPCV
  • AdLingo Fee
  • Adloox Fee
  • Adloox Pre-Bid Fee
  • Adsafe Fee
  • AdXpose Fee
  • Agency Trading Desk Fee
  • Aggregate Knowledge Fee
  • Billable Cost
  • comScore vCE in DoubleClick Fee
  • Data Fees
  • Data Management Platform Fee
  • DoubleVerify Fee
  • DoubleVerify Pre-Bid Fee
  • Evidon Fee
  • Integral Ad Science Pre-Bid Fee
  • Integral Ad Science Video Fee
  • MediaCost Data Fee
  • MOAT Video Fee
  • Nielsen Digital Ad Ratings Fee
  • Platform Fee
  • Platform Fee Rate
  • Premium Fee
  • ShopLocal Fee
  • Teracent Fee
  • Third-Party Ad Server Fee
  • TrustMetrics Fee
  • Vizu Fee
  • Audio Mutes (Audio)
  • Audio Unmutes (Audio)
  • Companion Clicks (Audio)
  • Companion Impressions (Audio)
  • Complete Listens (Audio)
  • First-Quartile (Audio)
  • Midpoint (Audio)
  • Pauses (Audio)
  • Starts (Audio)
  • Stops (Audio)
  • Third-Quartile (Audio)
  • Companion Clicks (Video)
  • Companion Impressions (Video)
  • Skips (Video)
  • General Invalid Traffic (GIVT) Begin to Render Impressions
  • General Invalid Traffic (GIVT) Clicks
  • General Invalid Traffic (GIVT) Impressions
  • Invalid Clicks
  • Invalid Impressions
  • Invalid Active View Eligible Impressions
  • Invalid Active View Measurable Impressions
  • Invalid Active View Viewable Impressions
  • General Invalid Traffic (GIVT) Active View Eligible Impressions
  • General Invalid Traffic (GIVT) Active View Measurable Impressions
  • General Invalid Traffic (GIVT) Active View Viewable Impressions
  • General Invalid Traffic (GIVT) Tracked Ads
  • Invalid Begin to Render Impressions
  • Invalid Tracked Ads
  • TrueView: General Invalid Traffic (GIVT) Views
  • TrueView: Invalid Views
  • Audio Media Cost eCPCL
  • Media Cost
  • Media Cost eCPA
  • Media Cost eCPA (PC)
  • Media Cost eCPA (PV)
  • Media Cost eCPC
  • Media Cost eCPM
  • Media Cost Viewable eCPM
  • Total Audio Media Cost eCPCL
  • Total Media Cost
  • Total Media Cost eCPA
  • Total Media Cost eCPA (PC)
  • Total Media Cost eCPA (PV)
  • Total Media Cost eCPC
  • Total Media Cost eCPM
  • Total Media Cost Viewable eCPM
  • Total Video Media Cost eCPCV
  • Video Media Cost eCPCV
  • Profit
  • Profit eCPM
  • Profit Margin
  • Profit Viewable eCPM
  • Programmatic Guaranteed impressions passed due to frequency
  • Programmatic Guaranteed savings re-invested due to frequency
  • Refund Billable Cost
  • Refund Media Cost
  • Refund Platform Fee
  • Audio Mutes (Video)
  • Audio Unmutes (Video)
  • Average Display Time
  • Average Interaction Time
  • Completion Rate (Audio)
  • Completion Rate (Video)
  • Counters
  • Exits
  • Expansions
  • First-Quartile Views (Video)
  • Midpoint Views (Video)
  • Third-Quartile Views (Video)
  • Complete Views (Video)
  • Fullscreens (Video)
  • Interactive Impressions
  • Pauses (Video)
  • Rich Media Engagements
  • Scrolls
  • Starts (Video)
  • Timers
  • Total Display Time
  • Total Interaction Time
  • TrueView: Views
  • Verifiable Impressions


By reducing the human friction points of buying and selling advertising efficiencies can be gained. Less paper shuffling and more action! We can gain efficiency beyond just automation though, with so much data to hand, being available in real time, campaigns can be optimised and machine learning applied in order to increase the ROI on your campaigns. Here at Advant Technology optimisation is our mantra, so we’re constantly assessing and analysing data to improve ROI for our clients.


We’ve explored targeting in a bit more detail in this post, in short though, rather than the old spray and pray method of advertising, programmatic allows you to be ultra-targeted pinpointing only those most likely to buy your products or services. The many targeting parameters available within a DSP help you pick and choose from the billions of impressions available across the real-time bidding ecosystem.


Machine Learning and AI in Programmatic


Programmatic is the process of using software to buy digital advertising. Here at Advant Technology, we specialise in programmatic advertising across the real-time bidding ecosystem. Real-time bidding is the process of buying ads on an individual impression basis. The process of a publisher auctioning an impression from their website/app and a buyer buying the impression is near-instant.


At Advant Technology we’ve employed independent artificial intelligence specialists who build models using data available from our DSP platforms. These models which go beyond machine learning help to increase ROI by using bid modifiers picking and choosing the impressions that are deemed most valuable. All of this can, of course, be done by a human but AI speeds up this process further increasing efficiency and giving our clients a competitive and unique advantage. Typically, the variables that AI will use to make decisions comes from aggregated across the following areas:

  1. Geolocation data (country, region, city)
  2. Time partition data (hour of the day, day of the week)
  3. Size of creatives
  4. Data regarding Operation systems (Family, version, carrier)
  5. Data regarding the browser (browser id, browser language)
  6. Data regarding gender
  7. Data regarding the device (device type [computer, tablet, smartphone..], device model)
  8. Data regarding the placement (placement id, placement group, publisher id, seller id, domain, fold position, mobile app instance)
  9. Data regarding the supply type (browser web, mobile web, app web)
  10. Data regarding the viewability and the completion rate(predicted IAB Viewability Rate, Predicted IAB Video Viewability Rate, Predicted Video completion rate)
  11. Data regarding the segments (when segment are owned by the seat and their data is accessible through the LLD)
  12. Data regarding the frequency/recency of impressions