Blog post

What is programmatic advertising, and how does it work?

Date of post

8 January 2024

Blog categories

Read time

6 mins
Nick Janaway sits at his desk in the Marketing Labs office

In this blog, we demystify programmatic advertising, looking at what it is and how it works. We examine the role of ad-buying specialists, why data is essential, and the benefits of programmatic advertising. 

Prefer to listen to this blog? Episode 12 of the Marketing Blabs podcast is dedicated to discussing programmatic advertising.

Link to listen to episode 12 of the Marketing Labs podcast

What is programmatic advertising? 

Programmatic advertising is an automated way of buying ad space. This can be achieved across many platforms and often uses machine learning software. 

In practice, this means relinquishing some of the things you used to be able to control and allowing the ad platform to make decisions on your behalf. 

Platforms like Google and Facebook have been moving away from manually controlled campaigns for a while now. If you have access to an ads account that you use regularly, you’ve probably dabbled in elements of automation without being aware of it. 

In simple terms, you give the platform your budget allocation (and some other information), and it will spend that money making sure your ads are seen by the right people on the right platforms at the right time. 

Demand-side vs. sell-side – the difference 

One of the benefits of programmatic advertising is that you can buy space across multiple platforms and in various formats, i.e. image, video, listings, etc, from one place. You can let the automated tools create your ad for you, or you can provide dedicated ads for it to use.  

This is a good thing, but it does mean there’s a lot to learn about how it works. A couple of the terms you’ll encounter frequently are demand-side and sell-side platforms, so let’s define those here. 

Demand-side platform (DSP)

A DSP allows advertisers to buy space across a breadth of websites at once, giving you access to a wider range of digital advertising inventory. The Google Display Network is an example of a demand-side platform. 

It comprises a vast network of websites, apps, and other online platforms where advertisers can display their ads to a broad audience. 

Supply-side platform (SSP) 

This type of bidding platform allows publishers to sell their ad space when a user visits a page on their site. Amazon is a good example of a supply-side platform (SSP). 

It’s possible to connect an SSP with a DSP using an ad exchange. 

An ad exchange is an online marketplace that facilitates publishers and advertisers exchanging ad inventory. It uses automated buying and selling of digital ad impressions through real-time bidding (RTB) auctions. 

Are people still involved in buying and selling? 

It would be a mistake to leave the buying and selling of ad space to AI alone. Skilled digital marketers are still required.

In theory, the time you save manually controlling your bids should be spent on planning and optimising your campaigns, but remember, you won’t be privy to every decision the system makes. 

Most of what goes on is hidden (one of the biggest drawbacks of programmatic advertising). You’ll be expected to put a lot of trust in the system, so it’s imperative that you monitor absolutely EVERYTHING. 

Keep an eye on the output and ensure it performs for you. You may need to adjust your campaign set-up if you suspect something’s wrong, including updating your audiences, brand safety guidelines, and ads. 

Why is data critical in programmatic advertising? 

Programmatic tools need to know who your customers are to get your ads in front of the right people. The more data you can gather and put into the system, the better decisions it can make.

But data is hard to come by, especially for young businesses and start-ups. There is often a gap between your defined customer and the reality. What you think you know may turn out to be different. 

So, how do you close this gap? Start by ensuring that your cookie consent and privacy notices are set up correctly so you can legally use the data you collect.

Today’s strict cookie policies on third-party tracking can make it hard to build up a picture of your audience, but it’s still possible to glean data this way for the time being. Using your first-party data, you can build look-alike audiences or retarget your known customers.

What is a look-alike audience? 

A look-alike audience is a group of individuals who share characteristics and behaviours with an existing audience or a specific set of users. 

Algorithms and machine learning tools can identify patterns, trends and similarities among an existing group and create a new audience that shares these commonalities with the source audience. This newly generated look-alike audience then becomes a target for advertising campaigns.

The benefits of programmatic advertising 

The future of ad buying almost certainly resides in machine learning, and there are plenty of reasons to be cheerful about that. 

  • Cross-device capability

With programmatic advertising, running campaigns across various platforms is possible. This allows you to serve ads in multiple places, reaching more people in the most cost-effective way. 

  • Extended audience reach 

By combining first and third-party data, you can target look-alike customers with preferences and behaviours similar to your existing audience. This will extend the reach of your ad campaigns and improve efficiency. 

  • Precision targeting 

Programmatic advertising enables targeted campaigns. It’s possible to leverage data to target specific demographics, behaviours, interests, and even individual users in real time, increasing the ads’ relevance.

  • Increased efficiency 

In theory, eliminating the need for manual intervention should save you time so you can concentrate on optimising your campaigns. The more you buy ad space this way, the more efficient you will become. 

And it doesn’t hurt to lean on a PPC services company like Marketing Labs when you need help.    

Case study – Google PMax 

Google Performance Max (more commonly referred to as PMax) is a Google ads platform that allows you to buy space across their inventory, including Search, Display, and YouTube. It uses machine learning to work out the best format and place to position ads based on your goals and audience.  

With the correct input, it can be a powerful tool, but to be effective, you must clearly define your goals and tightly control your audience

We use this system to buy ad space for a diverse range of clients across various industries. 

LuluB is a handmade jewellery brand in the UK. Using PMax, we’ve successfully: 

  • reduced the cost-per-click (CPC) with precise audience targeting and 
  • increased revenue by refining their ad creatives. 

The combination of more efficient traffic and improved conversion effectiveness meant a sizable increase in ROAS (return on advertising spend).  

Here’s an example of what you can expect from a typical PMax campaign in revenue growth. 

Screenshot of the results from a typical Google PMax campaign

In the screenshot above, the blue line shows purchases/sales, and the red line shows the value of the conversions.

The bottom line on programmatic advertising 

Programmatic advertising can harness powerful results, but you must manage it carefully and test variations to get the most from it.

Leaving the entire process to automation will end in underwhelming performance, with wasted investment and disappointing results. 

You need clear goals, a well-defined audience, and quality assets. To ensure success, it is essential to continually monitor the outcomes and refine your campaigns as they progress.

Do all of this, and you’ve got a recipe for success. 

Up next in our series of PPC blogs is Bing vs Google ads.

Post author

As head of digital at Marketing Labs, Nick is involved in delivering all of our services. He’s a man of many talents with experience in SEO, paid advertising and social media – to name a few!

More content like this

Get in touch with Marketing Labs®

Recent blog posts

Psst! Did you know we’ve got our very own podcast?