In today’s digital advertising ecosystem, various programmatic buying techniques can be used to reach a target audience. In which there are three most common models: Header Bidding, Waterfall, and Open Bidding. Although, the differences between the three may be confusing to people. This article will explore the three most prominent programmatic buying methods and discuss the differences between each.
If you’re to lazy to read the details, jump to the compare table here for TLDR version.
1. Header Bidding
– Definition:
Header bidding is considered “the upgraded version” of the Waterfall method. It is an advanced programmatic ad technique that enables Publishers to simultaneously offer their ad inventory to a number of ad exchanges and networks. With header bidding, bid requests are sent to multiple demand partners in real-time maximizing the ad inventory’s value
There are two types of Header bidding: Server-side and Client-side.
A browser-based ad auction is created using client-side header bidding. On the other hand, server-side header bidding starts an ad auction on an external server. As a result, server-side header bidding speeds up page loads and uses less processing power from the user’s browser.
– How it works:
To start with Header bidding, it is ideal to have at least 1 million page views per month, though it varies depending on your audience’s GEO. Plus, publishers must own an ad server or use a third-party ad server to deliver ads.
The bidding starts within the user’s browser as soon as the page loads. The code calls all demand partners at once to bid on the available impression. The highest bidder gets the impression, and the ad server delivers the creative to the screen.
– How to set up:
To set up, Publishers need a Javascript tag to build a header bidding wrapper; the most popular JS tag provider is Prebid.
1. Find demand partners who are willing to bid on your inventory
2. Set up a wrapper.
- A wrapper organizes all buyers and sets the rules for the programmatic auction.
- Wrappers function like a repository that contains code from various demand partners and determines the auction’s protocol.
- Set it up yourself or get it from a header bidding provider.
3. Integrate the wrapper with the ad server (ad server setup).Â
4. Set auction time limits as well as floor prices.
More about header bidding wrapper:
A prebid (header bidding) wrapper contains bids received from all the SSPs; it sends the winning bid to the Publisher’s ad server.
Without a wrapper, Publishers will have to integrate SSPs one by one, which leads to too many codes on the site.
The bids come through what are called Bidding Adapters provided by SSPs. SSPs and the wrapper communicate via these adapters.
– Advantages and Disadvantages:
Many Publishers prefer using this auction mechanism since it brings them greater yield. Adding just a single header bidding source can increase yield by up to 10%. This comes from the intense competition between advertisers for the inventory, increasing fill rates, and higher eCPMs.
Moreover, Publishers can sell inventory on a per-impression basis by pooling their inventory into a single server-side supply, providing them better insight into how much their impressions are actually worth.
By offering impressions to many demand partners in real-time, header bidding (server-side) helps increase the page load speed. Not to mention the prebid.js is entirely open-source and accessible for every Publisher.
On the downside, the technique is usually not very user-friendly. Setting up and maintaining the header bid wrapper is quite tricky.
An advertiser connected to multiple ad networks can unknowingly end up competing against themself. This will cause what is called duplicate bidding.
Lastly, the problem with page latency arises when using a client-side server since everything happens on the user’s browser, and combines with calling for many demand partners at the same time.
2. Waterfall
– Definition:
Also known as Daisy-chaining, Waterfall is a programmatic process of selling inventory in the order of priority between advertisers. Inventory is put up for auction, and sequentially, ad networks have a choice to bid for the inventory; the inventory is passed down until a bid meets the floor price.
– How does it work?
- The ad networks are usually ranked according to the average historic yield (CPM, fill rate, latency, etc..) they have produced for the Publisher. This means that an ad network with a history of purchases for a higher price will get a first chance at further impressions from the same Publisher.
- If the first network does not produce a satisfactory bid for some reason, the ad server calls upon another ad network to bid, and so on.
- The ad network bidding with a price higher or equal to the floor price gets the impression.
– How to set up:
Waterfall bidding is typically available within mediation platforms like AppLOVIN MAX, Levelplay by ironSoure, or the prominent Google AdMob. All three are PremiumAds-supported mediations.
- Get the passback tags from the platform
- Publishers set up the priority list
- Specify a price floor – the minimum acceptable price for the ad placement.
– Advantages and Disadvantages
This model solves the problem of unfilled inventory by offering it to one-by-one ad networks. It is also a great option for Publishers who prefer direct deals. An additional plus point is that it has an easy setup since it only requires pasting passback tags from one SSP (or ad network) to another.Â
However, as this model was one of the first to be born, it comes with several drawbacks compared to other mechanisms. First, Publishers suffer from low revenue because once it passes impressions from one SSP/ad network to another (known as “passback”), causing CPM price floors decrease, preventing high-paying advertisers from bidding.
Second, the long process of offering impressions sequentially to each ad network causes high page latency. Lastly, in contrast to header bidding, where there is a single auction, the sequential selling procedure produces substantial impression discrepancy.
– Waterfall mediation:
Ad mediation platforms generally prioritize and sequence ad networks by potential profit. When an app developer requests an ad, the highest-performing network is called first until its ad bidding pool is depleted, then moving on to the next highest-performing network, and so on. This ensures that the most beneficial ads are served first while lower-paying ads are requested last.
3. Open Bidding
– Definition:
Google’s server-to-server technology called “open bidding” enables publishers to invite numerous external ad exchanges to compete for their inventory. In a process known as dynamic allocation, real-time bidding takes place between demand partners and Google’s Ad Manager line items on a first-price basis.
– How does it work:
- Google Ad Manager ad server receives information after the Publisher sends an ad server.
- Ad Manager runs a unified auction to determine the best yield.
2a. Ad Manager selects the best-trafficked line item to compete in the unified auction.
2b. Ad Manager sends a bid request to targeted yield partners.
2c. Targeted yield partners run their own auction and return their most competitive bid to Ad Manager.
2d. Ad Manager hosts a unified auction and selects a winner.
- A creative or Mediation list is returned to the publisher.
– How to set up:
The process is done solely on GAM.
- Publishers must have access to Google Ad Manager
- Publishers with GAM for small business accounts need to manually enable Open Bidding via a Google representative or a Google Certified Publishing Partner (like PremiumAds)
- Inform the representative that Open Bidding is on for your account and provide details like domain and network ID.
- Set up Open Bidding.
– Advantages and Disadvantages:
One of the main advantages of Open bidding is its server-to-server connections, which ultimately decrease page latency, thus increasing ad viewability and revenue.
Open bidding is Google’s answer to Header bidding, when something is great, Google must jump on it and make it greater. Therefore, this mechanism also helps increase the competition for Publishers’ inventory, which results in higher ad revenue, without affecting user experience. Another huge plus point for the model is that it comes with exclusive Google demands from AdX and AdSense.
Not every product is perfect, therefore, Open Bidding also comes with some imperfections.
Unlike Header bidding with Prebid.js – open-source and free, Google’s Open Bidding has a higher entry barrier. It is only available to Publishers who have AdX access and use Google Ad Manager 360 (the premium and paid version of GAM for small businesses). Moreover, joining the above is not easy, Publishers need to meet certain criteria, or choosing the easier choice is to partner up with a Google Certified Publishing Partner like PremiumAds.
In addition, the process of bidding is entirely based on Google’s ad server to choose the winner, so it is less transparent than Header bidding since Publishers can’t determine with certainty why a specific advertiser won the auction.
That’s all for differences between the three auction mechanisms: Header bidding, Waterfall, and Open bidding. You can find below a table for better and more comprehensive comparisons.
After exploring each of the three major programmatic buying methods, it is clear that all three offer distinct advantages that can help Publishers meet their needs. Header bidding and open bidding are more advanced models, while the waterfall is slightly old-fashion but still, they are innovating. Ultimately, depending on the Publishers’ preferences, they can use any of the three methods to create great revenue, though they could use some ad ops help from experts like PremiumAds to boost it further.
