How To Use FouAnalytics to Reduce Carbon Emissions from Your Programmatic Media
Unlike the stupid claims from IPG Mediabrands offering clients "net-zero media buying" and other greenwashing claims about carbon emissions you can't measure or independently verify, here are simple, concrete things you can do to reduce the "carbon emissions" from your programmatic media buying, yourself, for free, using FouAnalytics (phew that was a long opening sentence).
Look for Supply Sources in In-Ad Accounts
Once you log in to the FouAnalytics dashboard, you will see one of the following data grids if you are using the macros supplied by DSPs -- custom_site, custom_supply, custom_ssp, custom_ADSTXT_AD_SYSTEM_DOMAIN (see screen shot below)
In the leftmost example above, the top 10 supply sources (highest volume first) are listed in the grid. The "25 more" on the bottom right tells you there's 25 more rows that will expand when you click that. That means this campaign has a total of 35 supply sources in use. Having 35 exchanges is massively redundant and is the root cause of massively wasted bid requests - every single ad opportunity is sent to multiple exchanges to solicit bids; and then every exchange sends back many bids from many advertisers. All of this consumes bandwidth and computing power leading to carbon emission, even though only 1 bid will win.
Reducing the number of possible supply paths by reducing the number of exchanges in your media buy will reduce the carbon emissions, significantly. Note I am not stating a specific amount (like grams of carbon), because I don't know what that amount is and have no way to measure it. Even the purveyors of carbon emission reduction vaporware are estimating it using algorithms, not actually measuring it themselves (do you see how easy it is for them to tell you completely BS, made-up numbers, just like many many of the adtech vendors before them?).
Having too many exchanges active in a campaign also leads to supply path leakage -- i.e. your ads going to "somewhere else" (not the sites you want). This is part of the "unknown delta" mentioned in the ISBA reports from 2020 and 2022.
Eliminate as many supply sources as possible
In the example below, supply path leakage is demonstrated. When I tried to target a single domain, but did not specify which exchanges (so all remained active), only 73% of my ads went to the domain I had in my inclusion list. When I eliminated all but one exchange, the accuracy went up to 97% -- i.e. 97% of my ads went to the 1 domain I specified. That "delta" of 24% (from 73% up to 97%) is the supply path leakage. Even if you specified the shortest possible inclusion list (1 domain), if there are too many combinations of supply paths, the exchanges failed to get your ads to the right place.
You can reduce as many supply sources and exchanges from your campaign by going into the DSP and unchecking as many of the exchanges as you can. For example, you don't need 35 exchanges active. You probably only need 3.
How do you choose? You can eliminate the supply sources/exchanges that have the most dark red. If your ads are already mostly going to one supply source, like Google in the top 3 examples (78%, 98%, and 82% respectively), you can easily turn off most or all of the other exchanges and not see any difference in volumes. In the bottom 3 examples above, you can easily just keep the top 3 - 5 by volume and still get your ads on most of the same sites. How do I know? Read on.
Most sites, especially the fraudulent ones, sell their ads through as many exchanges as they possibly can. The exchanges are also all interoperable, because they don't want to miss out on taking tolls from as many ad opportunities as possible. There are layers and layers of other shady things which I won't get into here. But suffice it to say you will get pretty much all of the sites and apps you want even if you reduce the number of exchanges from 35 to 3. Let me show you some concrete proof here. Taking 2 sites commonly seen in programmatic buys, yahoo.com and msn.com, you will see they each sell through multiple exchanges. In the case of Yahoo (left side), you can see that the top 2 represent 70% of the volume. In the case of MSN (right side), you can see that the top 4 exchanges represent 99% of the volume. Those are the exchanges you leave ON, and turn off the other 30 because they are not needed.
Block as many MFA sites as possible
You've heard of MFA ("made for advertising") sites right? Those are the crappy click-bait sites that use many flavors of shadiness to make ad revenue. What I bet you didn't know is just how bad these sites are. If you visit them manually, you won't see anything wrong. This is because special codes added to the url, trigger the avalanche of ads. On the left side there are only 544 ad server requests, 348 tracking requests, and 90 "other." But when special codes are appended to the url (right side) you can see the same page goes bonkers with ads -- 3,490 ad server requests, 2,387 tracking requests, and 685 "other" requests from the same page. How much do these types of sites contribute to the "carbon emissions" problem of programmatic advertising? Right. A Lot. These types of sites are clearly doing something shady. Do you want your dollars to be consumed by them (note that few to none of the ads are seen by humans).
Enormous tree graph of all the calls made from one page.
领英推荐
In the FouAnalytics dashboard, type in "confirmed-MFA-site" and click "filter." I will show you the percentage of impressions going to MFA sites. The screen shot below shows a campaign with 13% of the impressions being wasted on crappy MFA sites. If you CTRL+click the row with the 1 in it, you will isolate confirmed-MFA-site:equals(1) and you can look at the Source data grid to identify the sites to add to your block list.
If you've stuck with me all the way to this point, so you see how you can reduce the carbon emissions of your programmatic buying, without needing to pay for SPO (supply path optimization) or other BS ad tech services that tell you how many grams of carbon you saved (hint: they're lying to you, because they don't actually know). By using FouAnalytics, you can see which exchanges are active in your media buy, so you can eliminate most of them (e.g. from 35 to 3). You can move to an inclusion list approach to avoid most of the fraudulent sites and apps in programmatic channels. And you can block as many of the MFA sites as possible to reduce fraud and waste.
If you're not using FouAnalytics as analytics for your digital media, please let me know. I am happy to set you up with an account. FouAnalytics is free for small and medium businesses. I mainly charge large advertisers doing 1 billion or more ad impressions, in the form of an annual subscription, much like an annual Microsoft Office subscription.
Get in touch. For more background information, screen shots and examples, see the links below.
BACKGROUND INFO
Tour?of FouAnalytics Dashboard and Campaign Checklist
How?Site-Owners?Use FouAnalytics to Troubleshoot Bot Traffic
How to use FouAnalytics to Scrutinize Clicks from Programmatic Campaigns
"are you saying your tech is better and more accurate than the other verification vendors that have received 100s of millions of dollars in venture capital?" Yes.
Time to call BS on your fraud verification vendor(s)
Manager, Digital Auditing Services
1 年Pleasurable reading. Enjoyed mix of tongue-in-cheek and overt jabs. Thanks. Found one grammar thing..."But when special are appended to the url" might want to be stated as "But when special codes are appended to the url"
The Business Growth Locksmith | Connecting Home Movers To Service/Product Providers
1 年Brilliant overview - so many businesses today shouting about their ‘sustainability’ agendas and credentials yet very few consider the environmental damage associated with ad tech #justsaying