Storytime : Deconstruct to the Core Building Blocks
Storytime. Case studies. Real-life examples. Scientists are discovering that chemicals like cortisol, dopamine and oxytocin are released in the brain when we’re told a story. Stories are what make a point stick.
🎯 Mental model for this musing : Deconstruct to the Core Building Blocks
Storytime!
Two stories where first principles have led to deeper discovery and stronger outcomes :
- Idea to Execution : Audio Ad Analytics 🎧
- Data Analytics : Bot Activity across Online Ads🤖
Building Up | Audio Ad Analytics 🎧
What is it?
Advertising has helped sustain the ecosystem of journalism, storytelling, and creative process online. The measurement of ad analytics by an unbiased, third party source helps create trust and transparency in that transaction between advertiser and media partner, allowing for increased investment into that medium. At the time of this initiative, the ad industry had a sophisticated measurement solution for display and video ads, but there wasn’t any reliable measurement for audio ads. Yet.
Super high level, but let’s break down how we got there. Think : first principles.
Tell me a story :
Audio is so much more than just video without screens — it’s an incredibly powerful tool of connection.
In 2019, we began to explore what it would take to build a similar measurement solution for audio. This involved deconstructing the entire workflow of how ads are served within music streaming and podcasts, identifying the players in the ecosystem and their existing and potential roles in the audio ad serving ecosystem, and scoping out the current capabilities of each audio platform and tech vendor. This also involved an extensive customer research and discovery process to understand what metrics would be valuable for audio ad analytics, along with a hyperrealistic evaluation of development resources on both sides. This was turning out to be a beast of its own.
A key part of building high quality products and solutions is to become customer obsessed, while remaining customer friendly. This means deeply understanding the customers needs, while also understanding their limitations and motivations to meet those needs. Our audio platform customers did not have the luxury to build an integration from scratch, and to be honest, neither did our team. Build, buy, partner? It was time to make some decisions.
Building up : As a first step to figure out how we could enable ad measurement across audio ads within mobile app environments, we looked at how measurement was being done across video ads.
At the time, we turned to the IAB Open Measurement SDK, a collaboration across the industry to enable display and video measurement across mobile apps — so our team and others actively participated in working groups to expand capability to include audio measurement. Using the RICE model, we prioritized technical paths and key partners to integrate with as a start. Next, at high level, we laid out each of the existing video metrics and remodeled them for audio signals. Example : removing visual components in the methodology. Then, we analyzed the workings of our video measurement tag, and modified the code for an audio environment.
Fast-forward to the end : We launched a first-to-market solution through thoughtful collaboration, and quickly became trusted industry leaders.
Epilogue :
Using first principles, we not only took this from conception to revenue within 12-months, using minimal development resources, but also paved the way for additional innovation across the audio.
Data Analytics | Bot Activity across Online Ads 🤖
What is it?
Bots. Ad fraud. It’s the wild west out there! According to a study conducted by the University of Baltimore, ad fraud cost businesses an estimated loss of $35 billion globally in 2020 alone. Yikes! So yeah, it’s a pretty serious problem.
Some (very) high level background for those not familiar with the digital advertising world : Advertisers will generally run online ad campaigns across hundreds of publisher sites. So if you see an ad when you’re looking at recipes on one site and then the same ad when you’re reading the news on another site, it’s all part of that media buying strategy. For advertisers, it’s incredibly important to know their ad dollars are being efficiently spent, and their ads are actually reaching their target customers online — not a bunch of bots who will never buy their products or services. For publishers, these ad dollars are what helps them keep the lights on (aka pay the bills), and give us readers access to content for free. This is where ad analytics and measurement platforms come in to gather valuable, trustworthy data around advertising online, and ultimately help the ecosystem function more smoothly. Let’s dig into a specific example.
Tell me a story :
An advertiser noticed their Invalid Traffic Rate (the metric used to measure bots and malicious intent activity online) spike for one of their campaigns. At reviewing their data, they were able to see the bot activity was coming from one particular site — a well-known highly regarded premium publisher. Instead of relying only on the aggregate level data which would result in the advertiser canceling the entire media buy with that publisher partner, we were able to bring both parties together to drill down further and further into the data until we were at the most granular level. We found that there was only one creative ad in the overall ad campaign that was impacted, it was impacted on a very particular section of the site, across a specific browser type in a specific kind of mobile device, and was coming from a particular source of invalid traffic that makes up the aggregate number. The publisher acted quickly by isolating and targeting the specific problem, as opposed to making decisions on the higher level number. The advertiser implemented a similar process to work more productively with their publisher network moving forward, and this intelligence became a key part of how other more automated solutions were built.
Epilogue :
No ad dollars lost for the publisher, no premium audience lost for the advertiser, and no babies thrown out with the bathwater. A happy ending!
⚛️ Molecules are made up of atoms. The complex whole is made up of simpler, smaller parts. Deconstruct to the most granular elements, the core building blocks, the indubitable truths — and rebuild, reinvent, and reimagine from there.
💡For more on mental models : The Art of Problem Solving