How can Machine Learning be applied to Ad Ops? Talk with Chris Reid of Sortable
This week on the podcast, we talk with Chris Reid, who is the founder of Sortable, which uses machine learning and data science to optimize publisher advertising. Sortable is different than most monetization companies because they have build a large, back-end infrastructure to programmatically analyze users and show them the ad networks that will be most profitable to the publisher.
This was fascinating conversation for me, as I have long thought that many industries will head in this direction. The technology that Chris and his team have built allows them to massively out perform traditional optimization. They are seeing huge growth in their publisher base, and this episode should give you an idea as to why.
1:25 — What is Sortable’s approach to Ad Ops and Monetization
1:58 — What questions do you ask when determining how to monetize a site
2:31 — Examples of nuanced questions to ask a business
3:25 — How do these strategies change your approach?
4:19 — Assuming a site wanted a higher ad quality experience for users, what would you do?
4:48 — Controlling for bad inventory in ad networks
5:45 — Does Sortable operate sites or is it just client based?
6:15 — How does the work that you do with your clients inform how you operate your own sites?
7:15 — What was the major take away from you last year of working in this space?
8:17 — Dynamic Monetization
9:16 — Determining the value of users
10:13 — How did you go about building the infrastructure to value users at a granular level?
11:25 — Where in the optimization process are you applying this machine learning?
12:54 — Where do you see the affect of this analysis?
13:29 — How this approach is different than most monetization shops?
17:50 — Where does your technology interact with publisher technology?
19:55 — How does your tag system work? Are you running a real time auction?
20:50 — How machine learning is used to select networks