#63: Quo Vadis Green Shoot AdTech Interviews
With Rocky Moss, Co-founder and CEO of DeepSee.io
Reading Time: 7 minutes of seeing through the fog.
Welcome back to Quo Vadis Green Shoot AdTech Interviews.
In this series, we’re stepping into the shoes of “AdTech Equalizers” — the green shoot companies changing how the advertising job gets done. We delve into their worldview and unveil new strategies, tools, and thinking that make advertisers (and investors) better off.
Today we’re talking with Rocky Moss, co-founder and CEO of DeepSee, a leading provider of specialized analytics to maximize the effectiveness and revenue potential of advertising content.
Q: The MFA criteria recently put forth by the ANA, that DeepSee and Jounce helped devise, sounds like it needs to provide specific thresholds that would constitute a High, Medium, or Low MFA risk. Do you think that’s a problem?
A: Rocky Moss
I think the main issue with creating a simple grading rubric for MFA is that when we say “MFA,” we’re actually referring to a bunch of different behaviors that the industry has collectively decided to address at once. There’s a lot of baggage under that label, ranging from sites that buy 99% of visits, to sites that use AI to aggressively generate articles that game organic search to take visits from competitors.
I’m not sure it’s entirely useful to discuss so many disparate low-quality signals under one label, but I can’t argue with how effective it has been in reducing wasted ad spending globally. In terms of global impact on the bidstream, what’s now called MFA has a way bigger effect even than fraud, and I think that’s why we see legacy vendors like IAS and DV pivoting to roll something out for MFA detection (even though there’s not really a basis within MRC IVT guidelines).
Q: How did you get involved in the battle against MFA?
A: Rocky Moss
I got involved in this discussion a few years back when DeepSee published research on some of the worst arbitrage that we’d ever observed. That research got Chris @ Jounce and I talking about traffic sourcing’s role in the scheme, and that led us towards co-researching traffic sourcing for one of their monthly research wrap-ups.
I think we mutually would have been happy just calling out arbitrage as “arbitrage”, and moving on, but our research and others over the past few years have really pushed a long-standing frustration into the open. This low-value-high-reach clickbait has existed for years, and I think conflicting incentives are the main reason it hasn’t been effectively taken to task in the past. DeepSee and Jounce are independent, so we can speak plainly about what we’ve factually observed in our partners’ data.
As to the signals we helped the ANA develop, they are the culmination of all the above-mentioned research, across many different types of sites that don’t perform well for advertisers. I think the biggest benefit is that we can now start speaking the same language when it comes to why something is flagged as MFA. When a site gets blocked by an SSP / DSP due to perceived MFA status, the "why" of that decision can be communicated in relation to a concise set of metrics that are not siloed by any one measurement org. That’s key.
Q: So we have these signals that are often associated with poor ROI. Are we expecting buyers to determine their own acceptable thresholds for what constitutes MFA? Where can advertisers even get data to start modeling a solution?
A: Rocky Moss
I think buyers, or at least platforms, should indeed set their own policies around the metrics that matter to them. Any single one of the criteria we put forth could be reason enough for an advertiser or platform to block a site if severe enough. Buyers have long been curating blocklists full of sites they prefer not to run on, and this is simply an extension of that process.
Regarding the second question about who they trust for these metrics and determinations (supply partners or verification vendors), I imagine life will continue much as it has — platforms deciding whether it's worth it to in-house that measurement or go with a 3rd party vendor. The same types of inventory-quality decisions are currently made based on SIVT measurement, which is often a black box and completely unique to each measurement vendor so buyers are no strangers to these kinds of decisions.
We could obviously pitch ourselves as the source of that data, but I think it’s more important to get orgs thinking about these things themselves. Shop around, find the best source for you, and if you can’t find those metrics from existing media rating partners, push them to develop the capability! We want all boats to float as we raise the quality water level.
Q: Publishers have had it rough recently at the hands of brand-safety & anti-fraud tech. Forums are littered with stories about precipitous revenue drops caused by blackbox IVT tech from IAS, DV, MOAT or the like. How is what you’re doing any different, and what can publishers do to feel confident they won’t be flagged?
A: Rocky Moss
Firstly, Jounce and DeepSee have both committed to transparency around flagging status to any publisher who asks. I don’t want to be yet another palm to grease in the AdTech mafia. Publishing the criteria that we’re looking for already sets us apart from black-box IVT tech while the lack of specific thresholds actually makes it a bit harder for grey-hats to toe up to the line and not cross it.
The best way to not be flagged as MFA is to do what great publishers have always done—produce unique & compelling content, written by real people, and with monetization practices that respect the user experience.
Q: The reality for some publishers is that they must buy more traffic than they did in years prior to meet the same visitor quotas, does this mean they’d be flagged as MFA? What can be done to remove the stigma around paid traffic?
A: Rocky Moss
Changing the stigma around paid traffic, at this point, should very much be driven by publishers with evidence of how paid traffic drives outcomes for advertisers that are equivalent to or exceed organic traffic. This is a huge missing piece of the conversation. DeepSee and Jounce consider high rates of paid traffic sourcing a key indicator of MFA sites, simply because it's clear that this signal correlates with worse ROI whenever this data is made available to us. After countless audits of our partners’ data, it would take actual case studies to change minds on that matter.
Ideologically, I see nothing wrong with paid traffic to ad-monetized pages, as long as the landing pages act the same when visited via paid & organic channels. Manipulating content formats based on paid visits is a hallmark of a crap site. But we're realists, so we understand that great pubs sometimes have to buy traffic to meet quotas. If pubs are confident that this traffic performs at the same level as organic visits, then we should see publisher trade groups aggressively pumping out transparent studies that support this view. If they're not, why even talk about this?
One thing for sure, it’s definitely true that the way people find content is changing. Change is the only constant we can expect. We will continue to measure norms across thousands of publishers and when we see anomalies we’ll flag them as such. If the new norm becomes X% paid traffic, then the models will reflect that. As it stands today, buying almost all your traffic remains an anomalous behavior.
Q: Is there any hope for the long tail? What can new sites do to get noticed, without being flagged as MFA?
A: Rocky Moss
In the long tail, the uniqueness of the content is really important. For new publishers my advice is simple: Dedicate entire pages to your authors, tell your story! If you really want to stand out, you first should convince us you are a person! In the absence of historical advertising performance that we can look to, I’d advise supply-management teams to start by looking at the believability/authority of the author themselves based on links in bios & reverse image searches of various site assets like author headshots.
It’s never been easier to make a site that simply repurposes existing content vis-a-vis AI. For buyers, it’s easy to be jaded and say “there are ENOUGH websites out there, therefore I’m just going to stick to my target list of the Comscore top 500/1000,” but the internet is dead if we give into that impulse.
Ultimately, I want to see independent voices be elevated and be financially rewarded for their efforts. We’re definitely in an existential place as an industry, an inflection point where we’re either going to optimize towards who can create the most sites with the most seemingly accurate information (probably with generative AI), or we’ll find a way to help real humans monetize their labors of love in the long term. Companies like Mediavine and Raptive are definitely a key part of elevating independent voices, but even they likely struggle with maintaining quality checks across so many sites.
DeepSee Research You Might Like To Check Out
DeepSee.io improves ROI Compared With Audience Targeting Alone
DeepStreamer: Illegal movie streaming platforms hide lucrative ad fraud operation
Rewarded Traffic: The Inorganic User Engine Driving Ad Campaigns on Major Websites and Podcasts
A Case Study in Monetizing Piracy: MangaOwl and Chessmoba.us
Arbitrage Sites Manipulate Metrics Using Misleading Content Formats
Ask Us Anything (About Programmatic)
If you are confused about something, a bunch of other folks are probably confused about the same exact thing. So here’s a no-judgment way to learn more about the programmatic ad world. Ask us anything about the wide world of programmatic, and we’ll select a few questions to answer in our next newsletter.
Join Our Growing Quo Vadis Community
Was this email forwarded to you? Sign up for our monthly newsletter here.
Get Quo Vadis+
When you join our paid subscription, you get at least one new tool every month that will help you make better decisions about programmatic ad strategy.
Off-the-beaten-path models and analysis of publicly traded programmatic companies.
Frameworks to disentangle supply chain cost into radical transparency.
Practical campaign use cases for rapid testing and learning.
Disclaimer: This post, and any other post from Quo Vadis, should not be considered investment advice. This content is for informational purposes only. You should not construe this information, or any other material from Quo Vadis, as investment, financial, or any other form of advice.