Earlier this week Goodway Group published a survey which tipped programmatic display ad prices to jump by as much as 20% over the next 12 months in a report entitled What’s Programmatic Costing You?
The report looked at some of the dynamics likely to alter the cost of display ads bought in such a way, with report authors noting that "it's time to reset your pricing expectations".
Report authors at Goodway Group have forecast that North American publishers taking advantage of improved yield optimization technologies such as the continued adoption of header bidding, plus a generic shift from fixed CPMs to more dynamic pricing models (or dCPMs) will fuel this trend. The Drum caught up with Jay Friedman, Goodway Group, chief operations officer (pictured below), to further explore his views on the market dynamics that will cause prices to increase so sharply in 2017.
In a Q&A session, he shares his opinions on why this cost increase is not necessarily a negative trend for media buyers, measures necessary to limit fraud in the industry, as well as some of the buy-side’s fears over yield optimization technologies.
The report says that 67% of display ads will be traded programmatically over the next year, with that in mind, it's also worth noting that adtech has historically been seen as a tool of the buy-side of the industry (i.e. to drive down rate cards, etc). Bearing all this in mind, can you provide comment on the dynamics at play which are going to cause the forecasted 20% increase in the price of display ads in 2017?
It’s 80% header bidding. Now that ‘the stack’ is flattened and all inventory is up for auction, buyers are finding very good quality inventory and bidding on it. Yet, because the ecosystem has moved to a first-price auction rather than second-price, buyers are still over-bidding, which drives up clearing prices.
In most market places, a 20% increase in the price of anything over a 12 month period would likely lead to some sort of reaction from those see-to-be adversely affected by such a market dynamic. This can take place in the form of advertisers tightening their belts (or making some other kind of expenditure sacrifice). Would you care to comment on what kind of reaction we might expect from the buy-side?
This assumes that advertisers are buying an impression because the impression is what they want. This isn’t the case. Smart advertisers are buying the impression as a means to a conversion. While we’ve seen big price increases, we have not seen commensurately declining cost-per-outcomes. In fact, in many cases, because we’re better able to differentiate and purchase better inventory now, we’ve seen improvements in costs per outcome.
If a marketer was used to spending $10bn, getting 100bn impressions, and 2bn conversions, I believe they can still spend the $10bn and get the 2bn conversions or actually do better. They’ll just do it with less impressions along the way.
How much do you expect industry-wide initiatives, such as TAG's Payment I.D. scheme (which is designed to make it harder for fraudsters to get away with siphoning money off from the industry) will have an effect on contributing to this 20% increase?
I don’t think this will help or do anything meaningful. The single biggest thing that we as an industry do right now is to have the buy- and sell-sides agree to one layer deeper of transparency with each other. The buy-side needs to pass multi-bid data to the SSPs (at least two bids per impression) and the sell-side needs to pass lost bid, and bid density data. I actually think this could bring us back to a second-price auction, although I don’t think prices will reverse course now that we’re bidding on the entire stack worth of inventory.
On that note, isn't it likely that fraudsters will likely find ways to game the system again? What in your mind is the best course of action against all of this?
I do agree that more money means more fraud, in general. However, audience-first buying combined with more direct to publisher buying will lessen the open exchange exposure which is how the fraudsters take root. There is plenty of fraud outside of the open exchange, but they need the OE to take root and grab users. I don’t see these as being all that related.
You also mentioned header bidding in your assessment of the overall findings, what will adtech companies (and media owners) have to convince the buy-side of the industry to adopt it? For instance, some indicate that inadequacies in the header bidding stack can lead to scenarios where buyers (that run multiple campaigns for multiple clients) can end up bidding against themselves, etc.
They’ve already adopted it whether they like it or not! It’s not the buy-side’s choice as to whether or not they take part in header bidding. If they’re buying programmatically, they’re buying in and competing against others in header bidding. The ‘bid against themselves’ fear has been around since the beginning of programmatic in 2006. Something being not false doesn’t mean it’s true. It means it’s unknown. Bid collision is statistically insignificant. I’m far more worried about getting in a car accident on a given day than the impact bid collision will have on my budget.
A full copy of What’s Programmatic Costing You? can be downloaded for free by clicking on the link