When Juice Mobile first started securing premium mobile inventory for its clients, it had the same problem as everyone else. Advertisers weren’t willing to spend a large amount with any one publisher, and publishers didn’t want to play ball unless there was enough money on the table.
“The part that drove me the most nuts was, I had $50,000 in mobile, and I’d call publishers all the time and they’d hem and haw,” said Neil Sweeney, founder and CEO of Juice Mobile. He would try and split a client’s budget and expose $10,000 to each publisher – but that wasn’t usually enough to get them interested.
At the same time, he’d have to find which publishers in the right regions had the kind of inventory the client wanted – display, native, video – and whether it was the right size, shape, format, position, etc. For direct buys, the only way to get those specs was with phone calls. A lot of them.
“The agency was in a situation where they had to do a manual workflow for 10-15 different publications,” said Sweeney. He recalled thinking, “This is crazy, it’s so inefficient.”
Ad exchanges weren’t much help. The inventory Sweeney’s clients were looking for was high-end, the kind that publishers reserve for clients they have a relationship with. In mobile, most publishers still only sell remnant inventory on real-time exchanges – ad spots they haven’t managed to sell off any other way.
So Juice came up with a solution for discovering premium mobile inventory that would automate a lot of the mobile buying process, but stop short of letting machines make the actual decisions on which inventory to buy.
The resulting platform, Nectar, launched in spring 2013 in the Canadian and U.S. markets. In some ways, it’s similar to automated direct technology similar to that used by Centro and AppNexus’ Twixt, which basically automate the process of direct sales. But in building the platform, Juice applied a few key insights from programmatic, and came up with a new way of thinking about how buyers and sellers transact media. It turns out it’s not just a completely fresh take on ad sales – it also could solve a lot of problems that both advertisers and publishers have with programmatic.
Enter Nectar
Here’s how it works: The buyer sets up an order in Nectar, with the specifications for the kind of mobile impressions they want – format, geographical location, time of day, etc. – and the amount they’re interested in paying. Nectar then reports back any sellers in the region that have inventory meeting those criteria. The buyer picks the sellers they’re interested in, and the platform sends them an order that looks a lot like an RFP.
Next, the sellers submit rate proposals. This is where the insights from programmatic come in: Nectar uses data on all the sellers involved to give everyone participating in the RFP an idea of how they stack up to the competition. Though each of them is blind to who the other participants are, they can see how the inventory compares to the average performance level of the pool, based on historical CPM, click-through and conversion rates. Based on those stats, the system suggests the best rate they can offer to appropriately value their wares, and they can either let the system automatically submit that price or manually choose a competitive price they want to offer. Once all the offers are submitted, the buyer manually chooses how to distribute the buy between the sellers – they can split it up between all sellers equally, or choose the one they like the best and transact 100% of the buy with them.
The key strength of the platform is that at every decision point, both the buyer and seller get to make the decisions manually, rather than having a programmatic algorithm decide it for them. Publishers get to see how much money is on the table, and decide exactly how they want to price their inventory to get a piece of it. Buyers get to pick the sites they advertise on, rather than trusting algorithms to find the right kind of premium inventory.
Publishers that have used the platform said knowing what they’re up against is a huge problem in programmatic – one that Nectar neatly solves. “They provide real-time tools to the publishers in Nectar to make better informed decisions about how to release our inventory to select buyers,” said Sandra MacKechnie, president of digital advertising and strategy at Postmedia. “Nobody else is doing this right now – so that’s why it was a really compelling offering.”
Many publishers fear that placing their premium inventory on an exchange will devalue it – if people aren’t willing to bid high, the inventory will sell for very little, and over time that will drag down their all-important CPMs. On Nectar, publishers have the control to set their own prices, similar to direct sales – except they have a better understanding of the market, and they don’t have to worry about pricing their inventory too high to interest a buyer.
What’s more, Nectar rewards good performance: the more your inventory draws engaged audiences, the higher you can price it in the future. Juice boasts that the total monthly revenue drawn by all publishers on the platform has grown more than 300% in the last year, and that some publishers are seeing 4-5 times their initial earnings. MacKechnie said she has seen her CPMs increase significantly, and float substantially above exchange rates, since Postmedia got on board.
On the buy-side, the platform solves a notorious problem in mobile – incompletely filled orders. Going into a programmatic buy involves a lot of uncertainty, and one of the most basic risks is that other buyers will outbid you on all the inventory you want – resulting in an campaign that runs fewer than the desired amount of impressions. Low fill rates are a challenge particularly in mobile, where there are fewer exchanges and fewer auctions to bid on.
With a Nectar buy, the buyer locks down the inventory they want well ahead of schedule, so it’s not even presented to other bidders – just like an up-front direct buy. It’s a way to ensure that what you pay for is what you get.
The catch: It’s not real-time
What buyers sacrifice to have that control and certainty about where their dollars are going is the real-time aspect of programmatic. Because the impressions are bought before hand, they can’t tell who’s actually looking at the impression they’ve bought. Being able to exclusively buy impressions that are shown to a particular audience audience is one of the big draws of programmatic, and one that Nectar doesn’t benefit from.
Sweeney says that buying audiences doesn’t make sense on mobile for several reasons. For one, mobile is a far less data-rich environment, with fewer opportunities to track and identify audiences in the first place. Mobile applications don’t have cookies, and alternatives for collecting data on users are stymied by the fragmentation of the various mobile devices, operating systems and vendor SDKs that have to be taken into account. So a lot of the targeting data that makes real-time audience buying possible just isn’t available.
But there’s a deeper conceptual reason that Nectar isn’t based on finding real-time audiences in remnant inventory. Brands like Coke or Ford don’t just want to be buying a particular audience, wherever they are – they want to be buying a particular audience, on a quality publisher’s site, Sweeney says.
“If you ask most people why they’re buying real-time in the first place – are they buying real-time because it’s in the best interest of their client? Or because it’s easy, it’s efficient, it allows them to reclaim margin?” he asks. The engineers that have built solutions for programmatic buying have largely focused on price and scale – but have neglected environment, he says. Advertisers do care whether their ads run on The New York Times or JoBloBlog.com, even if exchanges and DSPs don’t. Audience and data still certainly matter, but the context surrounding an ad shouldn’t be overlooked.
In Canada, the vast majority of digital inventory is still bought direct from publishers. Premium advertisers that are buying programmatic are largely doing so on private exchanges, or by integrating directly with individual publishers’ ad servers to buy their inventory in real-time on a preferred basis. Although there are certainly exceptions, many Canadian premium advertisers prefer buying quality media on a guaranteed basis, rather than using data to track their audiences into the dark corners of the web (where they can’t necessarily follow).
Sweeney stresses that Nectar isn’t an alternative to programmatic (or direct) – it’s an add-on. “What we’re saying here is, there’s different ways to buy for different kinds of client,” he says. He doesn’t see any client on the publisher or agency side abandoning either programmatic or direct to work solely on Nectar; rather, they’ll all work together for different aspects of a single buy.
So far, that’s how publishers like Postmedia and Pelmorex are using Nectar. The platform is so new, and so unlike either direct or programmatic channels, that buyers are still taking their time to figure out how it best fits into their media plans. MacKechnie at Postmedia says that she’s seen some very big multinational brands buying ads on the platform, but so far it is definitely not enough to fuel Postmedia’s digital ad sales on its own.
A saturated market?
Bringing advertisers and agencies on board yet another new-fangled buying platform is no doubt going to be tough, and involve some aggressive evangelism. But Juice has a few things backing it up. For one, Nectar isn’t Juice’s entire business – it has a programmatic demand-side platform, Swarm – so the company doesn’t stand or fall on the success of one product.
And since Juice already has working relationships with major Canadian and U.S. publishers, they’ve been a lot more willing to try out Nectar. After just a year and a half in-market, Nectar already boasts 5.5 billion available monthly impressions – putting it in the ballpark of of a mid-size display ad network.
There’s also something strikingly intuitive about the Nectar platform. In television, advertisers vie for scarce publisher inventory, so it makes sense for them to try and outbid each other for prime-time commercial spots. But in digital, inventory is plenty, and ads are scarce – which should really mean that publishers are the ones competing to win over advertisers to fill their spots. That’s why Juice’s reverse-auction model, where publishers pitch advertisers, makes a lot of sense – it gives publishers a chance to make a case for a fair market price, based on the quality of their product, and it makes it easier for buyers to get the value they want. As Sweeney points out, both buyers and sellers end up with more control, and less risk.
In principle, there’s nothing stopping Juice from developing a similar platform for desktop advertising. But Sweeney says there’s a good reason he’s not looking into it: “Desktop is dying.”
“We treat desktop like we treat radio,” he explains. “It’s legacy media, that the consumer is porting away from. … Look at every major media company out there. Look at Facebook, Twitter, Instagram, Snapchat. Everything that’s disrupting in the media space, whether it started as mobile or its now calling itself mobile-first – it’s entirely pivoted to mobile.”
