The last few weeks have seen plenty of interesting announcements from our esteemed competitors, Yahoo! and Google. But two that you might not have paid so much attention to are are Yahoo’s launch of AMP! (those guys have got to get over this whole silly exclamation point thing), and Google’s new AdManager product. Both are aimed at making it easier for publishers to sell their ad inventory, though both are part of a bigger story.
Sometimes it seems like publisher ad serving is the red-headed stepchild of the industry – frowsty, boring technology used by publishers and ad sales houses to push inventory around from pillar to post in an attempt to get it off the publisher’s hands before it goes stale. But publisher ad serving has one killer attribute – whoever provides the primary ad tag on a publisher’s site can control (or at least influence) how that publisher’s inventory is monetized. Control that tag, and you’re the nightclub bouncer, magnanimously letting the pretty girls in whilst keeping the spotty boys out. If you’re just another ad network serving ads into that ad unit, you’re the spotty boy hanging about the back door. What’s more, whoever owns the primary ad tag gets to see all the traffic data, a treasure trove of information for anyone trying to build a behaviorally targeted ad network.
So, since Google and Yahoo now both have significant ad network operations (AdSense and YPN), it makes sense for both of them to offer free ad serving technologies to publishers. How do they stack up? Should they even be compared?
Yahoo has been working on a new ad platform/exchange for some time, under the title of Project Apex. Some noise was made about Apex shortly after Jerry Yang took over Yahoo last year. The offering is aimed at publishers, and purports to offer a radically simplified/more powerful way of managing and monetizing online ad inventory. Reaction to the announcement has been mixed; the Yahoo Publisher Network (YPN), which AMP is seen as an evolution of, has been dogged with editorial quality issues, affecting advertiser ROI and driving advertisers into the arms of Google. So a certain amount of skepticism attends this latest announcement.
There’s relatively little publicly available information to help differentiate AMP from YPN (and Yahoo aren’t likely to tell me much for the time being). The main theme seems to be that AMP is a much more ‘open’ platform than YPN, meaning that publishers can use it to direct inventory to any third-party ad network or advertiser, without having to have Yahoo ‘represent’ (i.e. take a cut of the value of) their media. The video preview of the technology also makes much of a publisher’s ability to cross-sell other publishers’ inventory as part of a negotiated sale, by tapping into Yahoo’s Newspaper Alliance network.
Additionally, there’s talk of an API, enabling ad networks to integrate with the platform. In this latter respect it’s hard to see a clear difference between this offering and the Right Media Exchange (RMX) – in fact, a previous working title for AMP was the Advertiser/Publisher Exchange.
It’s too early to say whether AMP is actually any good, though the demo looks pretty and appealing – the principle danger is that Yahoo is re-re-inventing the wheel here, potentially ignoring the YPN and RMX code bases and coding something new from scratch. And the effort does seem to be squarely in the negotiated space – i.e. the sort of inventory that is Yahoo’s bread and butter; we can count on Yahoo knowing a thing or two about this kind of inventory, but I wonder how it’s going to address the highest-growth part of online ad sales – the tail.
In September, Jerry Yang said that AMP would take three years, which seems to imply something of this nature, and which caused gasps from the likes of ValleyWag at the time. But now Yahoo says they’ll have something out during the summer. Presumably they’ll have a lot more to add even then.
A couple of weeks before the Yahoo announcement, Google announced the beta of Google AdManager. The timing of Google’s announcement was significant (or at least an unfortunate coincidence) because it almost exactly coincided with the completion of Google’s acquisition of DoubleClick. AdManager (which shares a name with a Microsoft product, Atlas AdManager, which I believe is the subject of some discussion of a legal nature between our two companies) is positioned as an easy-to-use, self-serve inventory management tool, allowing publishers to manage their reserved and discretionary inventory, and select ad networks for the latter (including, crucially, Google AdSense).
The motivation behind AdManager (and AMP) can be seen in a quote from MyYearbook.com on the AdManager website (my highlighting):
“The AdSense integration feature has been helpful with helping us optimize our remnant inventory. There have been numerous days where we made more money because Ad Manager was able to auto adjust and send AdSense more inventory.”
It seems like the primary positioning of AdManager is as a trade-up from AdSense – for publishers who have been using AdSense and perhaps another ad server (or doing something clunky like serving ads manually), AdManager represents the next step up, as this comment from this forum thread shows:
“What we used to do before is, insert the adsense code or any other affiliate banner code in the adslot MANUALLY! If you want to insert adbrite code there instead of adsense code, we will edit the code and upload it right?? OR some people also use other PAID third party hosted ad servers/their own ad servers!.”
The interesting thing about this announcement (and the reason the timing seems so unfortunate) is that DoubleClick already has a publisher ad server on its books – DART for Publishers (DFP). DFP ain’t no slouch – it’s the market leader, with about a 60% share (our own Atlas AdManager, by contrast, has about a 10% share amongst publishers). But Google AdManager and DFP fulfil very similar roles – helping publishers to define, predict and sell online inventory. So it will be interesting to see how this situation plays out now that the DoubleClick acquisition is closed.
What’s also interesting about this product is what it says about Google’s innate attitude to online advertising – that it is a process that should be simple and as automated as possible.
Compare and Contrast
So the bottom line is that these are two quite different products, aimed at solving a different set of problems. But these two products paint a fascinating picture of Yahoo and Google’s attitudes to online ad sales.
Yahoo seems to be thinking of itself as a semi-neutral ad network/exchange, enabling its publisher partners to build and sell their own on- and off-network inventory packages and then clear/settle them through Yahoo’s infrastructure. Even Yahoo’s Newspaper Alliance is itself a member of a larger network, quadrantONE. AMP will sca
le well at the top-end of the market, but its features will mean little to smaller publishers who just want to monetize remnant inventory simply.
Google, on the other hand, seems to view itself as the uber-network. This is understandable, since it has such a huge network of publishers via AdSense. AdManager seems primarily to be aimed at increasing the amount of inventory that is made available to Google’s network, as opposed to allowing publishers to build their own mini-networks – publishers are very much expected to be the end-nodes, in Google’s worldview. Google AdManager will scale well into the tail, but lacks the sophisticated required by larger publishers.
To be fair, Google does have more convincingly third-party sell-side assets in DoubleClick (including the DoubleClick Exchange), whilst Yahoo has its own network in the form of YPN, so these positionings are not exclusive, but I think they are representative of the attitude that both companies bring to this business.