The launch of Google AdWords and its subsequent rapid growth in the early 2000s helped spawn an entire industry around paid search advertising. The ecosystem that arose includes Google and its advertisers, as well as a slew of agencies, consultants, software platforms, data providers, third party tools and various combinations thereof in between them.
Almost every change or announcement that Google makes seems to raise the question of which of these middlemen Google may be trying to disintermediate.
When it was recently announced that Google would stop passing search query data via referring URL for paid search clicks, Marin Software was prompted to address whether this posed a serious threat to their management platform business.
They rightly responded that it “would not be that big a deal,” a sentiment RKG’s George Michie also expressed in saying, “This is an annoyance, but in the great scheme of things, it’s not a major problem.”
Just last week, Google’s “Step Inside AdWords” presentation announced a number of new features, including a set of “enterprise-class” reporting, bidding and management tools. This raised similar questions, but unless you were already losing sleep thinking about Google cutting you out of the loop, you’re probably not going to start waking up in cold sweats over this one.
Focusing on digital agencies and bid management platform providers, let’s examine some of the reasons why Google isn’t likely to put us all out jobs soon.
When Enhanced Campaigns was first announced, it was touted as simplifying the management of AdWords campaigns across devices.
While it may have done that (particularly for less-sophisticated advertisers), it also opened up a range of new possibilities for how advanced marketers could take advantage of existing segmentation options such as geography, device, audience lists and time-of-day, and it created a framework upon which additional levels and layers of complexity could be added in the future.
Under the legacy AdWords model, segmenting bids for facets of user context like device and geography meant duplicating campaigns with different settings. If we wanted to set different bids based on mobile vs. desktop, the US state the searcher is in and whether or not the searcher is a returning customer, we would have quickly multiplied the number of necessary campaigns by 200 times.
If we wanted to account for more granular geographies like city or zip code level, or even a radius around a set of locations, the number of campaigns we would need under the old AdWords model could increase by orders of magnitude. This was not manageable, so advertisers had to pick and choose a limited number of high-priority aspects of user context to account for.
With Enhanced Campaigns, advertisers can now act upon much more of the information we have at our disposal. If we want to adjust bids for a thousand different locations and multiple devices, we can easily add those as modifiers. But, this also introduces the complexity of calculating how all of those modifiers should be set when our performance data is spread so thinly.
This is a problem that agencies and bid management platforms have been working on and solving for over a decade, particularly with respect to predicting the performance of low-traffic keywords, an area with many parallels to predicting the importance of pieces of user context.
Google’s free AdWords bidding tools have only recently even allowed for adjustments based on the value of conversions, and they do not approach the sophistication of the better bid platforms out there, which incorporate many data signals that go well beyond account structure.
While Google could task a team of engineers tomorrow with bringing its free bidding tools up to speed with the best enterprise options, they could have done so a year ago, or five years ago. They don’t seem to be in a rush.
Were they to do so, that would just be one piece of the puzzle though, since even the best tools require smart people at the controls.
It seems much more likely that Google could seriously threaten the paid search software business than the management side. Solving problems algorithmically is in Google’s DNA, but a lot of the work that goes into managing paid search well involves the types of problems where human beings still hold the advantage over computers.
At its core, paid search management is about language, whether it is building out keywords, writing succinct ad copy that will stand out and appeal to users, or even tailoring a feed for Product Listing Ads. A software program, like Google’s Keyword Planner, can scan a URL or crawl an entire domain and generate an extensive list of potential keywords. Most will be good, some will be odd but harmless; but almost invariably, a subset will be downright dangerous.
An analyst can scan a web page and see a product such as an “electric guitar with free amp” and wouldn’t think to run the phrase “free guitars” on broad match. An algorithm just might. That’s a blunt example, but the necessity of understanding the nuances and meaning of language (and using that knowledge predictively) crops up time and again in paid search management. It can be very expensive to drop a computer-generated term list into AdWords, sight unseen, and hope our bidding algorithm will quickly separate the wheat from the chaff.
Even with bidding, human insight and oversight is invaluable. Bidding software is only as good as it’s written, but its effectiveness also depends heavily on the quality of the inputs we feed it.
Does your bidding platform know that Cyber Monday will be December 1st this year, instead of the 2nd like last year? Hopefully, but the paid search analyst sure will. Does your platform know that a product you sell was just touted by a big celebrity and its conversion rate will soar through the roof? Unlikely.
With paid search, or really any marketing technique, we also have to scrutinize the larger strategic questions that no software is going to answer for us. For instance, should we weigh customer lifetime value more heavily, or do we need to see a more immediate return on our investment? Should we steer our mobile customers to our app even if we take a short-term hit on conversions?
Google may be able to help in this area and the more day-to-day examples above, but it hasn’t been a core strength for them and it would mark a pretty major shift in focus.
Somewhat obviously, even if Google could provide top-notch, free AdWords bidding tools and maybe even manage them as well, advertisers would still need to manage Bing and Yahoo search along with a host of other channels.
Would Google be willing to provide a free tool that worked with Bing Ads? Tough to say. Would advertisers trust Google to allocate spending appropriately across engines? Most probably would, but not all.
Advertisers also use platforms and employ agencies to manage and report on multiple other channels. AdWords may be the single biggest line item in a digital marketer’s expense report, but we also have other display efforts, social, organic search, and CSEs to account for, and typically an attribution scheme that ties them all together.
Google is a huge company full of smart people; it can do pretty much what it chooses. If they don’t at some point make a big push to cut out some of the middlemen in the AdWords ecosystem, it may speak more to the economics of it all than anything else.
In 2012, the top 233 US search agencies covered by Ad Age’s 2013 Agency Report generated combined revenues of $1.03B. Google generated $21B in US revenues in 2012. Those agencies’ revenues amounted to about 5% of Google’s, which might seem pretty efficient to Google compared to trying to scale up a massive new service delivery business. Rather, they can count on the existing market to produce robust paid search programs and focus instead on their own strengths.
Also, although a reputable agency is not going to push their clients to spend more than they should, the interests of agencies and Google are well-aligned and geared toward seeking out opportunities for growth.
On the management platform side, it’s almost surprising that Google isn’t doing more, but then again, it may just be a tough business. In 2013, Marin Software had revenues of $77.3M dollars, but a net loss of $35.9M. If Google invested more in this space, they would do so to get advertisers to spend more on ads, but Google may again be content to let the market handle it rather than make a large investment themselves.
There’s also no guarantee that advertisers would drop their current tool, thus freeing up budget, even if Google had a comparable free option. Google Analytics is a great tool, but from my experience, most enterprise-level advertisers still use a paid option.
To close, Google’s official headcount stood at a little over 46 thousand in Q1 2014, but the livelihoods of many more are affected by the moves it makes. This includes those of us intermediaries and facilitators in the AdWords ecosystem, as well as the clients we work with. I don’t read recent events as suggesting any tectonic shift is happening soon, but I could be wrong. Thoughts?