Monday, October 12, 2009
With a few high-profile exceptions, water always flows downhill. If you want to calculate which way a river flows, all you need to calculate is which end is on the high ground, and which on the low ground, and there's your answer.
Similarly, the flow of advertiser dollars from other channels into paid search platforms like Google AdWords has been predictable over time because of one factor: measurable performance when compared head-to-head with other channels.
In accounts where that condition hasn't been satisfied, companies don't increase budgets for paid search. They don't always shut off their accounts, though. So the end result is an account that sort of wanders along, hoping for the best, performing far below potential.
At different points in time, this state of affairs might benefit Google a lot or a little. The times when it benefited Google less were when their relevancy incentives (mostly CTR-based) were set in a dogmatic sort of way. If your account was really lazily and loosely targeted, eventually stuff would get "disabled" or "deactivated." That would kill volume for you, but you were probably doing horribly anyway, so it was actually a savings. Google, meanwhile, didn't get to exact as much of an "idiot tax" out of you. So they had to rely on other mechanisms, such as ego bidding or a hot economy, to further their profitability.
It's not quite like that now. Google has a sophisticated set of mechanisms for selectively allowing you to screw things up. They may not particularly like your fuzzily-targeted ads, but they're willing to show a certain percentage of them in certain verticals as long as your bids are sky high.
The end result of a low-volume, fuzzily-managed campaign, of course, is a poor ROI for the advertiser. But here's the curious thing. If the company has invested something in people and plans to run ads in this channel, they don't shut it off completely. They treat it like their other underperforming channels: shrug their shoulders, hope it gets better, turn the budget down a bit so it doesn't cause significant harm. From day one, I've suggested that this isn't a true savings, but a squandering of potential. "All in" or "all out" should be the advertiser's mentality. It is difficult, but not impossible, to optimize your campaign so that increasing the budget is feasible. If that's not happening, why run it at all?
The fuzzy, meandering, high-CPC-low-return campaign is not a negative scenario - short term - for Google's revenues and profits. These inefficient campaigns collectively spend heavily, smoothing out the ups and downs of the keyword auctions where advertisers are managing more tightly to customer responses. That's one of the reasons why, in the near term, Google's revenues will continue to rise gradually or at least be flat, avoiding the severe hits that other advertising media have taken during the recession. Look for proof of this with Google's upcoming Q3 2009 earnings release on October 15.
At Page Zero Media we inherit such accounts fairly often. There are two common targeting errors that companies make in the early days of planning AdWords campaigns:
(1) Loose and broad targeting, particularly of the type that addresses a mass market when you're going for a niche market. A company will say to themselves (perhaps having watched Conrad Hilton in Mad Men saying he wants to put a Hilton "on the moon"): "We want someday to be the biggest and best insurance company in the world!" But for now, they're just focusing on being top-of-mind in car insurance for high-risk drivers, especially those under the age of 25. Somewhere along the line, possibly in a text message from a golf banquet with impressive friends, the CEO forced the associate to the assistant marketing director to try broad matches for the words "insurance" and "car insurance" in the campaign. They're still eking along, in select markets, costing the company $18 per click whenever they are clicked, eating into the allotted budget for all the good stuff.
(2) Insider thinking. Companies get into their own jargon, right down into the regulatory mumbo-jumbo from arguments in Congress about all the players in their industry. The next thing you know, there are ad groups containing all the jargon about high-risk drivers -- jargon that's only ever come up in those Congress debates, in board meetings, and expensive consulting reports. People click on the ads occasionally, but they're inevitably just looking for information, not a car insurance vendor.
As these cardinal errors continue to eat away at the overall budget, the ROI for the whole campaign looks poor. The budget stays where it is. And the company concludes that the channel doesn't match the hype.
I got a note from a colleague who has spent years building out paid search campaigns that have formerly failed for the above reasons. Here's his tale of woe - the industry sector changed to protect the innocent:
If I can lend a "yes this is the way it works" rule to any PPC work for [insurance] that we should inscribe into our core being: Seth Godin needs to be read by every one of the people who do campaigns for them. Most of the time (probably CEO driven) they focus on the whole "getting their name out there" thing. Each new one I see is a broad, undisciplined mess. They constantly target information seekers with words like "best insurance" or "top insurance companies". The people who click those 1) Already have a vendor and are satisfied with them; and 2) Are simply looking for ideas; 3) In no way are targeted properly. The kicker is that those words are $5, $6, or $10 a click. They might as well spend that money on a TV commercial because that is akin to interruption marketing, PPC style. You can get fooled as well, because if someone has an article on top insurance companies on their site, like our client does, people will read it and stay. The bounce numbers tend to be pretty low, so if you are looking at that only, you get a false sense that the words are working. But when you look at signups, they are very close to 0%. I am not sure many of them do it, but content and blogs should be king for these sites, for those words; and they should spend that money on writing content. If you have and write articles based on all the key buzzwords like "car insurance comparison", you should get bang for that on organic search for $0. Example: "Best [(competing but ultimately very different type of) insurance]" and so on. Our new client spent $22,000 on those words via adwords. They got one person to sign up for an account. ONE! [Edit: we did more checking, and it might have been two.] Conversely: Targeted to their selling proposition of [doing more research on individuals in high risk insurance categories to offer them a break on rates if possible], those words [all having to do with insurance rates, unfair insurance rates, demographics of insurance rates] and so on, with a good ad, had $4000 spent and they got 20 people to sign up. OK, long story short: For financial services, not unlike others but especially for them, being targeted to exactly what you do if you are spending $4 a click or more is so vital it is not funny.
Google has worked themselves into a situation where the "idiot tax" helps Google's bottom line. So although they'd prefer it if most advertisers improved the relevancy of their advertising, the current system is built to hedge against idiocy. That being said, then, Google doesn't stand to gain a whole lot as advertisers become savvier and more efficient. In the past, the genius of the ad platform meant that Google's earnings and profits raced ahead faster than expected. During the recession, they're outperforming everyone else. But as the economy recovers, this current efficiency also means that you may not see Google grow as fast as you expect. They've wrung a lot of cash out of relatively wasteful advertising. Less wasteful practices will definitely help individual advertisers... a lot. They won't help Google nearly as much.
Labels: advertising budgets, goog, google adwords
Saturday, March 17, 2007
So Jim Hedger nicely commented on my recent interview with Richard Zwicky over at Enquisite. I have a little riff on Jim Collins' hedgehog concept, rolled in together with some statistical viewpoints about the overall size of the SEM spend in relation to marketing and advertising as a whole - something I try to convey in seminars, book, etc.
Jim mentions this as "injurious" and "potentially limiting". On the injurious front, especially since his name is Hedger, I'll take that as tongue in cheek, of course :) . But the idea that the sum total of SEM does indeed have its limits is important, in my view. Advertising spend outside our realm is indeed seemingly limitless. I think it's important to talk about limits, though. If all you can spend on relevant clicks in a year is $1,000,000 - even as a large company with a big budget - then that's all there is here. I think it's important to admit that. SEM is a large and growing field that does specific things for a large number of companies. It's approaching 50% of online ad spend, depending on how you measure. But it's finite, and the daily and weekly impact can indeed be quantified.
I do think we could do to dig in with a bit more research about the thus-far underquantified impact of search listings in driving offline revenues and choices, though. Especially when it comes to local and long tail research.
Labels: advertising budgets, hedgehog concept
View Posts by Category