Just how excited should we be about new paid search interface features? At least as excited as we should be about new ad formats and new types of inventory. With a better interface, at least we don’t have to sell anyone on its value. We just use it.
Advances in platforms are a real help. Nothing makes that more obvious than a quick reminder of how badly the platforms used to suck.
Change is bumpy. But once we get through the learning phase, key tasks are simpler to achieve now. Data manipulation, a dream.
Overture Is Long Gone; Re-channel Wasted Energies
Perhaps the most extreme reminder of how cumbersome and wrongheaded much of the workflow used to be in the paid search world is the view from circa 2000, in the old GoTo/Overture days.
Overture spawned many myths that hamper the effectiveness of paid search campaign managers to this day. Because the system required all matches to be exact matches, the best strategy was to go for maximum keyword coverage, bidding long-tail keywords as low as possible.
I don’t even remember, and don’t want to remember, how campaigns and ad groups worked in those days. But none of the finesse of setting geography and other key parameters by campaign, and setting up ad tests by ad group, was available then. It was, to use a technical term, crap. (Not even counting the fact that the ads showed up just about anywhere but Google.) And in keeping with the Stockholm syndrome, some folks got quite attached to it.
Sure, you can still get a click for a nickel today…in your dreams. (And in Canada.) The long tail has some advantages, but there’s no guarantee those clicks will be inexpensive, nor that they will produce consistently better ROI (define) over and above capturing long-tail queries using fewer phrases, with matching options.
The other thing the old Overture auction had “going for it” was a simplistic ranking formula with no bid discounter, creating an opening for third-party bid management to close “bid gaps.” Because the newer ad ranking algorithms are more sophisticated, bid management can now focus on more advanced objectives.
Life is just better now. We’re not forced to be “good at” playing unproductive games.
It’s 2010: Paid Search Geeks Love Interface Upgrades
Google, in particular, has stepped up the usability of its platform. Building a great campaign is still hard. But refining it is now more enjoyable. For example:
- The “networks” tab makes it much easier to eyeball the performance of all the publisher sources in “automatic placements” (content targeting), and to exclude nonperformers.
- “View all search terms” in the keyword tab allows you to see what actual search queries are associated with clicks on broad matches, again, for possible exclusion of inappropriate matches.
- The “audiences” tab connects you with potential remarketing campaign setup. Beware of this if Google has already pitched your client or boss on it, however. Even if it’s a good idea, you’ll need to work harder now to convince them of it. (Nudge, nudge, you’re better at sales than Google is, because you’re less biased.)
- Filters allow you to set precise rules on your sort. For example, comb through a big list of ad creatives restricted to units with four or more conversions (so you’re not looking at statistically insignificant data).
The tabs appear at any campaign level you’re managing at: account, campaign, or ad group. It wasn’t always that way. It’s now easy to sort all keywords in an account by Quality Score, conversions, or CTR (define), if you’re doing a certain type of sweep. Account-wide management is particularly handy for networks. Cherry-pick the most urgent actions in less time, and take immediate action right there in the interface.
Google isn’t the only one with more integrated reporting. Yahoo Search Marketing, for example, offers indicators on its five-point Quality Index for ads in rotation, which can help you red-flag problems with ad relevance.
Cool Tool or Sales Tool? The Bid Simulator
Some new tools, while cool, are also unsettling indicators that the Quality Score algorithm today is opaque, in spite of the increased transparency in reporting overall.
The Google AdWords bid simulator tool, for example, gives you a visual and numerical projection for how many impressions and clicks you can expect to get for a lower or higher bid on a given keyword. You’ll sometimes see much higher numbers of potential impressions being made available if you bid significantly higher, even if your average ad position was 1.8 or higher. How? The old auction logic would have dictated that in such high ad positions, you’re close to the limits of what a higher bid might bring, volume-wise. Why is Google telling me I can double impressions by doubling an already high bid?
Under the new logic, Quality Score is determined per query, on the fly – each time contextually-based decisions are made not only as to how high your ad will be ranked, but whether it is delivered at all. Google doesn’t explicitly outline the proportional emphasis, but it continues to state (as it has since late 2008) that Quality Score now affects not only ad rank, but the probability of ad delivery on any given query.
As the space achieves maturity, advertisers and geeks alike will trade a bit of money for time and transparency. Google benefits financially, but not without offering some worthy shiny objects in return. As a practitioner, you can make use of the new functionality to maintain and refine campaigns.