With the proliferation of high-powered third-party tools, we could be forgiven for overlooking the loads of performance and personality you can inject into an account using the latest and greatest capabilities of Filters in Google AdWords. Give this a try! By creating bespoke outcomes using techniques you customize yourself, you’re not only providing a client or employer with the high-value-added service they deserve, but you can feel a bit more creative than you would using somebody else’s off-the-shelf methodology.
We didn’t used to think of “Filters” as particularly actionable, even though, of course, they are. But with today’s capability to select all the filtered keywords and take any type of action (similar to bid automation) on all of them, AdWords Filters behave nearly the same as third-party bid automation tools, or the “Automate” button in Google AdWords.
Using Filters as your automation method comes with a number of advantages. Foremost, you enjoy the flexibility and ease of sorting and viewing that is baked into the AdWords platform. But perhaps the greatest advantage of Filters is the flexibility in date ranges. For some reason, those creating bid automation software have loved “data look-back periods” of a week, a month, or (woo-hoo!) all 90 days…or “all time.” Notice a gap there? Between “90 days” and “six years,” for example? A whole bunch of your account might not have meaningful data from just the past 90-day period. And “all time” may be skewed, or overkill. With Filters, we can opt for a certain 18-month period and make our first couple of parameters related to minimum volume and cost, so we’re now managing a large, meaningful set of keywords or ad groups with a nice, statistically significant history. (To be sure, you should manage accounts frequently – daily, even – but for long-tail segments, you do want long date ranges.)
The principal drawback of using Filters as an automation method is the lack of a well-developed logging system and revert functionality such as that offered by AdWords in relation to the “Automate” button.
Without further ado, here are five fun tricks you can try. Tip: most filters can provide misleading results unless you are thinking of ways to further filter out or avoid anomalies, such as brand terms. It can be helpful to filter at the campaign level as opposed to the whole account level.
- Goosebump. Sometimes you need to give volume a boost, but you can’t always do this in the most perfectly rational manner. Special situation: a client in the travel industry is willing to temporarily relax the allowable CPA in order to hit certain short-term volume goals – to impress an investor, for example. The account’s performance over the past 12 months has been volatile due to steady improvement via optimization and the deployment of Enhanced Campaigns in recent months. We want to reach the short-term sales target, but while minimizing waste. We’ll go on data from just the past 120 days, with the parameters as follows:
- Conversions ≥ 1 (we’re taking any glimmer of performance as smoke to find fire).
- Conversions ≤ 15 (anything high volume, we already manage directly).
- Ad position worse than 1.6.
- Cost/conversion < $52 (target is $60; we’re cherry-picking the good stuff).
- Select all filtered keywords and increase bids by 10 percent.
Remember, you can do this with just a couple of keystrokes with the current version of filters. You don’t need to change 92 bids individually. You can also “preview” the change if you wish. The only major difference between this and a more conventional bid management method is that this method is focusing on a shorter date range than might be warranted to make a statistically significant call on what bid level is accurate for some of these keywords. But given the turbulence in the account in the past year, and the short-term nature of the client’s new sales target, taking account of recent conversion behavior is a decent heuristic.
- Quality watchdog. This one’s easy. In an account that has relatively misleading attribution in the first place, it’s been decided that pausing low-Quality-Score keywords is high priority even if some of those keywords appear to generate some business. The action taken is to be roughly threefold. First, arrest the drain on the account-wide Quality Score factor by pausing all keywords with QS of three or worse. Second, given the waste that was apparently allowed to go unchecked for two years, determine whether the current account manager is up to the job. Third, rebuild the account structure as warranted, resuscitating some of the keywords and showing them against more relevant ad copy, implementing Sitelinks, and evaluating landing pages. The simple filter is as follows:
- All keywords in account with Quality Score ≤ 3.
- Date range: past six months.
- Select all filtered keywords, and pause them.
- Keyword myth-buster. A truism around the micro-water-cooler that is the PPC division of your company has it that all keywords involving the word “buy” or “wholesale” will be worth more than product-related keywords not containing those keywords, so they should always be bid aggressively out of the gate. As it turns out on further investigation, that isn’t always the case. Phrases containing “buy” actually fail relative to other keywords when it involves the broad match type or a mobile device. “Wholesale” only works well when shown against relevant ads and landing pages. How do you challenge the “myth of the buy words” in just a few seconds? Across a whole account or campaign, simply employ the following filter. Again, this parameter is now easily findable in the menu that drops down from the Create Filter function.
- Keyword text contains: “buy.”
- Set the date range on this to the most recent one-year period.
That’s pretty much it. Now you scroll down to the totals at the bottom of the page and look at the relevant statistics (such as ROAS or conversion rate) for “all filtered keywords” and compare it to the mean for “all keywords” across the campaign or account. In an account with many moving parts, this can make a big difference to your methodology in setting initial bid levels for new ad groups, for example. To get more granular with your analysis, you could add parameters for > 30 clicks and match type=broad, for example.
- Group love. If you are stuck filtering just for keywords, you probably use volume thresholds to avoid managing keywords with too little data to be statistically significant. But are you forgetting you can also look at aggregate keyword data – in the form of ad groups – to look for signs of trouble that might not be adequately captured by keyword sweeps?
- Go to the ad group level and set your date range to the past year.
- Clicks > 100, or some arbitrary volume threshold that is not too high, not too low. (Definitely not too high, as part of the point is to catch underperformers that have been flying beneath the radar.)
- Assuming your target ROAS (conversion value/cost) is 2.8, filter for conversion value/cost < 1.4.
- These problem ad groups are now “singled out” for special attention. If your filter only shows 30 to 40 ad groups or less, you could immediately go in and make a number of bid changes in these groups – even on lower-volume keywords arbitrarily. If you have a large account and more than, say, 300 culprits on this list, you may be in for a month or two of more fundamental renovations on these ad groups.
- Match game ’13. Here, we’ll go on the warpath against broad-matched keywords that are underperforming, singling them out for an extra dose of the bid-down treatment as they clearly haven’t gotten the message they aren’t wanted. Broad match types can pull valuable impressions away from more specific match types, and that situation won’t abate if they’re hanging around with medium to high bids. In this case, rather than taking the situation gradually, you want to give it a little extra emphasis with across-the-board cuts.
- Set up your normal bad-keyword-sweep parameters, such as past nine months, cost > $75, cost/conversion > $40.
- Add a parameter, match type. Check only broad match.
- Decrease bid for all filtered keywords by 8 percent. Instantly, your account’s more specific keywords get more of a chance to shine, rather than having their thunder stolen by broad match types.
Over time, your overall ROAS should improve as you bid more accurately by match type as a rule, rather than taking everything case by case. Case by case makes sense on the surface, but if you’re like me, you believe that broad match should be bid above other match types only some of the time. If it’s most of the time, your account isn’t doing as well as it should in terms of relevance. This strategy does not substitute for deep dives into the Search Query Report. It also blends Broad Match Modifier with the more-frequently-misbehaving ordinary broad match, unfortunately. (Come on, Google!)
Many of these techniques are quirky and personal rather than being universal, obvious rules that everyone should apply. This goes to show how much scope there now is for customization and creativity in the use of Filters as bid rules. The Filter menu contains nearly every parameter under the sun, from max bid to impression share, keyword and ad text, engagement metrics like bounce rate, and beyond. Enjoy making your own rules. And in the spirit of PPC slow thinking vs. fast thinking, let System 2 reign over System 1.
Image on home page via Shutterstock.
In part one a few weeks ago, we discussed what brand TLDs (top level domains) are, which brands are applying for them and why they might be important. Today, we’ll take an in-depth look at the potential benefits for brands, and explore the challenges brand TLDs could help solve.
In 2017 it is essential that SEO professionals secure the buy-in they need from their business leaders so they can accomplish their professional goals.
Every year, Google's well-oiled digital ad machine generates tens of billions of dollars in revenue, making the search giant the biggest single recipient of digital ad spend.
Dating back to Ancient Greece and Egypt, monumental structures have relied on the strength of stone pillars, working together to support an immense amount of weight and pressure.