EDITOR’S NOTE: This post has received an upgrade. 🙂
Original Post Date: April 10, 2017
Let’s get these AdWords optimization rules out of the way early – I’m lazy as hell.
I don’t mean to say that I don’t like work; I do! It’s more so that I don’t like wasted work, wasted time or futile efforts. Therefore, I focus on the 80/20 rule when prioritizing optimizations efforts.
That makes it a little ironic that I work in PPC where the work quite literally is never done. When it comes to optimization of search campaigns, there’s no “checking it off the list for good.”
Behaviors change, competition changes, the engine and landscape change far more frequent than we’d like. As a result, there’s something to check and do every hour of every day.
But back to the root of things – Aaron being lazy. I’ve worked in an agency setting for the duration in my career, where clients are essentially paying me for my time.
It’s my responsibility to my clients to deliver the absolute best use of my time, so they get the most Aaron for their investment. Clients focus on ROI (return on investment); I focus on ROE (return on effort).
Having done this for quite some time, I’ve come up with a schedule for what you should look at in an AdWords account as well as at what intervals to maximize the value of your time and to maximize the return on every search dollar spent.
Below is a schedule of AdWords optimizations to review on a daily, a weekly, a monthly and a “quarterly” basis. The AdWords optimization list won’t be exhaustive, but it will help you to chunk out your time smarter.
I don’t want to waste time logging into accounts every day, navigating ALL the way down to each campaign and dealing with pesky things like math. Most of these tasks noted can be somewhat easily automated via alerts and/or scripts. They’re noted where applicable.
So without further ado…
Daily – Is Anything Broken?
Fluctuations: The most important thing to check with regard to AdWords optimization on a daily basis is just making sure that everything is doing what it’s supposed to. If an element’s performance changes by 10-15% day over day, I don’t really care.
If a keyword moves from spending $10 a day to $1,000 a day, I want an alarm – even more so if it’s in the reverse.
To make sure fluctuations are no more than daily tweaks, set up alerts for yourself to identify any major fluctuations that may indicate a disturbance in performance. You can read more about creating AdWords alerts here.
Budget Trending: Being an agency fellow, I’m held to stringent budget standards by my clients. Even if you run an account for yourself, chances are you have a set dollar amount to spend per month. Set up a budget pacing sheet for yourself to make sure you’re on target, and that you don’t run out of cash as the month closes. I tend to use the following equation for my budget pacing, as it accounts for week-over-week spend fluctuations:
While we’re discussing budget, remember to check Search Lost IS (budget) column to see potential impressions percentage, impressions that could be lost due to not having a large enough budget. Perhaps, it’s worth investing a bit more, if you find you’re missing out on a lot of impressions.
Figure out what time of day your budget is getting most depleted and you’re missing out on the most impressions.
If you’re past a beginner level, you might want to switch from default “Optimize for clicks” to “Rotate indefinitely” option for ad rotation.
If you want to test ads, you’ll want them to get equal opportunity to show up, to get a more accurate view of what’s performing better.
If you can’t raise budget when seeing you’re missing out on impressions, you might want to keep your ad delivery method to standard. If you do find you can add more to your budget, apply to certain keywords based on Google’s “Below First Page Bid” and set to Accelerated method of delivery.
Bidding: I won’t go too in depth into bid strategies as that’s a conversation for another day, but in short your bid strategy should seek to maximize volume from strong performers, improve efficiency for laggards and weed out (or kill) poor performers.
I’m a firm believer that nearly all bidding should be automated. There’s no need for you to waste time and headspace doing a simple Boolean equation in your head, even more so if you have a complex portfolio strategy. Too often, I’ll see accounts where users are doing manual bidding at the same time every week or month and missing tremendous opportunities where there time could be spent more wisely.
If you have a more complex account, by all means enable a portfolio type strategy to get the most out of your campaigns. More advanced tools like Kenshoo or Adobe offer portfolio tools which allow you to bucket a set of keywords with a target metric. Even using AdWords bid rules in the shared library are technically a portfolio tool which can take care of the bulk of the work for you.
That said, you’ll want to put in a system of checks and balances for your AdWords machine to make sure it’s working right. Set up daily alerts to let you know if top keywords fall below position ~1.2, as that’s a likely indicator that something’s awry.
Remember to check your Average Position on the campaign, ad group, and keyword levels, and perform bid optimization accordingly. If cost per acquisition (CPA) is higher than target CPA, reduce bids and put budget toward those that are more profitable for you.
Weekly – Are Things Trending as They Should?
Search Term Reports: Ah, the gold mines (and trash heaps) of SEM.
As part of your AdWords optimization, you should be evaluating search term reports as frequent as humanly possible, but you can very quickly get lost if you dig too deep. I recommend at least a light evaluation of search term reports on a weekly basis to harvest high level negatives.
Set a threshold for yourself to make sure you’re spending time efficiently. Depending on the account size, I generally won’t worry too much about anything with a few clicks or less.
Better yet, use an automated tool like Karooya to identify gaps for you.
If you do find negative keywords, you can add at the ad group or campaign level.
Ad Test Review: You’re running ad tests, right? Ok, good.
As part of your AdWords optimization, you should review at least weekly to make sure the tests are trending in the right direction. While you’re unlikely to have any data with statistical significance at the one week point, you’ll be able to identify how things are going and how long you have to make a decision.
If you do identify a clear winner in a week, even better. Use a simple statistical significance tool to make sure your data is more than just a fluctuation. Reaching 90% significance is usually good enough for me, but 95% gives me real confidence that I’ve made the right decision.
Same concept goes for Ad Extensions. You should spot check your sitelinks, callouts & snippets on a weekly basis to make sure there aren’t any surprise disapprovals, since those won’t show up in standard alerts.
Make sure that if you have an expiring or promotional sitelink, you’re checking in to replace it after it goes dark as well. Nothing is worse than only having two sitelinks for your beautiful ads.
Placement Reviews (and Bing Search Partners): If you’re running on the Google Display Network or on Bing’s Search Partner Network, you’ll want to review placements frequently to ensure your ads are showing to relevant users in relevant placements, and that you aren’t showing anywhere you don’t want to.
Similar to search terms, these can get hairy in a hurry. Set a decision threshold for yourself to make sure you don’t wind up spending 3 hours to save $5.
When evaluating both, make sure that you’re evaluating the root domain and not a subdomain. In both instances, you may find placements that show a url such as 1234412312.keywordblocks.com (a common Bing partner). While the data for this particular placement may look minuscule, as you roll it up to every iteration of keywordblocks.com, the volume may have more impact.
Monthly – Does the Data Warrant Changes?
Super Search Term Reports: While your weekly spot checks caught a lot of things, that little threshold we set may have allowed a few bad seedlings to grow into weeds. Running a longer term lookback will help catch any issues that flew under the radar during shorter term AdWords optimization reviews.
On the same token, you can use this time to look for any converting search terms that you aren’t actively bidding on. Review the search term report for anything that has converted with more than a handful of impressions and add them into your campaigns on exact match.
Don’t worry about adding everything that converted, especially if there was only one impression and one click. Those terms will likely get hit with Low Search Volume tags, and never be seen again.
Ad Test Decision Time: I recommend doing a more in-depth evaluation of your ad tests every 30 days or so. At this point in your AdWords optimization, if you hit any level of statistical significance, you should feel confident declaring a winner. If you don’t have enough information to make a decision you may never make one, so you’ll likely be safe stopping the existing test and starting anew.
Bid Modifier Evaluation (or how I learned to stop fighting and love the Dimensions tab): On a monthly basis, you conduct an in depth review of the major available bid modifiers to see if you can isolate any poor performers to pull back on/cut, or better yet any strong performers to maximize.
Review day of week and time of day performance both on an account level and a campaign level to see what you find. Usually, the account wide assumptions will hold true for every campaign, but better to be safe and check everything.
Review geographic performance the same way when you’re engaging in AdWords optimization. I recommend reviewing on the DMA (designated market area) level rather than the state level, since many defined DMA’s can perform drastically different within a state. I’m looking at you, Texas and New York.
When you’re reviewing, make sure you’re looking at relative metrics rather than raw number of conversions or clicks. New York, Texas, California and Florida will always have the most volume, because they have the largest populations – that doesn’t mean they’re the best performers.
Instead, look at the metrics that really reflect where you’d like to shift your budget – Conversion Rate, CPA and/or ROAS – then adjust your budgets accordingly. If you find a clear cut winner, you can do more than just increase bid modifiers – you can break it out into it’s own campaign.
This is a good time, in the AdWords optimization process, to review demographic performance as well, as there can be a major difference between HHI (household income) levels or gender/age. A reminder that in order to collect data on these categories, you need to enable the audiences like so.
While we’re on the topic of audiences, it’s important to not only define targets, but redefine them periodically, if needed. Cost and CTR could be good and bad depending on conversion rate and amount you earn from each click.
Think of the lifetime value of a customer. If they have a higher cost per acquisition, but then have great potential to make multiple highly priced purchases from your brand once secured, it might be worth the upfront costs.
Kill or Keep Search Partners: I always recommend starting with search partners both for Google and Bing; however, they don’t always perform as we’d hoped. Take a look on a monthly basis to see if your KPI’s dictate a change in strategy.
Results can vary in Google Search vs Search Partners, so you might find you want to enable/disable the Search Partners option.
Check Mobile Assumptions: I hear this mobile thing is gonna be big. Chances are, however, you set a mobile bid modifier down from desktop/tablet based on lower conversion rates.
On a monthly basis I recommend challenging your assumptions to see if a mobile click is still worth as much (or as little) as you’d assumed. And don’t forget cross device conversions. Here’s a link to a post on the Acquisio blog to help you value cross device conversions into your bidding strategy. So, check your mobile data often, as mobile behavior is sure to change.
Sometimes, we have to dig a little deeper to get to the root of a problem. One might look at a campaign that isn’t performing well and assume running for mobile traffic was a problem, when in fact, it may just be that placements on apps was where lost clicks were taking place.
Cut/Change/Create: We spent all month waiting for our brilliant bidding bots to maximize the value of every single keyword. Did they do what they were supposed to, or do we now have a bevy of keywords sitting in position 900 with impressions that never get seen? Every month or so is a good time to do a deep clean on your account and trim any fat.
As part of your AdWords optimization, I recommend killing anything that’s performing 3x worse than you’d hope. If your CPA goal is $100 and terms are converting at $300 after bid modifications, you’re probably safe to pause them.
Same goes for high impression, low click keywords. No need to bring yourself down with low quality scores – if you find a keyword that has a significant number of impressions (say 1,000+) and few or no clicks/conversions even after your house cleaning, you should clean them out to preserve and protect the rest of the account.
Quarterly – Is Our House Crumbling Down?
If you follow the cadence noted above, chances are your account will look drastically different from where it started even after a short 90-day time period.
Imagine if you go even longer – you’ll wind up being the guy with that house on the block with too many mismatched additions.
I’m not advocating a full blown account teardown every quarter – this about laziness and efficiency, remember? But rather, challenge the assumptions that you made when you built the account every 90 days or so to make sure they hold true when engaging in AdWords optimization efforts.
You may find that a specific region or keyword performs so well that it warrants its own campaign — or perhaps, women in California respond to far different ad copy than college aged males in Florida.
When you find these significant pockets of top (or under) performers, it’s likely time to evaluate your segmentation and perhaps the account as a whole. For further reading on performance based segmentation, check out my post on Search Engine Land.
Wrap Up on AdWords Optimization
Remember, an account manager’s work is NEVER done.
You’ll likely find as an account matures that this cadence may be a bit of overkill. That doesn’t mean you should stop doing it though. It just means you’ll get more and more efficient with your time, freeing you up to grow the business in even bigger ways.