I was recently speaking to a fellow search marketer who was stumped as to why all the key performing indicators were moving in the wrong direction on one of his accounts.  We checked the the main culprits to make sure nothing was inadvertantly changed.  For your own reference, those main culprits are typically:

• Campaign settings accidentally changed
• Landing page no longer functional, or loading very slowly
• Added a term as a keyword instead of as a negative or vice versa (forgetting to put the minus sign “-” is a big head smacker)
• Reasonable ad copy that follows best practices

Once we determined that these weren’t the issues, I asked him: When was the last time you updated your bids?

“Oh, I have my bids automated so I can spend my time on more important things”

Ah Ha!  Bid automation can be your best friend and worst enemy at the same time.  Most people who have done PPC for any length of time can tell you how time consuming manual bid adjustments can be, especially on accounts that are more than a couple hundred keywords.  Don’t get me wrong–manual bidding is way easier and less time consuming today than back in my day when I had to walk up the Overture hill both ways to the Google school in the Miva snow.  However, I continually find that the professionals who rely solely on automated bidding will sooner or later find themselves stuck in what I like to call the death spiral.

The death spiral occurs when automated bid rules take your bidding limits to the extremes in either direction.  For example: I tell a program that I want to bid between $.50 and $1.00 with a target CPA of $50, and to check back every 7 days.  When the program runs and sees the CPA is $20 instead of $50 with bids at $0.60, it’ll raise bids to get higher positions to achieve more volume, which typically raises the CPA.  But in 7 more days the CPA is $45 with $.75 clicks, but since we’re still under $50, it raises bids again and will almost guarantee my CPA busts over $50–chances are way over.  In the opposite sense, if my CPA is $75 with $.60 bids, it’ll drop the bids as low as it has to to achieve that target CPA, but wait!  What if by lowering the bids the position drops so much you lose some converting traffic?  Then your CPA will actually go up instead of down. Once that happens, you hit the death spiral and will never reach your goals. Doh!

Anyways, back to the problem at hand.  By letting a program do all the bidding for you, regardless of how long you’ve been involved in search and know the rules, you will begin to lose focus on what is the ideal combination of bids and ad copy for a given keyword and become more reliant on the program to continue.

Consider this analogy: A baseball player can hit home runs all day long against a pitching machine, but to continue to be good at batting against real pitchers, the batter must also practice his stance, how he holds the bat, how to twist his wrists as he swings, etc.  When a player begins to show weakness in any of these and can’t hit the ball against a real opponent, he’s sent down to the minor leagues to practice the fundamentals.  That’s what manual bidding is: a basic fundamental that all search marketers need to excel in.  Automation can only follow rules, search marketers can break them when they see opportunity, or to divert sudden disaster.

“Hold on,” you say, “I have thousands upon thousands of terms.  There’s no way I can do manual automation for all of the terms.  I don’t have enough hours in the day.  And even if I did I wouldn’t have time to do anything else in the account.”  That may very well be true, but in that case I want you to manually adjust bids on your top 10 terms and leave everything else on autopilot.  Check up on those auto pilot terms at least once every 30-45 days and make some manual overrides as you spot issues.  But whatever you do, don’t put your whole faith in a bidding tool. Doing so risks the success of the account on the assumption that you were 100% correct when developing the assumptions required to write the rules to base bid changes on.  I’ve been doing this a long time and will be the first to tell you I’m never 100% correct.  (The crazy advertisers who don’t know what they’re doing and bid $10 on terms will throw any assumptions you have out the window).

After my “ah ha” moment my search marketing colleague went back and analyzed which keywords were converting that no longer are, and looked at the bid changes.  Sure enough, he was caught in the death spiral, thanks to a lack of conversions that was due to a hosting error with his client’s site.  The site being down meant people couldn’t convert, which drove the CPA higher, and the bidding program put several moderate-performing keywords to non-performing levels. The impact from the sudden aggregate performance loss from several mediocre terms became very big very quickly.

Please don’t read this as me hating automated bidding, because I strongly believe it has its place, but I disagree with relying on it exclusively for the purpose of working on more interesting aspects of search marketing.  If you’re a search marketer with your head in the clouds thinking about the next greatest test that will blow everyone away, don’t forget to come back down to earth and do some of the nitty gritty work that we all need to do.  Your clients will thank you, and you’ll be a better marketer.