PPC Bid Management Strategy – Bid Optimization of Non-Converting Keywords
Most of my PPC campaign management is done by hand, without too much intervention by bid optimization or bid management tools. The following suggestions come from direct experience working with thousands of running keywords. What generally happens is that keywords begin to fall into groups:
1. Monthly converting keywords – keywords that convert each and every month.
2. Irregularly converting keywords – keywords that convert well one month only to not convert at all the next month.
3. Rare converters – keywords that rarely, or never convert over extended periods.
You can optimize a campaign after just 30 days, but you’re most likely to drastically cut group two to rarely convert, and cut out the rare converters entirely. With just 30 days of data, you can’t know whether a keyword will be an irregularly converting one. If a keyword costs $600 one month and does not convert, and you go to cut the bid so it won’t be displayed as much, that keyword may spend $10 the next month and not convert. Had it spent another $600, though, it might have converted a dozen times. Some keywords, believe it or not, actually do this.
On the other hand, if a keyword converts well in the first thirty days, you could increase the bid, optimize ad copy, improve landing pages, and improve its ad position, only for it to spend a fortune and not convert well or at all the next month.
My experience: Optimizing keyword bids too early can quickly establish a positive ROI for negative campaigns, but with smaller profit margins because you’re cutting rare converters and not taking chances on keywords that did not convert in the first thirty days, but could have later become irregularly converting keywords.
I don’t believe anything is a trend until it has at least three data points. Two months of data does not show a trend, it just shows one result versus another. At three months and longer, you will begin to see trends. So just because a keyword converted both months or not at all may not be indicative of anything. Still, optimizing after 60 days of data should focus primarily on keywords that have not converted at all, or converted both months.
If a keyword is spending lots of dollars, and its average rank is 1.6, but it didn’t convert either month, that keyword isn’t doing its job. You can always optimize ad copy and landing pages to try and get it to convert, but when it comes to bid optimization, this is the type of keyword that probably just doesn’t convert, and should probably take a cut in its bid. The positioning is very telling about whether a keyword will convert.
A keyword that converts both months and has an average rank of 4.7 may convert twice or thrice as much if it were to show more frequently in the first and second ad positions.
My experience: Increading the bid on keywords that have mediocre average rank is a gamble after 60 days, because it’s often an expensive keyword, but the gamble could be the key to a smashing hit of conversions.
Also, it’s easy to get nervous about high volume keywords that convert one month and not the next. By being upfront with clients about the bid optimization process, I’ve been fortunate to allow some of these keywords to run longer than my gut told me to. I highly recommend this route because it allows keywords to begin to show a trend with 90 days of data, and maybe putting the keyword in the irregularly converting instead of rare converters groups.
Optimizing after 90 days or longer is ideal. With 90 days of data, you can see three months’ average CPCs, three months’ average rank, and three months’ conversion data. A keyword that had a CPC of $1.24 then $1.35 then $1.48 with a steady or dropping average rank and relatively stable conversions may be a candidate for an increased bid, because there is greater competition for the keyword. It may also be a candidate for a decreased bid, though.
My experience: The average rank tells the story in conjunction with the other data points. An average rank dropping from 1.6 to 2.1 but still converting may be able to withstand a decreased bid and maintain conversions, even at the risk of dropping to an average position of 2.4.
NOTE: This post did not address ROI or ROAS or CPA, it simply approached bid optimization/bid management from a converting vs. non-converting point of view. Always optimize keywords while factoring in whether they are providing a satisfactory return on investment/return on advertising spend and/or cost per lead/acquisition.
If you’re looking for more on ppc management strategies, check out my post on faster ppc management.
September 15th, 2008 at 7:45 am
Thanks for the great post and i agree with many of the points. Since i’m also managing some of the campaigns manually and i have tried some of the techniques mentioned her and it worked out well for me.
June 20th, 2009 at 8:09 am
Thanks for the excellent post.