One session that I wish I had attended at SMX West is Avoiding PPC Pitfalls. One of the speakers, Amy Konefal of Closed Loop Marketing, put her Powerpoint online so I read through it. I’m not sure how I feel about the things that she points out.
The basic premise is solid — broad match can kill your ROI. Okay that is 101 when you are training someone to manage pay per click. Amy says that if you broad match the term “car parts” you will show for “fish” queries. I have been playing broad match game, it is a mix of Jumble and a Will Short puzzle, for a few years and I can get to fish in three steps: Car Parts –> Spelling Filter –> Carp Arts –> Theme Filter –> Carp = Fish. Coincidently that logic makes it show for nature painting and sculpture. I have seen a number of strange results over the years from pay per click testing and that scenario above is not one of them.
Certainly the car parts to fish is an outlier to make a point. The point seems, from the powerpoint slides, to be that exact and phrase based matching have a higher number of relevant terms. That is true. It is also true, as Amy includes, that eliminating broad match can leave money on the table. What I don’t see in the slides is the topic of negative terms. Amy’s slides recommend phrase matching and use of exact matching on longtail terms. This is very time intensive and phrase matching still allows for diluted relevance that is the major downside of broad match.
So how do you correct dilute relevance without sacrificing volume?
Use negative terms following the rules that Amy discusses. Take a conceptually broad phrase or broad match term and subtract every concept that you don’t want. Instead of making an additive bidding structure make a subtractive one. What Amy describes is building a number of overlapping phrases that cover your money making zone. This leaves money on the table for you because you are losing out on the unique phrases being searched that don’t match and we know that Google has been steadily changing their model to show preference to the bids that cost you more money. So overlapping terms may be a source of hidden cost inflation.
A well managed campaign should be looking for the raw searches that are off-base and adding negative concepts that cut away the fluff terms. The important caveat is that you are employing phrase and exact matching on terms that are closely related to your intended target. So if you are bidding on car parts on Google -fish is good but -auto accessories is probably cutting into your valuable related terms. Use the phrase matching to remove -”wheel covers” should have you still bidding on terms like steering wheel without showing up for steering wheel covers. In either scenario it is important to put in the time to control your ad distribution, but there is more than one way to skin a car part — or fish.
Tags: Google · pay per click · PPC2 Comments
Related Posts
How Much PPC Competiton Do You Have? Part 1>>How Much PPC Competition Do You Have? Part 3>>
2 responses so far ↓
Hi Carlos - thanks for taking the time to look through the slides from my Avoiding PPC Pitfalls presentation. I think some things are lost in translation though.
I actually don’t recommend overlapping phrases as you mentioned above, or simply going with Phrase and Exact match as a sweeping generalization. A large point made was to ‘buy the most valuable traffic first’. What I mean by that is that if you are an advertiser that is in an industry where your keyword universe is $20K worth of traffic, yet you only have a $5K PPC budget for example, then you should consider Exact and Phrase match on your most strategic variations first - then tap into Broad if you can afford traffic beyond that. Typically Exact and Phrase, in our experience, produces higher ROI.
However, if you can afford the higher VOLUME that Broad can contribute then Broad Match with extensive negatives is often the way to go - but you have to diligently monitor it. If you set it and forget it, you will lose a lot of money given how liberal and unpredictable Broad Match has become over the past year.
By and large, the main takeaway (I hope) was that our situations are individual and our accounts unique - so test and track to determine what the best strategy is for your company.
Amy - thanks for the clarification. I definitely agree. Each situation has to be treated as unique. But I don’t think most people have the breadth of data to figure out what the most valuable terms are starting out.
Many of the no-brainer terms on the additive side are not as effective at the goal as negative terms are. Think of a tennis ball very quickly you can subtract court, uniform, shoe, racket, & match. That cuts out a huge swath of content that doesn’t bring you immediate value. You are right the subtractive method is more appropriate for a campaign manager that is attentive to the program.