The Most Important Words in SEM

By Steven Bauer

The more industry blogs I read, the more I see a common theme. That theme follows thinking that I have been promoting in travelling around the world spreading the gospel of SEM – the three most important words in Search Engine Marketing: TESTING, TESTING and TESTING.

I would like to amend that statement; there are TWO important words in Search Engine Marketing – TESTING and BENCHMARKING.

I hope we all understand the importance of testing. As an industry, we do a lot of testing. We test creatives, landing pages, keywords, and max CPCs all the time. Testing allows us to help our clients by knowing the best creative, the best landing page, the best keyword and the best bid to ensure campaigns perform well. But how do we know they are the best? The answer . . . BENCHMARKING.

Benchmarking is a baseline used to evaluate the performance of account activity.

Benchmarking is not a one-shot process you do when you launch a new campaign. Are you doing a creative test? You need to benchmark the current performance in order to understand if the test is successful or not. Are you doing a landing page test? Ditto. A keyword expansion? Account structural enhancements? Ditto, ditto.

How much data do I need? That depends on the brand or product’s sales cycle, latency, and seasonality; but the standard answer is, “a minimum of two weeks of consistent quality data” is needed to establish a benchmark for clicks, costs and conversions.

Testing AND Benchmarking, the two most important words in search engine marketing.

3 comments:

Aaron Goldman said...

Hey Steven - I agree TESTING and BENCHMARKING are 2 of the most important words in SEM, but let's not forget about APPLYING. I've seen too many companies out there testing and testing (and even benchmarking) only to garner so much data they don't know what to do with it (what I like to call "overanalysis paralysis.") I know one of the things you've spent a lot of time doing here at RM is teaching best practices for rolling-out learnings after testing. Let's make sure the importance of that doesn't get lost in your overall message.

Cheers,
AG

Steven said...

Hey Aaron - my perception of the general marketplace is the majority of people only think about testing and not how to evaluate that testing. Therefore, my emphasis right now is to plant that seed for the importance of BENCHMARKING in search process (paid and natural)
More to come. . .

Dave McAnally said...

"I've seen too many companies out there testing and testing (and even benchmarking) only to garner so much data they don't know what to do with it (what I like to call "overanalysis paralysis.")"

That happens all the time with natural search. The evolution of our analysis (which, regardless of its girth, is generally still essential to driving a project) has been geared towards making it as specific and actionable as possible so our clients don't get mired in details. Not to say we compromise the extent of our analysis, but the more we can show a client what actions should be tied to a given data point, the easier it is for them to prioritize it among the millions of other things they have on their plate.

To Steven's point, I think benchmarking is absolutely critical for any natural search project. If we aren't realistic up front about what we're up against, then we can't possibly be realistic in how we measure success.

 
Copyright © 2008 Resolution Media, Inc. All rights reserved.