Awesome Ad Groups: Small is Good

Last week I described the process of building great PPC search-campaign keyword lists. I walked through an exercise of defining personas — customer groups and their associated needs/desires — and building lists of words they would use as search terms. Assembling these word lists using Excel led to using them as “seeds” in a powerful software tool called “The Permutator” that creates all possible combinations of the seed words.

Here’s the seed list associated with the Roger persona I described last week:

Seed Words and Keywords

Intelligently combining these seed words results in a list of 540 keywords — you can download the Permutator keyword list and an Excel spreadsheet containing the 540 keywords here.

So we’re close to finishing our exercise and launching — just pop the 540 keywords into an ad group, write a great ad, and rake in the profits, right?


One of the biggest mistakes made by amateur advertisers is creating ad groups that are composed of long lists of relatively unrelated keywords, and writing ads that try to cover every concept in the keyword list. Then they wonder why their ad groups perform so poorly — CTRs under 1 percent, low ad positions and high CPC charges.

The good news is that the secret to getting great CTRs, ad positions, CPCs and quality scores is pretty simple: the keywords in an ad group should be tightly related to each other and to the ad text. How tightly? Make this your mantra:

(Almost) every keyword should appear in the ad text.

Why is this important? Several reasons:

  1. When the keywords appear in the ad text, the search engines “bold” those words, attracting the searchers’ eyes.
  2. When a searcher sees her search term bold-faced in the ad, there’s a greater likelihood she’ll choose that ad over competing ads that don’t include the term.
  3. Such ads get the best CTRs — which the search engines reward with better ad position at a lower price.
  4. Over time the keywords and ads in such ad groups will accumulate better and better quality scores — which helps the search engines decide which ads should be favored over competing ones.

That’s why ad groups created with our mantra in mind (repeat after me: “(Almost) every keyword should appear in the ad text.”) are the ones that get the best results, like CTRs in the 5- to 25-percent range.

But how should we decide which subsets of the 540 keywords we created should be aggregated into separate ad groups? My suggestion: group keywords according to where you think the searchers are in the research-shop-buy process.

Though I’ll use B2C terminology here, realize that these principles apply equally to searchers whose needs/desires are satisfied by B2B products and services: most people go through several “search steps” before making a final decision:

  1. Research. People use broad search terms to research available products/services/solutions in broad categories.
  2. Shop. People use specific search terms to look for alternatives and specifics among competing products/services/solutions.
  3. Buy. People use very specific search terms to approach and reach a decision. Often these search terms include specific company names, domain names, and model numbers.

Here’s a table showing search terms used at each of the three stages in the process:

Research Shop Buy

Next week I’ll describe the way to use this understanding of the research/shop/buy process to split up long keyword lists into tight, effective ad groups. As always, let me know your comments and questions.

Join us for SES San Jose, August 18-22 at the San Jose Convention Center.

Related reading