Keyword Research via AdWords testing

[Chris Ratchford]

I know this isn't a local specific topic, but I'm curious to know if any folks here have setup AdWords campaigns for strictly keyword testing.

I've done it for non local clients and I'm speaking a client's conference later this week.

I'd like to hear any feedback from people here with their experiences.

Thanks!

 

[Eric Rohrback]

I think buying ad space to start out is a great way to see how people are searching, then you can use that data to optimize your site. SEO takes a long time to see the fruits of your labor, PPC is pretty immediate. Use PPC for the first couple of months to get some solid data, then tweak your site as appropriate. Since Google is quietly edging out organic data, this is one of the better methods to see true search queries.

We suggest this method with many of our clients, and they seem pretty happy with the results. They get phone calls/web conversions immediately after starting the campaign, then in a few months they're in a good spot organically as well.

 

[Chris Ratchford]

Hi Eric,

Thanks for your comment! Have you ever ran an AdWords campaign just for impressions, not clicks?

An SEO consultant here in Charlotte taught me this technique. The goal is impressions (not clicks). What you are trying to do is come up w/ the highest impression keywords and write the website content.

The reasoning behind this is that keyword tools (even Google's KWT) doesn't provide exact volume of searches, only estimates. You'd run this AdWords campaign for 1-3 months, then take the best keywords and write your content (and title/descrip tags).

It also is a great method to determine if you need to create new content or website. Spend $$$ on keyword research before investing a lot more $$$$ on a website.

 

[Dave]

I've done it. I've done it on and off for a couple of years. As far as I'm concerned adwords itself is the best keyword tool for a local campaign.

I've experimented with it in a lot of different ways. I can't think of a better keyword tool for assessing volume of traffic ...than trying to get content that matches that traffic.

Now you may not have to do that. Here is something I gathered from using the new keyword planner.

One way I tested it was against one of my campaigns. I added about 8 exact match high volume keywords and matched the geo area against a couple of campaigns. The campaigns run 24/7 with high budgets. They usually don't run short.

That means I'm seeing ALL the impressions.

The data for those keywords essentially EQUALLED my impression data for the last 30 days.

WOW!!!! google never gave such granular data.

So without even running a test campaign you might get really great data simply by doing your keyword research, identifying the high volume possible keywords and then add them via [exact match] to see what has been driving impressions over the last month.

I only tested it a couple of times but that was pretty stunning.

I'm going to repeat something. They are doing this now because they have destroyed great keyword research.

I looked back about 1 year. Take out all the not provided's from organic search and all the "direct traffic" from IPhones....and basically I think I see only 50% of the keyword data I used to see hitting my site.

notwithstanding ..using adwords for keyword research works very well. Jiggle around with your keywords so they don't show in the top 3 or even lower. With that kind of presence it doesn't cost much.

 

[Dave]

One of the more productive uses of this technique is to use it when considering expansion.

Say I've an smb in one California city, say San Jose. I'm thinking of trying the smb in another city or two. Run a ppc campaign for the smb in some other cities. You can vary your choice of keywords. Say your a wedding photographer in San Jose. You are thinking of expanding to Oakland, San Fran or other cities.

Set up an adwords campaign for each extended area where you are considering a new location and test various keywords on it ; keywords with geo terms and keywords without.

Its a tremendous way to measure potential size of market place.

Say you are looking to expand into a larger metro area. Run a keyword campaign there. Then check adwords to see which part of the metro region generates more impressions. Assume the city will generate the most. But which parts of the suburbs expands the most???

very effective little tool and helpful to planning. Want to check branding phrases? Run a campaign wherein you use exact match for various competitors. Bam!!!!! great way to see who has the most branding and name recognition!!!!

 

[Laustin1878]

Great thread and great information guys, thank you. I have used this in the past and as noted already, you are getting factual post key term data as opposed to pre-data that any tool will provide. Keyword tools typically provide predictions and estimations where post data is actual data.

One question and I don't mean to hijack, what have you guys found to be the least amount of money to throw at a test like this? What have you found to be the most minimal budget that will provide you with enough data to make an educated decision? Many smb's don't have big pockets and as with everyone, want results yesterday. How do you pull this off with small budgets?

 

[Dave]

Great thread and great information guys, thank you. I have used this in the past and as noted already, you are getting factual post key term data as opposed to pre-data that any tool will provide. Keyword tools typically provide predictions and estimations where post data is actual data.

One question and I don't mean to hijack, what have you guys found to be the least amount of money to throw at a test like this? What have you found to be the most minimal budget that will provide you with enough data to make an educated decision? Many smb's don't have big pockets and as with everyone, want results yesterday. How do you pull this off with small budgets?

Each business will be different. It also depends on your goals. Frankly I dislike spending money on the research. Since in all cases I'm testing things that I never tested before...I really don't know how the ads will be placed. If I only want the posts to gather impressions and not clicks, I might write a lousy title, get a low quality rating.

Then what I want to do is adjust the campaign so I'm showing low among ads. It won't get many hits. But I do have to spend time on it and fine tune it.

For some experiments with certain goals I got my answer in a week. For others I 've tested over weeks or months.

An smb spending $3k/month has a far greater capacity to test these things than an smb spending $600/month. Also if you have a high roi you can afford to experiment more. If the ROI is low....an experiment of this type will probably lower it

 

[Linda Buquet]

The reasoning behind this is that keyword tools (even Google's KWT) doesn't provide exact volume of searches, only estimates.

Keep in mind the new Google Keyword planner provides quite a big more data, especially based on location.

Plus you don't need to know exact counts, you need to know relative comparative popularity. You need to know if Atlanta Cosmetic Dentist or Atlanta Cosmetic Dentistry or Cosmetic Dentists in Atlanta gets more searches, for example.

Here's my post if by chance you have not seen it. There's a good Dental example.

Hot for Local! NEW Google Keyword Planner - KW Research by City

 

[Dave]

I tested the keyword planner in a different way. One experiment was to test specific keywords for existing campaigns. That is simply a different option among the methods to use the planner than how Linda tested it.

When I ran it against my existing campaigns I inputted a group of [exact match] keywords in brackets and broad match keywords.

I used the exact same geographic boundaries for the campaign.

The startling data was that the info was exactly what I was seeing in adwords. The scope of the data was the last 30 days.

Why does google show data for the last 30 days? I don't know. But if you use the google adwords advisors they'll take you through some experiments and they'll tell you to set the campaigns for the last 30 days. I guess google likes that time frame.

I suspect the planner as Linda showed it in her experiment are examples of estimates. All the numbers are rounded. As she said though, the scope of the difference is the important data.

You might want to take the data from the planner in the way Linda showed it...then add those keywords as a check. One by one. Use [exact match] for the critical phrases and use broad match. Try them both. See what kind of data you get.

As for the experiment in using adwords to test potential keywords I've used it in a couple of ways:

A) test concepts of words that are different than the main keywords but are relevant to the business...maybe not directly but tangentially. Previously I had tested phrases against one another using google trends. Running a test campaign was far more granular, specific and believable.

I've tested keywords for possible expansion. I've also tested for branding phrases.

Any other uses you can think of????? I'm all ears.