Over the last couple of years I’ve noticed a couple things in the search space: 1. keyword tools are unreliable 2. PPC traffic can show some valuable search trends. Well here’s the point: PPC and SEO help to inform each other consistently.

There are reasons why SEO and PPC should always be in cahoots: On the SEO side, a simple look at the analytics queries that come through and spend time on site are usually from keywords that are relevant and have a value. If there is evidence that some terms that show up organically on page 2 or deeper in the SERPs are getting hit are are converting, then these are that have value for PPC. We can immediately pay PPC to be seen next to the top ranking organic sites showing for these competitive terms. Of course, because this is PAID traffic, and because CPCs can be so wildly different for each industry or vertical, ROI calculations should be applied constantly.

Now it gets interesting. On the PPC side, and before SEO goes to work creating targeted landing pages, title tags, keyword rich content on the site, etc., wouldn’t it be great if there was an idea of ROI that stood independent of search volume statistics? Well, unfortunately this is the risk of PPC at every turn in the game. Yes there are competitive analysis tools, bid optimizers, keyword research tools, and on and on… These amazing tools, while infinitely helpful for a modern marketer, just do not round out the actual field of search. When it comes down to it, every search is still based on user generated content, and this content, even if sometimes predictable, is just as often counterintuitive.

I propose that while entering an SEO commitment, you should be testing the viability of the keywords you want optimized on your web pages by employing PPC. This way you can build out the particular keywords on the main targeted pages of the website while keeping tabs on the PPC keyword statistics which perform best for your site’s ROI.

Through Google, this should be done in a couple ways:

1. Use your usual keyword research tools to come up with a nice grouping of unique keywords in order to test how they react in the live market. Remember to only use the exact match option for these keywords, be sure to bid them to a desired position, and most importantly, make sure content network is set to “off”. Obviously this could get pricey as you want to ensure that your keywords show up before the fold on page one and keep in mind that you want enough data in order to make decisions, but this method will give you an idea on how often your keywords are currently queried and how well they convert for your site.

2. Set up a separate ad group for testing. This adgroup should be phrase matched and include peripheral keywords that are more open ended. Pull a Search Query Report every week or so. Depending on volume and conversions, this adgroup should yield a bit of information. Use these newly found keywords that have converted for your site in the exact match campaign - overall, you may be surprised by what has brought in revenue.

Remember to continue to test new keywords, and certainly don’t forget to put a good spin on your ads. Don’t forget to test ads and don’t stop coming up with new approaches to test.

After a couple of months you can take your PPC metrics and compare them to your old keyword research tools… odd how PPC traffic and traffic estimate tools seem to vary so much sometimes. no surprise… no surprise… what the f….?!?!?!

Now you can take both statistics into account when optimizing your pages - using PPC to find keywords that are actually profitable for your site can save a tremendous amount of time on the SEO side. There are times when the highest volume keyword that a site could rank for is not the most profitable, and these are the times for which we need to test. Suddenly the PPC advertising budget can become an investment in a long term search strategy.

Finding the proper keywords is the cornerstone of any successful SEM campaign. I imagine every PPC manager out there begins a campaign with keyword research in order to find what keywords are going to generate the desired traffic to a website. This will typically include picking basic, generic keywords and finding similar search queries which will provide the basis for an extended, or long tail, keyword universe.

While gathering these initial nuts and bolts of a campaign negative keywords will also be identified, ensuring ads will not show if certain keywords are added to the query. Some of these negative keywords will be obvious, while others can be surprising as there are now so many unique industries and websites catering to niche markets. This type of research is invaluable since it is impossible to know a priori every possible query that could trigger an ad that, if clicked, will cost money resulting in no ROI. (Of course negative keywords are only valuable in a campaign if broad or phrase match terms are used - though that discussion will have to wait for a future post.)

The true value of a keyword lies in its ability to put relevant information in front of the searcher. Sometimes this goal breaks down, such as when there are conflicting search queries. For example, if someone searches for “green paint”, this can have a couple of different meanings: Green as a color or Green as in eco-friendly. If you are bidding on this keyword, it is important to identify not only the semantic value of the keyword; user intention can determine whether a certain keyword is profitable to use in a PPC campaign, and certain keywords may become more competitive due to semantics. In addition to semantic issues there are also long tail keywords which can easily trigger an ad if broad match or phrase match is used. There are literally millions of examples using Google Adwords Broad Match option - The keyword “Al Green” could show ads for “Al Green CD”, “Al Green Tshirt”, “Al Green MP3″, etc, depending on the market. It is always important to recognize the goals of the campaign and to truly know what is being sold and why… sure. But how is this kind of research best done when every keyword research tool out there starts to fail and an account is already live?

This is where Google’s Search Query Report can come in handy. This report can fill in the gaps for your PPC campaigns by providing the PPC manager with actual user generated search queries used in the live market. Unfortunately, Google Analytics does not show this even if you have a linked PPC and Google Analytics account. The terms that show in Analytics will merely show results on the keywords as they are entered into AdWords. If you had bid on the keyword “widget” as a broad match term, Analytics will show metrics for the keyword “widget”. But this is regardless of how many searchers not only saw the ad for that broad or phrased match keyword and then clicked even though they may have searched for a term like “green widget”.

Now what if the company doesn’t sell green widgets? Worse yet… what if the company or PPC manager had never even heard of a “green widget”? There is no way to stay ahead of every industry curve, and since many advertisers find themselves in a live market it is best to get the real data while it’s still out there and available. Now, pull that report and see what you’ve been missing.

Just don’t blame me when keyword popularity changes and we need to do a reverse keyword audit on your PPC campaign.