This may seem like old hat to many SEOs, but it's a tip that never fails to get an "oh yeah!" during client meetings. The concept is simple - in any given search engine optimization campaign, you are naturally going to form a list of high-traffic, (perceived) high value keywords that are an idealistic goal for your site to dominate. For a site like SEOmoz, those might be the highly competitive terms like "SEO" or "Search Engine Optimization," while in a field like BuddyTV's it might be "tv shows" or "tv news."
The problem is that while these keyword searches seem like no-brainers, ranking for them can take a remarkable amount of effort on both the content and link building side. To warrant that investment, you need to know, from a business perspective, that financial returns will accompany the rankings. One great way to do this is to use paid search to investigate the likely ROI of visits from those keywords. Buy the keyword traffic for a few weeks or a month and measure visitors via a segmented tracking campaign (check out this post on action tracking to learn more). If the visits that arrive via those searches convert well and produce value, you know that a serious investment is warranted. If, however, they turn out to be tire-kickers and have a low propensity to produce returns, you can re-focus on higher ROI targets.
There's just a few valuable tips to bear in mind when you're pursuing this process:
- Paid search traffic can behave differently than organic traffic, so don't take the figures at 100% accuracy. Build in some room for error, and you'll create far better expectations.
- When crafting your PPC campaign for test purposes, make sure to narrow to exact match so you don't accidentally measure traffic that's coming in for longer tail or modified versions of the search query. It's great to do this and measure response in a PPC campaign, but with SEO, you won't be able to naturally rank for those same variants unless you identify and target them individually.
- Make sure to narrow to a geographic area, especially if your keywords contain any potential local intent or local modifiers. Otherwise, you can seriously over/under-estimate.
- Keep seasonal variation/flux in mind. Use Microsoft's Keyword Forecast or Google Insights for Search to help out. Volume fluctuations usually indicate shifting intent as well, so purchasing keywords in a down period can hamper the accuracy of your forecasts.
That's it for this week's headsmacker. I've got a very personal post I worked on during my plane flight back from LA this weekend coming soon (hopefully tomorrow), and we're also launching our new blog etiquette guidelines and some explanations this week, so stay tuned!
BTW - If you somehow missed it, go back and check out Danny's brilliant post from last week on analyzing the Top 100 Blogs. It flew under the radar a bit, but is worth a thorough examination.
Excellent point -
Makes me think of uses of paid traffic for other things too like using Stumbleupon paid traffic for a quick glance of if a page sucks or not.
Pay like 50 or 100 bucks to Stumbleupon and the reports they give you of how many people liked it or disliked it (and how many didn't care) are a great resource.
Can you explain a bit more on what paid Stumbleupon reports are? I am new to social media and did not know that they have paid featured. If I am misunderstanding.. its the sudefed from the cold. :) ~ Ginger
Hi Ginger,
Stumbleupon offers a paid service where you can take different pages and have them displayed to stumblers.
They show the thumbs up and downs of people who visited your site
https://www.stumbleupon.com/ads/
It is handy to get a quick view of what people think of your page and it drives alot of quick traffic to your site for many great categories.
Here is a demo of the product
Thank you! I just signed up for an account and as usual am staring at this thing wondering what all I can do with it. You just saved me a lot of time. Appreciate it. :)
Another excellent tip Rand.....Quick question though...Is this process viable for small to medium businesses with small SEO budgets?
What sort of daily budget would you look at setting to really get a good idea of the amount of traffic?
@Thetjo - Liked the look of the microsoft tool...thx
Kel
Kel - sadly, there's no hard and fast answer. It depends entirely on the level of advertising competition for the keyword phrase, the bids they've put in and the amount of traffic. The nice part is that during the test campaign, you can control costs fairly well - so the only investment question is whether you've got enough to build a statistically significant model.
@Kel
I usually try to run these PPC tests on content networks, which are usually more affordable. A test on Yahoo and Live is a potential, after all, you just want to find out if a keyword works or not.
In order to make it even more cost effective go for exact matches and avoid too broad terms - which would have lower ROIs for small businesses anyway. Exact match on a longtail will allow you not only to judge the volume, but the quality of the volume.
Once you see the benefit of the narrow keywprds, and optimise for them, you can take steps to target volume for the broad matches.
For small businesses such a staggered approach reduces the cost bundle.
@Rand - Controlling the costs is a plus and I suppose when you think about it....even if you set $20 a day with $2 max bids on an exact match keyword and your money gets gobbled up inside an hour..you could safely say you're in the zone.
@ Rishil Thanks for chipping in...I've really enjoyed reading your small business SEO posts...it's really helped alot. The Content Networks...would they be places like Facebook? Previouslly I've been judging keyword strength based on competitors keywords...
Cheers Kel
I am glad you enjoyed the series.
Regarding content metworks, these are any sites that feature adsense on them. They would show your ads if there is a good concentration of your target keywords.
Regarding low budgets, try and geographical limitation. In my experience cpc drops with super targetted campaigns.
ofcourse - adsense...thank-you
Please keep the small business series alive....they really are awesome reads and provide a bucket load of info for guys like me on the otherside of the world trying to break into the same market..
K
@Rishil,
I'd be wary of any data coming from testing keywords on the content network.
Content network users are in a different stage of the buying/marketing cycle then users on search engines (be it organic or paid).
It also depends on your website though I guess - if you're looking to attract customers in the early stages of the buying stage, for example you run a review site, then it may make sense.
This is actually a technique we employ all the time. As a small "botique" company that offers speaking and training services, it is worth it for us to test the terms first.
Our PCP budget is fairly high for a small company.. in total about 5 grand a month. When you consider that one sale equals 3500 to 7500 for one gig, that budget is not so bad.
Most of that money goes to our most competive terms obviously, but especially after spending some time at the SEOMoz advanced in Seattle, I am more willing that ever to test the "long tail" on pay per clicks. If something turns out to be *HOT*, you bet I'll optimize for it.
~Ginger
"Our PCP budget is fairly high for a small company.."
I didn't know people actually budget their drug use.
LOL..ok..you caught me posting on sudafed.. Seattle gave me a cold.. Figures that my first few comments here would be that way.. :) ~ CreativeSpellar
I have a real life example from a small startup business. Budget on PPC keywords was $20 per day. We learned that the key word we thought would be great produced tire kickers and did not convert well and discovered a few other key words that hadn't thought to be the top ones were better. It was worth the money spent although we still have a lot to learn and areas to develop further.
As Tim Ferriss mentioned in his interview with Derek Sivers, he ran a Google AdWords campaign to determine the best book title for "4-Hour-Workweek" based on which ad had the best click-through rate.
In addition to the tools mentioned, check out the Microsoft Online Commercial Detection Tool. It's a great tool for measuring the commercial value of a given keyword or website.
"check out this post on action tracking to learn more" there should be a link in this line, right?
1. Make sure to add google analytics tracking to your landing page
2. Change settings in google analytics to be able to track exact match phrase traffic from broad match targeted keywords.
3.Bid on Exact Match keywords you find relevant
4.Bid on Broad match keywords to increase your list of exact match keywords, this list you will get through google analytics results.
This way you dont loose any exact match keywords you may not have come up with in the first place.
This may belong somewhere else. If so, please point me in the appropriate dirrection. My company sells a high end service so how do I measure analytics other than through having them do something such as fill out a form. I have never signed up for google analytics because I was under the impression that it was for people who were selling products and wanted to see a direct correlation with their PaidForClicks.
If there is a paper on this or some information that would be helpful can someone direct me that way please? ~ Ginger
Actually, analytics tracks pages that you have their tracking code on. As long as it's on the page, Google will track the number of users who surf onto it. It also gives you keywords they find you with using the search engine.
I think the only pages it can't really track are e-commerce and secure https pages. Someone correct me if I'm wrong.
@nonlinear - thumbs up for you point re tracking exact match search queries in Analytics.
For anyone interested, GA doesn't do this automatically, but thankfully there is a simple way to pimp your reports to track the exact query people have been searching for prior to clicking on your PPC ad. I haven't tried it yet but given the limitations of Adwords' Search Query Performance Report I think its just what I've been looking for!
Hey you've forgotten to put the link to that post on action tracking.
We were just debating the effectiveness/costliness of this type of testing the other day. Like everyone else, we have some very competitve keywords that we target and are looking for a good general tool for determining an accurate volume of searches on particular words.
Since we coulnd't find anyhting that we thought was particularly accurate, the idea of testing with AdWords came up. Put in the "exact match phrase," build a terrible ad, see how many impressions we receive and hope we don't get too many clicks!
We have limited budgets to work with so the cost of testing with AdWords is prohibitive unless we can limit the number of clicks we receive. Anyone have any other methods/tools of determining accurate KW volume?
None of the big name tools we use for KW research does a good job of this due to sample size, and I'm guessing Google doesn't share this info.
You do know of Google's keyword tool, righ?
Yeah, but it seems to be highly overestimated and very round. Right now we basically use a conglomeration of several tools . . . that gives us a decent ballpark, but nothing accurate in terms of the number of searches 8-(
Rand great post, I often times forget about the value of PPC for testing keyword ROI. I guess that's because I stay so focused in the organic world, that I kind of think as paid search as secondary. But it really can be invalueable to your organic efforts, especially if you are trying to pitch a client that you are aiming to provide the best possible traffic to there site. Because we run into clients that want to rank for "blue widget on sale" but "blue widget on sale" provides 0 traffic and no value to them. however "blue widget extravaganza" provides tons of traffic and these people are converting and taking the desired action. Because no matter how good they think they know their customers and what they want, they often times really don't think about it as a searcher. Thanks again for the great reminder how PPC can help our organic campaign.
thanks rand. very informative post. Certainly is a good idea to track the performance of adwords, especially while waiting for organic listings to take effect for new clients/keywords
Great Headsmacking Tip...it honestly hadn't ocurred to me, but totally makes sense. ROI is the trickiest part of the keyword selection process in my opinion, and this makes it much more of a science than an art.
Hey Rand great post. Someone already mentioned this in the comments above but the link for "(check out this post on action tracking to learn more)" is missing. I would love to see the info that the post has.
I'm guessing it's this one:
https://www.seomoz.org/blog/actions-worth-tracking
Thanks rand, great advice. I used to work for a company whose clients were all localised, and so promising a page one spot to them was no big deal if they expected people to put in the town name in Google. However, working for a nationwide company formation company now, I've realised what an artform it is when you're up against a whole country of competition, rather than a small town!
Big ups for the Headsmack :)
Love the series. Thumbs up from me as well.
I total agree that traffic behavior, conversion and bounce rates can be quiet different between paid and organic rankings and in most cases there are also seasonal changes so a short ppc test may not give any accurate information depending on the product. But still very useful information.
I love it when you're reading a post and a question pops into your mind, then a few lines later the answer appears :P
"Paid search traffic can behave differently than organic traffic, so don't take the figures at 100% accuracy. Build in some room for error, and you'll create far better expectations."
PS....how do you quote a portion of the entry in your comment (the right way)?
This is just MY humble opinion, but I would use both Google Analytics and a Hitslink Enterprise edition tracking account. Why?
I've noticed that Google Analytics will only show you the already entered keyword that was triggered for a specific query, but will not show you the actual search query typed in. You will see specific queries on the organic side however.
But I totally agree with this post. It's especially useful to SEO for a high cost, yet high-yield keyword, to save money when you don't have $5/click to spend.
I'm glad that I can finally contribute. This is fun!
Hi Clif,
Google gives the exact search query that was entered if the settings are modified.
Or did i misunderstand? Can you get the queries that ads appeared but were not clicked on?
From what I understand, queries for which you weren't clicked on can either be acquired by
1) Google Webmaster Tools. Found that out from the Q&A featured somebody clued me into
2) Hitslink Enterprise, if you also happen to have high SEO rankings. So if your PPC ad doesn't get clicked on, maybe your organic ad did.
As far as a setting in Google Analytics, I'm not really sure.
Rand. Excellent post. Even though it seems obvious, it's a definite had smacker. With all the information and strategies that is out there, sometimes we fail to see the easiest things.
Antonio Days.. Instead of using some zippy one liners... how about you share your thoughts and enlighten us.
Great point. Aside from getting a ballpark estimate with the various keyword tools a small segment of the budget allocated for 'research' can be money well spent.
Coming from a strict marketing/advertising background, it's unfortunate that my HTML coding abilities are a little on the weak side. Enter Drupal CMS and I have a fallen in love. We've been making the switch to Drupal for a large majority of our clients so they can easily add and update content. What I've come to realize is that it's fantastic for SEO as well.
Using Drupal CMS (or a CMS like it) I initially optimize top level pages with keywords solely from keyword research (they may end up being great keywords so why not?). Once a PPC campaign is up and running, I can add in appropriate keywords based on PPC results and ROI. I'm seriously in love with this system. It takes mere seconds to update a page ...
Great tip. Even with "the numbers" exposed finally in the Google keyword tool, they can be inaccurate or deceiving.
Nothing beats real life testing on your own site - then you can see some much more accurate data. Of course, this also takes a lot more time :(
Thanks Rand, this is a great post for people like me who are living in organic traffic only.
My dream is to get to organic traffic only so I don't have to pay google. While we have made progress we have a long long ways to go.
Useful article for people who is new in Seo.
But people can use : https://adwords.google.com/select/KeywordToolExternal
Google Adwords Keyword Tool.