Do you manage too many keywords in your PPC campaigns? Feeling a bit overwhelmed?
There’s a lot to be said for running small, tightly optimized campaigns with short keyword lists instead.
When Chris Anderson, a columnist at Wired magazine, wrote about the Long Tail back in 2004, the concept was seized upon by the search marketing community. The Long Tail outlines a niche sales strategy whereby a vendor can sell a wide range of niche items, in small volumes, which collectively add up to more revenue than their big sellers. Think Amazon. By covering many niches, you make more money.
Search marketers seized upon the Long Tail concept because it dove-tails nicely with search strategy. You can use an infinite numbers of keywords, some of which may only receive one click a year, but added together, they provide a lot of traffic at low cost.
This theory works best in SEO, where there is nothing to manage after you’ve published a page, but in PPC, covering a lot of keyword terms can create management overhead, and affect Quality Score, which drives up your costs.
Which Terms Drive Performance?
Your top 5-10% performing keywords are likely generating almost all your sales. The PPC Long Tail, all those groups of low-traffic keywords, are probably generating nothing but mental overhead. Such campaigns can be tricky to manage well.
Your Quality Score may also suffer if you run long keyword lists. If you’re using an exhaustive list of terms covering areas where there is little buyer activity – the do-it-yourself brigade, for starters – your click through rate could suffer, which can affect your Quality Score. Your minimum bids could rise, so running with the Long Tail could in fact cost you.
Go through your lists and make a note of the low traffic keyword terms. Can any of these keyword terms be covered by broad or phrase matches? What about a combination of broad & phrase match with the addition of some negative keywords? Would you lose anything by doing so? Is the existence of these keywords helping or hindering you ability to meet your sales objectives? Look at the terms that generate your conversions. How many really perform? 20? 50? Would you be better off focusing all your mental energy on these keywords? Are you wasting time testing long tail keywords and ad copy that will take a long time to prove their worth? Is it time for a PPC spring clean?
Running Long Tail Strategy
Of course, some people swear by long keyword lists and running a huge number of groups. This strategy can and does work. Keep in mind that campaigns that receive huge volumes – millions – of clicks at the top end can include a lot of low-performance keywords further down the tail without it affecting the Quality Score too much, but smaller operators may not have this luxury. Few click-thrus, across a wide campaign, can hurt the keyword terms that perform well.
Long Tail keyword terms can also be useful for testing purposes. There might be gold down there somewhere! Again, it’s all about how much time you want to spend on testing and management of terms that deliver limited testing data over long periods of time.
ROI
Whatever method you choose, the important factor to look at the return on investment.
When calculating ROI, don’t forget to build in your keyword management time, and the opportunity cost of that time – would you have been better off managing some other aspect of the campaign, such as landing pages? Use of the Google Desktop Ad Manager or the Google API can improve efficiency, but frequently it is best to focus on improving lifetime visitor value and conversion rates before digging too deep into longtail keywords.
Most importantly, is your Quality Score affected by having too many low paying keyword terms?
What strategies do you use? Do you go for the short, focused campaign in terms of keyword lists and groups, or do you like to cast a very wide net?
This is some of the best evidence I’ve heard AGAINST doing a lot of long tail phrases. I’ve always been of the mentality that “it can only help” to have a longer keyword list, but I was ignoring (1) the value of my time and (2) the potential damage to my CTR.
That being said, I still see value in one of the primary ways I use PPC which is in market research. I don’t always trust the traffic estimation tools, so I like to run an ad to (1) capture actual search volume (impressions) or (2) test an offer before I optimize for the word.
02.23.10 at 8:45 am
Great post. I am facing exactly this issue these days.
Running a campaign for a client that does MBA consulting for leading programs around the world. On one hand, the general keywords such as MBA or GMAT are way to broad and can bring lots of irellevant clicks + very expensive! (even with more than 100 negatives). On the other hand, the 2-3-4 keyword phrases are hardly searched for and creat such low CTR’s in this account which for sure hurts the quality score. Any tips?
02.23.10 at 12:47 pm
Interesting post, particularly considering it contradicts much of what was stated in a Search Engine Land article of the same day (although that post was by Wordstream who have a vested interest in promoting the long tail).
I’m curious if you have any evidence to back up the comment that keywords that have low search volume adversly affect quality score – why would this be the case?
02.23.10 at 6:09 pm
Hi Giovanna,
I think a real key for effectively managing long tail keywords (in the event that you choose to) is creating clusters around the long tail data. Even if you’re not doing this in your Ad Groups (many times it won’t make sense to avoid having both high and low volume keywords within an Ad Group) you want to be sure to have a means of analyzing the data in clusters; RKG had a great post about it here: http://www.rimmkaufman.com/rkgblog/2009/04/13/keyword-tagging/. As you mention it’s really difficult to measure and manage long tail keywords unless you come up with a means for grouping them meaningfully.
RKG also had a good post with some interesting data that basically concluded that the long tail is huge for some advertisers, not-so-huge for others: http://www.rimmkaufman.com/rkgblog/2010/01/04/ppc-head-tail/.
Thanks!
Tom
02.24.10 at 12:25 am
Hi Giovanna,
You’re completely right. Long-tail management can be extremely profitable but it often comes at a huge expense in terms of time required for setup and optimisation.
One method I often use to find a balance between the granularity of a Google AdWords account and the time and effort required for its management is the 10% Clicks Rule http://www.alanmitchell.com.au/techniques/the-10-percent-clicks-rule/ .
By looking only at ad groups which receive over 10% of your clicks, and comparing the ad group’s search queries to the ad group’s keywords they’re being matched to, it’s possible to quickly and easily identify profitable long-tail opportunities, without spending huge amounts of time researching keywords which will only get a handful of clicks per year.
Cheers,
Alan
02.24.10 at 3:35 am
Hi Giovanna,
It stronly depends on the inhouse capacities (hours often cost less money), the size of the campaign, the number of competitors and the volume. Can’t blindly say Long Tail is over-rated.
Regards,
Wieger
02.24.10 at 11:29 am
Its all about different views and strategies. The point stated here is all opposing long tail keywords whereas a mix of regular and long tail keywords will get you branding, reach and conversions together.
02.26.10 at 7:52 am
I think this is true and want to test it. Every time I add a slew of long tails, nothing happens.
I’ve got 50+ campaigns with 50+ corresponding landing pages.
Each campaign has between 25 and 150 keywords.
Are you recommending I go into these campaigns and delete the poorly performing, or not-performing-at-all keywords?
If so, should I just base this on their quality score and if so, where’s the cutoff? A score of 5 and below?
Thanks!
RM
03.07.10 at 9:56 pm
Excuse me if I seem a bit thick (I’m only a very occasional PPC user) but I’ve never quite understood the point of a long-tail keyword list anyway.
I have run some AdWords campaigns targetting broad match phrases like “chocolates” along with a host of narrower, exact and supposedly long (or at least longer)-tail phrases such as “box of dark chocolates”. The thing is the long-tail phrases invariably contain a word which is broad matched by me and every other person on the planet trying to sell chocolates, so as far as I can see there is no cost advantage in having a whole list of exact phrases on the grounds that you are competing on price with the shorter, broad-match phrases anyway.
Or am I missing something important here?
03.08.10 at 2:18 pm
Generally you are spot on Richard… the best broad matched core keyword ads compete along the tail as well. In many ways broad match + negative keywords is one of the best ways to buy keyword data and really expose yourself to most of the market (without needing to spend tons of time on campaign building and such). This strategy works great where markets are illiquid with strong profit margins. But where the profit margins are lower sometimes granularity can be beneficial.
The area where some tail keywords (or even separating out a few of your most important keywords into their own campaigns) can be beneficial is when user intent is classified into a variety of more specific and targeted buckets. That can lead to a more efficient campaign on some areas. PPC is a game of margins, and so a near-break-even broad ad might make 10% or 30% more money if you…
03.09.10 at 9:10 am
Hi I am new to Google PPC and have just found this blog. I have (I think) a simple question, that I hope someone doesn’t mind answering.
I have both long tail and one word keywords. I read on Google help pages to be careful of duplicating your keywords as this can cause a bidding war.
If I need ‘Brighton curtains’ and say ‘buy curtains in brighton’ should I use curtains each time when adding my keywords to my campaign. or should I be adding ‘brighton’ ‘buy’ ‘curtains’
Any help on this would be appreciated, many thanks
Tina
05.19.10 at 9:36 am
It is fine if there is some level of keyword overlap, but ensure you are tracking conversion rates and bidding accordingly. Typically the broader versions of a keyword do not convert as well as exact match or phrase matched tail keywords might.
05.24.10 at 4:01 am
Sorry I’m a little late to the conversation, but it’s such a good topic.
In the world of Quality Score, I think you should:
-start your campaigns with a fairly short list of tightly themed keywords (what you might generate form the free AdWords keyword tool)
-use all three match types (Broad, Phrase, & Exact)
-set the appropriate negatives (use your knowledge, AdWords tool, WordStream’s free negative tool, etc.)
-stay on top of your Search Queries by query mining (ask your search queries if they have enough good performance data to become a keyword, or if they have enough bad performance data (or completely out of context) to become a negative keyword).
I just posted a Query Mining Diagram on my new blog:
http://www.chadsummerhill.com/images/ppc-search-query-mining-workflow-using-adgroup-triplets-2/
I think the right question to ask of a search query that’s had some success is whether or not it generates enough data to make decisions on. If you add it as a keyword it’s so you can manage it, or take action on it specifically. Flooding your account with thousands and thousands of unproven keywords with every possible permutation doesn’t make sense to me. Getting granular is about taking more control (i.e. move a high volume head term into its own adgroup so you can have a separate Conversion Optimizer Bid, etc.), but there should be a reason to make your account more granular, like this long-tail search query is generating enough conversions and cost to warrant specific action.
Use tightly themed adgroups, all match types for your wide net, and watch your search queries.
07.30.10 at 3:50 pm
newbie here doing some research on keywords
Im curious as to how you select all 3 matches – broad, phrase and exact when the drop down list only allows you to choose one at a time. Do you add the same keyword 3 times with each of the choices selected ?
02.25.11 at 6:14 pm
This really changed my take on long tail keywords. Like the first poster said I had always had a “it can only help” attitude. I do agree that focusing on them for SEO is a great tactic however in my experience with small SEM programs it can be a tremendous strain on resources and affect the quality score.
07.13.11 at 1:35 am
Most keyword phrases are highly competitive and difficult to rank for because everyone is using them. Regular keyword phrases have the greatest competition.
Longtail keywords are generally keyword phrases greater than 3 words. Longtail keywords are word phrases that are becoming more prevalent in online internet marketing. Other keyword phrases are over used and longtail keyword phrases offer an alternative at getting more traffic from words that are not used very much globally. Longtail keywords are valuable but not highly competitive which makes them a great resource to use in website meta tags and article writing. There are longtail keyword generators.
11.19.11 at 6:17 pm