Search Engine Watch
SEO News

Go Back   Search Engine Watch Forums > Search Engine Marketing Strategies > Web Analytics & Tracking
FAQ Members List Calendar Forum Search Today's Posts Mark Forums Read

Reply
 
Thread Tools
Old 09-13-2006   #1
Northwind
Member
 
Join Date: Aug 2006
Posts: 31
Northwind is on a distinguished road
Mining the tail?

Gang - I'm still reading lots of press around 'tail' keywords and the ROI they bring to a campaign, and I'm curious. First, is this a strategy anyone really takes into account for their own campaigns on a regular basis, or just engine hype? Also, what products track these (other than maybe hitwise) or are you running inhouse software to pull these words. I've seen nothing, and consider passing it on to the IT guys.

Thanx,
NW,
Northwind is offline   Reply With Quote
Old 09-13-2006   #2
startlaunch
Member
 
Join Date: Sep 2006
Posts: 21
startlaunch is on a distinguished road
Sure, I use them.

Lets look at an example for the ringtones.

All you have to do is think about it logically. First, do a little bit of keyword research to find every variation of the keyword "ring tone" including misspellings, variations, multi-word variations etc.

After this point you should have something like 3000 keywords. Next, do a search on "music artists" and gather a list of every artist you can - pop, rap, country, everything. There's got to be at least 2000 artists out there. Next write a quick script to append each artist to each keyword.

Next do a search for every album and song. There you'll find another 15,000 sources of keywords to combine with your ringtones keywords.

Now lets do the math. (3000 x 2000 = 6 million keywords) + (3000 x 15000 = 45 million keywords) for a grand total of 51 million keywords.

2/3 of which hardly anyone else is targeting.

Theres your long tail plain and simple. And yes, there is plenty of money in the tail.

Hope this helps,

StartLaunch
startlaunch is offline   Reply With Quote
Old 09-13-2006   #3
vayapues
10 kinds of people in the world. Those who know binary numbers, and those who don't
 
Join Date: Jan 2006
Location: Salt Lake City
Posts: 322
vayapues is just really nicevayapues is just really nicevayapues is just really nicevayapues is just really nice
Marketing on the longtail is only important if you want to increase your conversion rate, while decreasing your costs.

Otherwise, I wouldn't worry about it.

My personal experience is that folks are more likely to purchase if they come to your site from a long tail search, because they are further along in the buying process, and because you have exactly what they are looking for.
vayapues is offline   Reply With Quote
Old 09-13-2006   #4
startlaunch
Member
 
Join Date: Sep 2006
Posts: 21
startlaunch is on a distinguished road
Well said vayapues.
startlaunch is offline   Reply With Quote
Old 09-14-2006   #5
losloslos
Member
 
Join Date: Sep 2005
Posts: 37
losloslos is a jewel in the roughlosloslos is a jewel in the roughlosloslos is a jewel in the roughlosloslos is a jewel in the rough
The only problem I have found with going after tail terms is that the broad version of the keyword can steal the traffic of the longer tail terms. If you do expand your keyword list to include thousands of words i would suggest keeping some of the one or two word keywords phrases on exact match, that way it will allow the other terms to get the specific traffic.
losloslos is offline   Reply With Quote
Old 09-14-2006   #6
Northwind
Member
 
Join Date: Aug 2006
Posts: 31
Northwind is on a distinguished road
long tail tools

I understand what startlaunch says, had the pleasure of doing something like this myself. Are there other ways to build these word lists or tools other than pure shoe-leather? I guess much comes from a knowledge of your target market and how the words can be misspelled, combined. We're also pulling some of them right from our logs.

NW,
Northwind is offline   Reply With Quote
Old 09-15-2006   #7
jasper
Member
 
Join Date: Aug 2006
Posts: 7
jasper is on a distinguished road
Depends on the Search Engine

Here's my experience:

The longtail works well with Yahoo. Why? They show the exact term first regardless of bid price. They then supplement the remaining spots with broad matches.

For example, if you type in "5500 watt portable generator" into Yahoo. It will show your 10 cent STANDARD MATCH before showing someone else's 50 cent ADVANCED MATCH. This could change when Yahoo changes their system later this year.

For Google, it doesn't work so well. Why? They use a click thru rate in their calculation.

For example, you create an ad for "5500 watt portable generator" for Google. The ad is shown at best once a week or once a month. It simply doesn't generate enough traffic (i.e. clicks) to stand alone on its own.

In Google, you're better off finding multiple two-or-three keyword phrases (i.e. portable generator) and broad match them. Don't forget to strip out as many negative keywords as possible. This way, you'll generate plenty clicks to matter.

As for MSN, I'm still trying to figure this one out. Tried the long tail, but didn't get any traffic on longer terms. I'm trying the Google strategy now. Maybe it will work better. Not sure yet. I'd appreciate hearing from someone who has figured this one out.
jasper is offline   Reply With Quote
Old 09-15-2006   #8
imarketnc
imarketNC
 
Join Date: Aug 2006
Location: Morrisville, NC
Posts: 4
imarketnc is on a distinguished road
Jasper, question:

With Google, if you increased your spend for that long-tail term, wouldn't you be able to get around the low clickthrough rate? I thought it was a combination of both what you're willing ot pay and clickthrough rate. If you tilt the scales in one direction, wouldn't that help?
imarketnc is offline   Reply With Quote
Old 09-15-2006   #9
jasper
Member
 
Join Date: Aug 2006
Posts: 7
jasper is on a distinguished road
What's the Objective? Sales or Cheap Ads

From our perspective, we try to stay in the #3 position for all terms. It usually costs more to start in and keep that position. Once you secure a top level position, you can gradually reduce your bids because your CTR climbs.

If the objective is to spend as little as possible on a keyword, you can bid a dime on some obscure phrase.

Here's the problem:

It will never establish a high click thru rate because it will be buried on page three of search results. Other advertisers broad matches will perform better than your highly optimized exact match.

So, instead of creating 100 variations of the term "something widgit", you simply broad match (or phrase match) the term.

You're going to get some irrelevant traffic. That's why you need to strip out all of the negative terms. Get rid of "parts" or "service" or "repair" if you don't do it.

The key is to obtain qualified traffic that converts. No just cheap traffic.

When you first start with this strategy, you'll most likely pay way too much for the keywords. The reality is that your ads will be running for years.

As you establish a decent clickthru rate, you (a) get reduced pricing on your keywords and (b) get a high ranking resulting in lots of traffic.
jasper is offline   Reply With Quote
Old 09-17-2006   #10
MikeLevin
Mike Levin
 
Join Date: Jun 2006
Location: New York City
Posts: 7
MikeLevin is on a distinguished road
> Also, what products track these (other than maybe hitwise) or are you running inhouse software to pull these words. I've seen nothing, and consider passing it on to the IT guys.

Pulling your long tail keyword list is pretty easy with the right tracking software. The problem is qualifying the words in this potentially very long list. If you target terms that are already working well for you, you're wasting your time. If you target terms that are unlikely to ever happen again, you're wasting your time. The idea is to zero right in on the sweet spot in-between, and preferably on the terms that are right on the edge of performing much better for you.

There are free tools specifically built to do this. Follow the link in my profile for one of them.
MikeLevin is offline   Reply With Quote
Old 09-18-2006   #11
Wilksy
Ben Wilks
 
Join Date: Mar 2006
Location: Various locations around Australia
Posts: 120
Wilksy is just really niceWilksy is just really niceWilksy is just really niceWilksy is just really niceWilksy is just really nice
One of the best posts I have ever read

Brilliant. This one's total gold.

The Definitive Secret to SEO Revealed
http://www.searchengineforums.com/gazette/issue-216/
Wilksy is offline   Reply With Quote
Old 09-18-2006   #12
Northwind
Member
 
Join Date: Aug 2006
Posts: 31
Northwind is on a distinguished road
tools

Thanx Mike, thats what I was looking for - some references to tools that can do this type of thing. Again, for us, it comes down to (a) is there value in the words we can uncover, and it sounds like there is for specific engines, and (b) are there tools out there that can pull these for us, and I now have some ideas there too.

NW,
Northwind is offline   Reply With Quote
Old 09-18-2006   #13
fathom
Member
 
Join Date: Jun 2004
Location: Nova Scotia, Canada
Posts: 475
fathom is a jewel in the roughfathom is a jewel in the roughfathom is a jewel in the rough
Quote:
Originally Posted by Northwind
Thanx Mike, thats what I was looking for - some references to tools that can do this type of thing. Again, for us, it comes down to (a) is there value in the words we can uncover, and it sounds like there is for specific engines, and
Enormous value... "one customer at a time though". However, in my experience such long tail phrases tend to be "money term" - or putting it another way - "an easier sale".

A while back in the gifts industry I ranked 3 websites for 6 month in all 3 search engines for >Gift Baskets< e.g. #1, #2, #3... not a single sale was recorded for sites and few on the third - even though the traffic was through the roof... bearing that in mind that is an enormous loss in productive time. (while it may not be the same in all indsutries and segements and the website "could have beenthe problem" my bet is on "the searcher really didn't know what they wanted "searched broadly" found some ideas and back to the search engine... thus since it is far easier to target vast swaths of phrases - productivity on '1' here and '1' there isn't a major problem beyond wrapping your head aroun it.

Quote:
Originally Posted by Northwind
(b) are there tools out there that can pull these for us, and I now have some ideas there too.

NW,
A tool is a tool... it is best not to focus "only on the data being pulled from any". Connors Communication's tool HitTail is a value addition but every tool has its limitation.

Matching this with AdWords conversion tracking, Analytics (and other of the like), and log file analyzers that also associate queries to conversions - when used together they are powerful allies to not just finding terms easy to rank for but also limiting to those that actually convert sales (allowing everyone else to waste productivity on those that don't).
fathom is offline   Reply With Quote
Old 09-18-2006   #14
Marcia
 
Marcia's Avatar
 
Join Date: Jun 2004
Location: Los Angeles, CA
Posts: 5,476
Marcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond repute
I've got a hard time keeping up with content development and page creation just from eyeballing the keyword phrases used and watching traffic patterns, there's such a volume of information that can be gleaned from a hand check of the stats - even with just cpanel and awstats. Basically, I analyze my target audience and the way they think and search on a continual basis, and it's about 90% intuitive on my part.

Agree 100% with fathom about general keywords, something like "gift baskets" pulls a ton of traffic but isn't a high converting target like specifics are. I just went through having 3 listings #1, 2 and 4 for single keywords that would be highly coveted, at one of the engines, and it did nothing whatsoever for sales, which just about always come from the specifics. I've right now suddenly got a #5 for a 3-word keyphrase at Yahoo, actually the *best* keyphrase possible for that site, and it isn't bringing conversions, it's just bringing a lot of traffic, window shoppers who haven't refined their decision making process through to the buying stage in the sales process yet. It isnt the site, because the specifics do just fine.

With another site (a niche directory), there's a #2 and #3 at Google for the top keyphrases - it means nothing for selling stuff or making money.

When I find "sweetness" for search terms on a site, I build around that, and try to create content using my people's own mode of thinking and using language. Traffic means zilch without conversions (which is why I'm such a big fan of the Eisenbergs, BTW).

It isn't so much getting a prolific keyword list, IMHO - that's easy-peasy from simple stats if they're looked at intuitively. What I find is that it's the actual writing of the text on pages that pulls in a lot, by writing descriptive "copy" using language that's natural to my target audience - including descriptive words, adjectives and adverbs, etc. It's also learning to insert words into the copy that the engines can pick up and return to users' searches as phrases - with or without exact matching, and watching to see how well exact match is working with the algos at any given time. Sometimes less is more

Added: incidentally, I'm referring only to organic search, not PPC.

Last edited by Marcia : 09-18-2006 at 05:55 PM.
Marcia is offline   Reply With Quote
Old 09-19-2006   #15
Gooner151078
Online-Marketing-Manager.co.uk
 
Join Date: Sep 2006
Location: UK
Posts: 155
Gooner151078 is a jewel in the roughGooner151078 is a jewel in the roughGooner151078 is a jewel in the roughGooner151078 is a jewel in the rough
More longtail information?

As described above, relevant content will accentuate rankings for terms at the end of the tail, whilst PPC can be used as a quick win. It is also worth remembering that what might be your long tail term may well be another advertisers' key focus.

Incidentally, the longtail theory extends beyond the realms of online marketing. Check out info from Chris Anderson if you're interested.
Gooner151078 is offline   Reply With Quote
Old 09-20-2006   #16
Marcia
 
Marcia's Avatar
 
Join Date: Jun 2004
Location: Los Angeles, CA
Posts: 5,476
Marcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond repute
An important point is that if you don't have a phrase or word on the site already, it won't show up in referrers because, being non-existent, it hasn't pulled any traffic. Thus, the need for rich vocabulary usage to begin with, and that starts with knowing your target audience, which is more intuitive, right brain copywriting (and research) than left brain stats.
Marcia is offline   Reply With Quote
Old 09-20-2006   #17
egain
Member
 
Join Date: May 2005
Location: Lancashire, UK
Posts: 134
egain is on a distinguished road
Think Marcias spot on as far as mining the long tail from an on-page SEO perspective is concerned. I have evaluated a number of client and potential client websites recently and the number of people that miss out small tricks that would significantly increase the number of "longtail" terms the site can be found on is astounding.

It is very easy to get carried away chasing a number of "key terms", (PS not talking about me here - but there are those out there that chase traffic vs conversions) and lose focus of the bigger picture and the benefits that can be mined by utiliseing a number of other factors in order to increase conversions and ROI, such as long-tail. However the impact of fully evaluating and mining the longtail in the long term should not be undervalued

Last edited by egain : 09-20-2006 at 09:44 AM.
egain is offline   Reply With Quote
Old 09-20-2006   #18
MikeLevin
Mike Levin
 
Join Date: Jun 2006
Location: New York City
Posts: 7
MikeLevin is on a distinguished road
When we engage clients, we clearly split keywords into two groups Benchmarks and Actuals. Benchmarks are a limited number of strategically chosen keywords. These are the ones that we track positions (whether search hits occured or not). They're the ones brought up in business meetings and everyone likes to pat eachother on the back over.

Then, we break out "actuals", where we don't position-track unless a search hit actually occurred. Enough data is there to tell you which are the best candidates for natural search targeting. But its just as easy to get carried away with chasing the long tail once you catch the fever.

So, keep one foot planted firmly in the benchmark keyword world. Only allow 100 or so terms into this list, because it's "expensive" to track and of less value than often believed. And also keep one foot in the long tail. Consider tens-of-thousands of phrases if you can, but only chase those that you have good reason to think will produce traffic AND convert.
MikeLevin is offline   Reply With Quote
Old 09-20-2006   #19
fathom
Member
 
Join Date: Jun 2004
Location: Nova Scotia, Canada
Posts: 475
fathom is a jewel in the roughfathom is a jewel in the roughfathom is a jewel in the rough
Quote:
Originally Posted by MikeLevin
When we engage clients, we clearly split keywords into two groups Benchmarks and Actuals. Benchmarks are a limited number of strategically chosen keywords. These are the ones that we track positions (whether search hits occured or not). They're the ones brought up in business meetings and everyone likes to pat eachother on the back over.
Interesting... I use the analogy of halftail mining (which I think is what you are driving at)... there are many terms "useless for meaningful conversions" but when matched with associate terms (e.g. the rest of the tail) there is massive conversion value.

In much the same fashion the PPC managers use split run campaigns that same philosophy can be used in SEO strategies. (particularly useful for synonyms, regional, supplementary, complementary, and even off markets).
fathom is offline   Reply With Quote
Old 09-21-2006   #20
Northwind
Member
 
Join Date: Aug 2006
Posts: 31
Northwind is on a distinguished road
tail metrics

I'm also interested to understand what metrics if any, folks are using to evaluate tail terms. Sure we have frequency, but has anyone begun to factor other more traditional analytic KPI type metrics into term performance.

Knowing for example, that a certain set of my tail terms is more applicable to my heavy users, or that terms referring directly to my SKUs tend to have high rejection rates, leads me to treat these terms differently?

Thoughts?
NW,
Northwind is offline   Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off