This session will address how marketers utilize the long tail of the Internet and search engine marketing to identify and reach consumers who are interested in what a company has to offer, but don't fit the traditional definition of the demographic audience. The end result is a more strategic, cost-effective search marketing campaign. We'll discuss the "long tail approach" from a tactical standpoint, including strategies that fit the users of this emerging market, as well as examples of why this works in terms of increasing sales.
Highlights of the presentation include:
- Keywords: Capturing consumers with tail
- The power of secondary search engines
- Case study examples of the concept "in action"
- Mary Bowling, Senior SEO, SEO Blizzard Internet Marketing, Inc.
- Brock Purpura, CEO, Etology
- Aaron Shear, Partner, Boost Search Marketing
- Stephan Spencer, Founder & President, Netconcepts, LLC
Familiarize yourselves Chris Anderson's long tail theory. Read the book. In sum, the majority of actions in our case take place on a small number of keywords, and keywords that are more specific and broader.
Mary Bowling begins her presentation. Specializes in travel and hospitality, that's why local is so interesting to her, and we will be covering long tail local search. Local search is search made with the intention of finding something within a specific geographic location. Online local search has the intention of completing the action online.
30-40% of all searches are estimated to have local intent. Engines are committed to interpreting the intent and want to satisfy the query. Evidence is the 3 or 10 pack of local results.
What is long tail of local search? "Denver + plumber". Other queries include "leaky faucet", etc. which indicate looking for a plumber. Nearby towns, neighborhoods, and other identifiers such as area codes or exits are included. All these possible queries, are not frequent, but people are searching for them. Must add brands to the list. They might add "Kohler" or "Delta". All these keywords add up together to create this enormous long tail. Why is this important? Because competition is less, and cheaper on PPC. Easier to rank for in organic - for same reason - less competition. People tend to want the high volume terms. Long tail traffic can really add up, and is generally very targeted. All these factors make the long tail attractive.
An example - a PPC ad group - shows slide with nice numbers of conversions and CTR's. Tips: Group properly. Don't set and forget.
Optimization is relatively easily. Create a new page, link to it with long tail term. Often see results within weeks. Add relevant terms to title and text to existing pages. Place term on a relevant page with good PR. Not as powerful. Best way to optimize is to think of categories to put long tail terms. Examples include - services, products, brands, etc.
The traffic can really ad up. For some of her clients, it adds up to a significant portion of traffic. In one of her case studies, more than 40% of traffic came from the tail, and was really targeted. Conversion rates for this case study was below 0.5% on short tail terms, whereas the long tail terms did much better.
Tip: Use blogs to harvest long tail traffic. Quick and easy to put up posts, and don't need tech experience. Blogs get spidered quickly, and can rank quickly. The right plugins can help - such as ones by Stephan Spencer.
Use long tail terms in your listings in Google Maps, Yahoo! local, Localeze, etc. Find the ones that are best for you. Recently, Google added attributes to your business listing. They will allows you to name attributes such as "town served", "descriptions", etc. Take advantage of that.
Next up is Brock from Etology.
Will discuss long tail from ad network prospective. Loves this group because we are here early and ready to learn!
A spiel about Etology. Founded in '05. Self serve ad network. Serves 1 billion ads/day. Network of 20k sites.
Shows a slide with a triangle representing all the sites on the web.
Tier 1 sites are the Alexa 100's. These sides command CPMs of $5-$10. The tier 2 represents the Alexa 100 - 5000. CPMs of $1-$5 generally. Usually they have internal sales force or use agencies to sell inventory. Under tier 2 is the rest of the web. The millions of other sites. Dependant on ad networks such as Adsense. Major decrease of CPM rates. Some get in the pennies per CPM. This is where the long tail exists. Price drop between tier 2 and 3 is key factor of opportunity.
Why is called the long tail? Pulled real time data off their system. Plotted the data from largest to smallest website. Shows the traditional head curve with the long tail. Key point of this slide is to show the aggregated volume of the long tail can be greater than the head curve.
The opportunity? Ran tests in multiple verticals. One test was in the auto industry. Ran same ad on tier 1 site that's well known as well as a tier 2. Found 10-15 small tier 3 sites such as blogs that were virtually unknown. Ran the same ads across all sites and found that the performance was statistically the same across all tiers. Shows that you can hit your performance metrics at a fraction of the price. In this case, 5% of the CPM! There is large opportunity in this price gap to leverage.
What are the action items? How to leverage the long tail? Goal is to build a vertical micro network. Go to the tier 3 sites, and find a handful that are specific to your industry. Contact the site owner. Start small. Use a flat rate - try a small test period. Collect the data, and take the next step. If it makes sense, repeat.
Another strategy is to use a custom ad size. Even the small sites are using standard ad sizes because they use doubleclick, etc. An example would be a link in the nav bar. Another strategy is to prepay. Pay upfront is very appealing to publishers. Build up the trust with the publisher, so that you can get him as a reference, or visa versa. From there, take advantage of pricing gap, building the micronetwork vertical, and then you'll be able to leverage the long tail.
Thats all! Thank you.
Aaron Shear from Boost is up.
How many people here are in ecommerce? Large show of hands. How many are shopping engines? Smaller show.
It's interesting how this market (ecommerce) has not accepted SEO in their main events of day to day marketing. Most ecommerce sites he finds are reliant on paid search vs. organic.
Goes over long tail principals. Head term: "Printers". Fat Tail: "Wide format color laser printer", Long tail: model number.
Finds that most ecommerce sites find the most page abandonment in the head. If someone comes for "printers", they bounce in seconds.
Common problems with the tail. Overly descriptive product names. Sometimes the page might be "too targeted". A solution is to use multiple title variations for same product. You know which keywords convert from your logs. Try to automate the process of putting the right phrases to the pages.
Recommendations: look at the shopping engines, look at the logs, look at your paid search campaigns. Alot of the info is in front of you.
How many terms per page? Ideally 5 permutations at the top level. 5-20 per category. Again, logs provide great insight.
Make sure that the words are close to the top of the site. Focus on interlinking. Makes sure that the links to words are repeated on appropriate pages.
Having a hard time getting indexed? Shopping engines as supplement. Learn from the shopping engines - they tend to dominate the space.
Look at your affiliates. They can take advantage of ecommerce sites that have poorly named pages.
Optimal traffic mix. From an SEO view, over 90% of your traffic should be from tail terms.
Next is Stephan Spencer of Netconcepts.
Has lots of slides and warns us that we won't keep up. So go visit the power point URL to download it. http://netconcepts.com/learn/long-tail.ppt
Polls audience on what they want to hear from this presentation. Has too many slides so wants to narrow the presentation.
The long tail is difficult because need to focus on scalable tactics. Need to use automatic processes to avoid manual work.
Page Yield Theory: Looked at over a million terms, and found interesting trends. Put together an interesting study. The idea is that a large volume of visitors that have been crunched. 80-90% on average of pages that were crawled by spiders were not delivering visitors. This page yield metric could be performing but are not. They just sit there and do nothing for you. How do you increase the % to perform? Many do not measure this. This is a huge opportunity.
Shows case study of 20% yield rate which is pretty good. Goal is to increase that to capture more of the longer tail. Look at non performing pages. Look at pages that are being hit by crawlers and not users. Can't touch every single page of a large site, and this is one way to prioritize. Another metric is phrases per page. Focused on being broader in appeal. How do you get a greater penetration per page in terms of visitor activity? Add more phrases to the page. An example is to leverage user generated content to capture more of the tail. Very powerful if you can leverage it.
Tactics to increase long tail performance: Experimental approach to SEO, as well as an iterative approach. SEO is an experimental science where you need to take hypothesis and test and prove them.
"Thin slicing" - a term from Malcolm Gladwell's book "Blink". An art expert trying to check if a work is forged or real, needs to make a quick gut thinking decision. A good approach for title tag and header optimization. Shows a Wordpress SEO plugin for editing title tags where you can quickly make changes. Can even do this with URL's. In a MarketingSherpa study - a long url is a deterent to clicking. Really want to optimize this.
Optimize internal linking structure. How to best pass link juice across the site. Want a flat structure. Pagination kills SEO because you are basically spreading crawl equity too thin. Try experiments using nofollows on "view all" links or "next" links. Poor anchor text. Really critical that you get attribute navigation right.
That's all - we're out of time. Email email@example.com if you want the research report.
Contributed by Avi Wilensky of Promediacorp.