Oct
9
Keyword Research For Recommendation Engines – Views
Filed Under Recommendation Engines | Leave a Comment
In the previous article, we discussed keyword research for PPC. Now, after covering the need for proper keyword research for SEO and PPC, we move on to recommendation engines.
Recommendation engines are services that try to present you information you will like. Kind of like a search engine, but based on user feedback. YouTube for instance, is a recommendation engine, as it values user rating. Other examples include: Digg, StumbleUpon, Reddit and many more. Most video stream sites frequently resemble YouTube, so we won’t bother addressing them separately.
Some recommendation engines have build in categories, while others use a tag based system. We will focus on the tag based ones, especially YouTube and StumbleUpon, for the simplicity of explanation. Of course keyword research is relevant to category based engines as well, especially since they usually have a search box option. Tags simply make for clearer examples, and the lessons are still useful for all recommendation engines, tag based or not. This article is about keyword research, so we will not discuss specific promotion issues, like making sure not to “discover” the same domain repeatedly on StumbleUpon, or YouTube’s upload policies.
Article Number 3 – Keyword Research for Recommendation Engines – Views (First Step)
We all know YouTube, and many are familiar with content sharing sites, such as Digg, but few understand the true power user opinion has in those sites. The key to all recommendation engines is that users rate content for quality. Thus, misleading users is extremely counter-productive. It’s true for all SEM generally, but with recommendation engine optimization, this issue becomes absolutely critical. Remember – the most important thing about recommendation engine optimization is making sure the users are getting exactly what they expect. Posting videos on YouTube with fake tags or titles may give you a wave of views, but it will end quickly and you will fail to get the most important thing – user approval. On Digg and StumbleUpon this means rating (diggs/thumbs up). On YouTube rating is less critical, but subscriptions play a vital role. Besides, even on YouTube ratings do count, especially if you want to get featured.
Why getting subscriptions is more important than getting views? Because subscriptions, favorites, or in case of Digg and StumbleUpon – friends, gives you long term exposure and authority. Your YouTube subscribers will see your future uploads on their YouTube main page, and your StumbleUpon subscribers are more likely to stumble them. In addition, StumbleUpon evaluates your opinion (stumbles) with your number of friends and followers in mind, as well as the success of your previous stumbles. Therefore – satisfying the users is most important.
Now that we know our objective, how can we accomplish it? We want to make sure that:
a) Our content matches the expectations of the users who view it.
b) As many people as possible, out of those interested in our content, get exposed to it.
This is the point where keyword research kicks in. We have the typical exposure vs. traffic relevance dilemma of keyword research, but the results fluctuations are higher, as our content can go all the way up, or all the way down in a matter of hours. This means we absolutely cannot afford mistakes. Before submitting content to recommendation engines, its marketing plan has to be fully formulated, with well defined target keywords. There is no place for changing strategy and targeting other keywords later on. SEO allows prolonged planning, with adaptions on the way. Recommendation engine optimization demands clear cut focus from the start to the finish. So basically, keyword research is the most dominant factor of all.
Obviously keyword tools that were designed for SEO and PPC are not suitable for recommendation engine optimization. Saying this may counter our interest, after all, Keyword Shack is a keyword research tool, and the previous articles put services like Keyword Shack in favor over free online tools, but it’s important to be clear on this: Keyword Shack is not designed for recommendation engine optimization. Our free tool, Combination Combination, can be used to save some time, though it still requires careful planning and thought.
Well, how do we make sure that users viewing our content get what they expected? Using relevant title, description and tags is critical of course, but not always trivial. For example, if you wrote an article about domain names, and post it on StumbleUpon, you might think that “domain” is a relevant keyword, but “domain” may also refer to real estate. On the other hand, if you look at the pages ranking top for “domain” tag on StumbleUpon, you’ll see they are relevant to Internet domains. So the users who surf the “domain” tag finds domain names relevant enough to “domain”. Checking StumbleUpon tag ranking is done by typing “http://stumbleupon.com/tag/[desired tag]” in your browser while not logged in.
Looking at the current ranking of a certain tag, or the description of similar high ranking content, is a way to learn about the taste of the users. But if you try to mimic top ranks directly, competition awaits you. So the best thing is to identify potential straight forward tags about you content, and also add a bunch of tags that are relevant to you, but distinguish the difference between you and your competition. This way, you can gain momentum that will help you take on the top ranks.
For example: We have a page with jokes that needs to be promoted on StumbleUpon. Niches like “funny” or “humor” are competitive, especially “humor”, because it’s also a category. Most of our jokes are actually black humor. The tags “black humor” and “dark humor” have traffic and less competition then “funny” or “humor”. StumbleUpon allows up to 5 tags, so we will choose: “black humor”, “dark humor”, “funny”, “jokes” and “humor”. Now we have a fair chance of ranking high for black and dark humor, get traffic from StumbleUpon, and acquire more positive stumbles. With good feedback, our page will move on to dominate “funny” and “jokes”, and eventually it even has a fighting chance for “humor”. Of course, this is only true assuming we actually have good jokes, and that we do our promotion right.
The key to using tags correctly is understanding what tags bring traffic that will give you good feedback, and then choose a combination of top target tags, easier to dominate tags, and middle way tags.
With the title, however, there is much less place for innovation, as it is limited in length, and must describe your content very accurately. Never lie in your title! Placing the wrong tag gives you bad feedback from that tag’s traffic. Writing the wrong title will piss off all your future traffic. Writing a good description is a combination of properly describing your content, while using promotional, keyword driven language. This is typical SEO-ish stuff.
So far we covered the basics of one submission, stuff that many of you already know. The next stage is gaining followers or subscribers. What does keyword research has to do with it? Apparently, a lot. The part of keywords doesn’t end with a success of a single submission, more about it in our next article: The second step of keyword research for recommendation engines – followers. Subscribe to the RSS feed to be among the first to read it.
