When you use a search engine, what are you hoping to find? A page that uses the keywords you've entered, or a page that answers your question and leaves you better equipped to do what you're trying to do?
If you spend a lot of time writing for search it's easy to forget that keywords are just a tool that search engines have developed to connect people with useful content: people aren't interested in keywords, they're interested in useful information, whether or not it uses the exact same phraseology as their search.
The fact is that people put one thing into a search box but might actually want something different: if I search for 'sports holidays in Mexico', a page about 'Cancun diving vacations' might be exactly what I'm looking for, without including a single one of my search terms. It's obvious to us that these two are related but it's very difficult to model computationally. Techniques such as latent semantic indexing attempt to improve the situation, but as yet search engines still aren't very good at telling useful information apart from filler.
SEO gone bad: an example This means that it's easy to produce a high-ranking page that features the keywords I've searched for and is superficially about' my subject, but that offers little or no valuable information. Here's an example, (right) which appears in the top 10 results of a Google search for Technics digital pianos'.
The headline is promising: 'Choosing a good Technics piano.' It's a reasonable assumption that if I search for 'technics digital pianos' I do want to know more about what makes a good one. Unfortunately that's roughly where the good stuff ends.
There is a listing of model numbers, but no way of telling them apart: we get simply 'Popular Technics digital pianos include...' How do I choose between them? Well, depending on your budget, you can choose one of the above Technics digital pianos' There's no information on which are more expensive or where they fit in my hypothetical budget.
Is there anything else I should know?
- You should also read customer reviews. That way you'll know what various owners think of different pianos.'
- If you're thinking of buying a digital piano, great sound should be a priority'.
Thanks for that. The problem with this page isn't really search-related. It's that it has little of value to say. The publisher's strayed from their area of expertise, where they genuinely have some insight and can add some value, and written something that could have been put together by almost anyone who understands the words technics digital piano'. It's as if most of the text is only there to link strings of keywords together in whole sentences. This is repeated for other brands of piano throughout the site:
'Kawai digital pianos are ideal for the stage, home, school and church. You should feel delighted by the amazing features that they're loaded with, no matter what your needs are.'
'It's obvious that there's a wide range of digital Roland pianos to choose from. You should be able to find one online very easily, as many musicians find it more convenient to shop online for their musical instruments.'
... and so on. This is a shame because the site does contain some really useful information: even on these pages, there are nuggets here and there. But they're buried under mountains of search-junk which drags the whole site down with it. This would be a much better site if it had much less content.
Search engines have tried to solve the problem of information overload, only for it to reappear at the next step in the user's journey. What I really want as a searcher is not to see my keywords reflected back at me. I already know my keywords! I want a page that takes them as a starting point and then adds to my knowledge or helps me complete a task.
The whole process of keywords > search > results page > clickthrough is only the start of giving your users what they want: if you don't have anything to say when they get to your page, it shows.