The following strategy increased by 40% the volume of organic clicks from Google in a matter of 3 weeks, following a long, steady decline:
Do you have an analytic strategy (success story) to share with us? Feel free to post it here.
Related article: 10+ Great Metrics and Strategies for Email Campaign Optimization
This is a mixture of tactics, and without a control (very difficult to do as you obviously cannot use the same content, you cannot measure all of the competing websites activity, and you do not know what Google has changed - the environment is changing during the experiment) you cannot be sure which of these made the difference, if Google have just tweaked a setting that changed the rankings negatively in the past, and then reversed it or tweaked something else that caused the 40% uplift. Did your tactics make a difference, and if so, which one(s) were worth the time/money investment? Was it just chance, or random noise in the >200 factors in the Google ranking algorithm? This is why SEO has such a bad reputation - because "experiments" with no controls are really just opinions, and often do not work when applied by anyone else (sometimes at great expense in SEO fees, and certainly expensive in wasted time).
Only Google knows what will increase your website ranking, and they have a vested interest in keeping that knowledge out of the hands of people who will use to manipulate rankings. The only way to be sure of getting traffic from Google in the search results, is to buy it via Adwords. Much of SEO is an exercise in self delusion (in my opinion - no statistics here to validate this argument!).
Google will change the weightings on the factors you used as soon as it sees them returning results not favored by the humans visiting the websites - so publishing anything that works is almost by definition ensuring that it's utility has an even more limited lifespan.
The best way to increase your Google search rankings long term is to do create what Google's search users want - so craft useful content that your readers want to share, comment on and respond to. The more time they spend on your page versus a competing page in the results, and the less people return from your website to Google to continue and/or refine their search, the better you will rank in Google. If you ran Google, isn't this what you would measure, and use to rank websites - using customers behavior to drive your algorithm optimisation? Whatever their customers want Google is determined to provide, to maintain their ever shaky competitive advantage. If they don't do this, then one of their current or future competitors will.
The above suggestions are proven tactics, and there is a straightforward explanation as to why they work. Data science (including SEO) is not just about data analysis and multivariate testing, but also discovering stuff that works without even looking at the data, a bit like some stock traders who "feel" the market and compete with pure geeks that apply techniques learned during their Ph.D. at Carnegie Mellon.
To put it differently, if you have opinions / visions that consistently work and don't have to spend hours implementing and testing them on real data, then you are more efficient (that is, you cost much less money to your employer) than the guy doing all the lengthy tests and bringing the same improvements.
If your opinions are consistently right, then I'd still be inclined to test why - understanding the underlying mechanism responsible for the success opens the door to improving performance. Unfortunately with SEO, most tactics that succeed are short lived, as the environment adapts to counter your good fortune (either google changes the weightings of the factors that lead to your success, or your competitors take advantage of the same "loopholes"). The stock market suffers from a similar problem - the environment responds to the trades, as other traders either follow or counter your trades. This is one of the (many) reasons the stock market is so susceptible to crashes, and why economics is not a science that can use mathematics effectively to predict the future effect of any strategy - the people affecting the strategy, and the markets response are in a constant feedback loop - so the cause and effects cannot be disentangled.
Yes, to consistently succeed, you must constantly adapt and stay in sync with the market (stock trading) or with Google (SEO). It's easier if you use robust techniques less likely to be influenced by daily changes in Google's algorithms.
Thanks for the quality posts, Vincent and Salvatore...
In my experience, SEO is certainly a "moving target"...and one that requires constant adaptation. Google has (and will continue) to prune their algorithm to produce a quality product, meaningful search results.
My tactic is to find and hire a good SEO consultant. So how do you find one? Easy, just search Google for "best SEO consultant". Whoever lands on page 1 for that keyword phrase must be doing something right. ;)