TL;DR (takeaways for the impatient)
Improving website performance is not an exact science. Customers and prospects vary from industry to industry and from business to business. The answer to making an individual website perform better is to test different ideas and measure how visitors respond. To do this testing successfully, a tracking and analysis system must be in place.
While we make recommendations for changing your website based on advice from the search engines, noted experts and other contributors, as well as our own experience, we're not suggesting that you dump your existing website. Instead, we suggest implementing our changes as part of A/B testing. (see testing, below).
Like any business process, improvements come from observing behaviors, in this case, those of web visitors. Changes are not made on whims, but when we see that current offerings and methods are not working as well as we would like. Proper tracking and analysis tools let us know (in some cases) what a visitor was looking for when they arrived on our site. They also tell us when visitors (in most cases) return to our site, as well as what pages they visit, in what order, and how long they spend on each page. This can greatly aid us in arranging and adding content, references, and other aids to the buying process.
Some analysis tools, such as Google Analytics, provide you and your SEO/Usability consultant with the ability to test changes in layout, content, content architecture, design, and so on. In its simplest form, this is called 'A/B' testing. Instead of changing a web page, a second version of the web page is created with the desired changes. A test is set up with Google Analytics, which will automatically and randomly serve either the original page or the changed version of the page when a browser request is received. After a period of testing, you and your consultant can compare visitor activity on the two pages, and determine if the changed page should replace the original. In this way, your site can be continually tested for small and large changes, and let your visitors tell you what works best for them. When we work with a client on any of our optimization projects, we use this A/B testing to test all of our recommendations.
Usability.Gov (a service of U.S. Health and Human Services) on "Web Analytics Basics" http://www.usability.gov/what-and-why/web-analytics.html
Lindsey Nelson, Digitalistmag.com, "5 Reasons You Should Use Web Analytics" (11/15/12), http://www.digitalistmag.com/technologies/analytics/2012/11/15/5-reasons-you-should-use-web-analytics-021100
Content Marketing Institute, "12+ Ways to Use Web Analytics for Better Content Marketing" (8/27/15), http://contentmarketinginstitute.com/2015/08/web-analytics-content-marketing/
Sebastian Tonkin, Caleb Whitmore, Justin Cutroni, Performance Marketing With Google Analytics, Wiley Publishing, Inc., Hoboken, NJ (2010) ISBN: 978-0-470-57831-5