I always find its hard to start when you are stuck only looking for the best solution - just get everything on the screen and start, don't worry about the end p

I always find its hard to start when you are stuck only looking for the best solution - just get everything on the screen and start, don't worry about the end product till you know all of what you have to deal with. Overload your test pages with sample data (max pictures, lots of "results", full detail forms, etc.). Now that you really have your problem defined. Time for fun! - refine HTML and CSS, keep it simple, If you are doing responsive stuff start with the smallest display mode possible. Shift-control-m in Firefox gives you a small screen simulation to start but also test on an actual phone/tablet browser to see how it works in the real world. As you hit snags, check out what other sites do to handle such situations, note what you think works and what you could do better then implement your spin on such things. Don't be afraid to switch approaches, make an archive/backup and re-factor as needed. The more experience you get doing it the better.