There are four areas in which I can help a business improve their online operations.
Each area is part of the cycle that all businesses should go through as they optimise. It is best to have the each of the earlier levels in place before focusing on the next; trying to optimise before a strategy is decided upon never works well. It is not necessary to start from scratch each time, so if you already know your strategy and have the measurement processes in place there is no need to go over old ground.
These projects identify goals for what a business should be doing online.
By answering questions like "should we maximise revenue or profit?" or "how do we separate ourselves from the competition?" the goals for how a website and online marketing channels should be helping a business can be stated clearly.
I do not like to answer questions like "which social media channels should we be using?" at this stage. Instead it is better to answer "how can social media be used to meet business goals?" and then answer further questions through testing during the optimisation phase. Doing things this way relies less on making guesses and more on figuring out the way the world actually is. I am a fan of being lean where possible.
A good strategy will make the goals for a website clear. Measurement projects enable strategic goals to be monitored using web (and other) analytics.
Sometimes this can be quite easy; most web analytics tools track things like revenue without too much difficulty. Often it can be much more complicated, especially for business models that don't rely on lead generation, CPM advertising or ecommerce (all of which are easy to track in the main tools).
Analysis of appropriate data is the key to producing actionable insight for optimisation.
The nature of a good analysis can be wide ranging so it is hard to simply define them. Instead, it is easier to state the qualites necessary of the analyst:
I am well suited to analytics projects involving web analytics data, particularly those associated with search as this is where most of my experience lies. I will also take on projects involving interesting data sources (like social graphs).
All of the other stages are useless without action. The key to optimising items identified during analysis is to get it done quickly so that the cycle can begin again.
Common areas of optimisation that I am good at are as follows: