At a minimum, you should be able

[kexej28769@nongnue]

Do: Make sure you have data.
Ask the client what they’re using to monitor performance. If they don’t have the basics, suggest setting it up or incorporate it into your proposal as a first step. Define the data you need to analyze the site by asking the client about their goals, working with them on how to measure those goals, and then determining the tools and analytics setup you need. Those essentials might include:

Webmaster Tools setup. I like to have at least Google and uae number data , so I can compare across all search engines to help determine if there is a spike or drop in both search engines, which could indicate the cause is due to something happening with the site, or just in one search engine, which could indicate the cause is algorithm related.
Organic search engine traffic. to see organic search traffic by page type (for example: service pages vs. product pages). Ideally, you can also filter by things like URL structure, country, date, referrer/source, and be able to run regex queries for granularity.
User testing and focus groups. Optional, but useful if available and can help prioritize. For example, focus groups on one site showed us that people are more likely to convert if they can see a certain type of content that is not necessarily SEO-friendly. If they are more likely to convert, they are less likely to return to the search results, so adding previously less-prioritized content can have a double benefit for the site: more conversions and a lower bounce rate back to the SERPs.
Don't: Make empty promises.
Simply put, please, SEOs, don't make any promises. Hopeful promises make SEOs sound like snake oil salesmen. This is a real problem for all of us, and you can help change it.