Selected answers from the Dumb SEO Questions Facebook & G+ community.
Michael Martinez: It`s impossible to do if you don`t record initial conditions. You must set some benchmarks so you know what you`re trying to improve and how much improvement there is.
I usually look at historical data to identify when there were seasonal trends or special events that affected traffic.
On the basis of that, I`ll watch for improvement in search referral traffic over prior similar periods.
I use only data from Bing Toolbox and Google Search Console. Third-party analytics data (including Google Analytics) is not reliable for accurate measurement.
You can also use raw server log data but that can be problematic, so not quite as accurate as the search engine Webmaster dashboards - but still more accurate than the third-party tools.
Ethan Lazuk: You first need KPIs that are relevant to your work and goals. You the SEO may care about keyword rankings, site traffic, map pack positions, or page speed. A business client might care more about conversions, like calls, contact form fills, or product sales. Maybe you’re trying to build a brand and get into the knowledge panel, or grow non-brand traffic with content marketing. KPIs vary based on your vertical, strategy, and goals. That said, if you have no KPIs yet, focus on tracking your data accurately (GSC, GA, Bing WMT) and recording site changes, then you can compare performance later when you choose metrics. Lastly, when comparing data, consider all the factors you can: site changes you’ve made, weekday vs. weekend traffic, seasonality, changes to SERP layouts or features, recent or historical core updates, new competitors in the market, global pandemics and radical shifts to user behavior(!), etc. Double lastly (and hugely important!), correlation doesn’t equal causation. Hope that helps!
What is your experiance with measuring impact of SEO initiatives? 
