In this article Loren McDonald tells us that a good email report shows where you’ve come from, what you’ve done
along the way, and clearly highlights where you want to go. A good
email report is like a map, showing you the specific turns, starts and
stops necessary to drive you toward success.

So what are the ingredients of a good email report?

1. Benchmarks. First, understand what your specific
benchmarks are, and, if applicable, how those compare to others in your
industry vertical. What benchmarks do you want to work from and try to
improve over time?

2. Response Models. How do customers typically
respond to your email messages? Do you get 90% of your conversions in
the first 48 hours, or is your product/service such that longer-term
consumer thought is needed? Defining your response curves and
understanding them as being fluid — ever-changing and in need of a
watchful eye — will make sure that when you pull data, you’re getting a
clear and full picture of the true impact of your program(s).

3. KPIs. Define your key performance indicators, or KPIs. David Baker wrote an article
on this a year ago, and I’d suggest you read it for a good
representative sampling of typical KPIs. In general, these are the
metrics that are important to your specific business. If you are
looking to create brand awareness or engagement, your indicator might
be opens, clicks or referred friends. If you want to drive sales, then
you want to look at conversion rate, revenue and cost per conversion.
Whatever these metrics are, they should be actionable and mean
something — both to those who manage the programs, and the higher-ups
who need to buy into your efforts.

4. Putting it together. It is important to measure
performance campaign over campaign. If you don’t have a historical
reference of your email campaign performance, please start immediately!
This can be as easy as creating an Excel spreadsheet where you can
track mailing over mailing. Keeping a record will allow you to see
overall performance trending and understand seasonality, list hygiene
issues and/or deliverability red flags.

Once you have your mailing-over-mailing scenario set up, you can
start to create pivot tables as a way to sort the data and look at more
specific bits of information. Over time you can look at data dimensions
like:

• Segmentation. How do different customer segments do
  over time? How does their performance change based on the messaging
  they receive? How do different segments grow or shrink over time? What
  is your most valuable segment?

• Messaging. What messages resonate with particular segments? What are your "winning" offers, CTAs, general messages?

• Creative. Which templates work best, what format?
  How do particular users interact with emails – where do they click,
  what behaviors are created?

Source: Email Insider