Understand your reports with Amanda AI

Reading a report by hand works just fine — we cover that in interpreting reports and metrics — but arrobaMail can do something better: read it for you. Amanda AI looks at all your numbers, cross-references them, spots what worked and what didn't, and lays it out in plain language with concrete recommendations. In this guide you'll learn to get the most out of that analysis and, above all, to have a conversation with it.

The underlying idea: you don't need to be an analyst to make good decisions. You need to ask the right questions, and Amanda helps you answer them using your own data.

1. Wait for the analysis (at 7 days)

Unlike traditional metrics, which you see instantly, AI analysis shows up seven days after sending. Why? Because opens and clicks take time to accumulate — analyzing two hours in would give you an incomplete picture. Once that week passes, in your campaign's AI Analysis tab, the interpreted report is waiting for you.

2. Read the score and the summary

The first thing you see is a score from 0 to 100 that sums up the campaign's health, along with an executive summary of a few lines that tells you, in plain language, how it went. That's the headline: if you have thirty seconds, this alone tells you whether the campaign performed well or needs attention.

3. Look at the four dimensions

The analysis breaks your campaign down into four dimensions, each with its own rating (Good, Fair, or Needs Attention):

• Deliverability: did your emails reach the inbox?
• Engagement: did people open and click?
• Content quality: did the message and the call to action work?
• List quality: is your base healthy, or is it dragging bad contacts along?

This tells you where the problem — or the strength — actually is. A campaign can have flawless deliverability but weak content: the breakdown puts it right in front of you.

4. Follow the prioritized recommendations

Here's where the real value is: Amanda doesn't just diagnose, it tells you what to do, with actions ranked by priority (High, Medium, Low) and the benefit you can expect from each. Start with the High-priority ones — they move the needle most for the least effort.

Tip: don't try to apply all ten recommendations at once. Take the first one, or the first two, for your next campaign, measure the result, and keep going. Changing everything at the same time makes it impossible to know what actually worked.

5. Ask about whatever you don't understand

The report isn't static — you can talk to it. If something doesn't add up, ask Amanda and it answers based on your own data. A few questions that genuinely help:

• "Why did clicks drop on this campaign?" → it identifies whether it was the CTA, the number of links, or the content.
• "Why wasn't this particular email delivered?" → it explains whether the address was blocked, had a prior bounce, or is on a blocklist.
• "Which segment of my audience is the most valuable?" → it cross-references device, location, and behavior to show you.

It's like having an analyst next to you who already knows your account and answers instantly.

6. Apply it and measure again

The analysis is only worth what you do with it. Pick one concrete change from the recommendations, apply it to your next campaign, and once the new analysis is in, compare: did the score go up? Did the dimension that was flagged for attention improve? That cycle — analyze, apply, measure again — is what improves your results campaign after campaign.

Common mistakes to avoid

• Ignoring the analysis because "you already saw the metrics." Metrics tell you what happened; AI analysis tells you why and what to do about it. They're not the same thing.
• Applying everything at once. Without isolated changes, you can't tell what actually worked. One per campaign.
• Not using the chat. It's the most powerful part and the most underused. Ask — that's what it's there for.
• Treating the score like a grade. It's not there to make you feel good or bad — it's there to show you where to improve.

Next steps

1. If you want to understand each metric by hand, see interpreting reports and metrics.
2. Put what you've learned to use and let Amanda build your next campaign in using Amanda AI.
3. Take your analysis further by connecting your account to an AI assistant in connecting arrobaMail with Claude (Cowork).