Your first report card
Your first Bot Report Card lands as an email link. This page walks down it from top to bottom so you know what you are looking at, using a real card as the example.
The card is a snapshot. It describes the dataset you uploaded and the period it covered, and it does not change afterwards.
The header
The title combines your bot's name with the period covered, for example Banking Demo 2 (Aug 18 release) (Mar 23 – Apr 22, 2025). The pencil icon next to it lets you rename the card. Worth doing if you upload regularly: naming a card after the release it covers makes a list of cards far easier to read six months later.
The panel on the right restates the bot name, the industry you chose at setup, the period, and how many conversations were analyzed.
If your upload was larger than your plan allows — a longer period or more conversations — a note here says the data was truncated. The card then describes only the oldest part that fitted, not everything you uploaded. See Plans and limits.
The Inquio Score
The gauge is your overall score out of 100, with a letter grade beneath it. The color follows the grade, so a red gauge and an F, as in this example at 56, means the assistant is not doing its job.
One number cannot tell you what to fix, which is why the card continues.
The four dimensions
Directly beneath the gauge are the four components of the score:
| Dimension | Example | Weight |
|---|---|---|
| Resolution accuracy | 47 (F) | 40% |
| Containment quality | 55 (F) | 25% |
| Dialog fluency | 97 (A) | 20% |
| Safety and compliance | 28 (F) | 15% |
This example is worth reading carefully, because the pattern is common. The assistant scores 97 on dialog fluency: it converses smoothly, handles turns well, and reads as competent. It scores 47 on resolution accuracy and 28 on safety and compliance, meaning it is fluently telling people the wrong thing and behaving in ways it should not.
A high fluency score alongside low accuracy is the most misleading combination in the product, and the reason a single overall number is not enough. Read the tiles before you draw any conclusion from the gauge.
Select any tile for the detail behind it.
Defender and Sales Radar
Two preview boxes show what was found outside the score itself:
- Defender — safety incidents and attempted attacks. In the example, 117 incidents and 78 attacks.
- Sales Radar — business leads and churn signals. In the example, 12 leads and 65 churn signals.
These are counts only. Learn more opens the full screens.
Summary
A written account of what Inquio found, naming the strengths and the problems and putting the numbers in context. Read more expands it.
Read this before the issues list. It explains why the score came out where it did, which the issues on their own do not.
Identified issues
The most impactful problems found in your conversations, each with a severity badge, a title, and a short description of the pattern. See details opens the issue, the conversations behind it, and the recommended fix.
The button at the bottom reveals the total. In the example, three issues are shown out of 19 found.
Issues are ordered by impact, not by how alarming they sound. Start at the top.
Benchmark comparison
Your scores against the average for your industry. In each pair, the colored bar is you and the grey bar is the industry.
In the example: overall 56 against 70, resolution accuracy 47 against 68, containment quality 55 against 71, dialog fluency 97 against 74, and safety and compliance 28 against 69. One dimension well ahead of the field, three well behind.
This is the section that tells you whether a mediocre-looking score is actually mediocre. Some dimensions run lower across a whole industry, and being level with your peers is a different situation from being 40 points behind them.
Recommendations
Specific, actionable fixes, written against what was found in your data rather than general advice. Each names what to change and why.
See all issues and recommendations opens the full list.
What to do first
- Read the summary.
- Look at which of the four dimensions is weakest, weighted by its share of the score. A 15-point gap in resolution accuracy costs you more than a 30-point gap in safety and compliance.
- Open the top issue and read five or six of the conversations behind it. The pattern is usually narrower than the issue title suggests.
- Act on the recommendation, upload fresh data after the change, and compare.
Resist fixing the thing that looks worst. Fix the thing that costs most.
What a first card cannot show
Trends. A trend arrow compares your card against the immediately preceding interval of the same length: a card covering 30 days is measured against the 30 days before it. On a first card those preceding days are not in Inquio yet, so there is nothing to show. Arrows appear from your second upload onward.
Defender may be empty, and that is a good result rather than a missing feature. See Defender.
On the Starter plan
The card is complete, but the links out of it are not available. See details, See all issues, Learn more on the Defender and Sales Radar boxes, and the conversation count all prompt an upgrade to Growth or Enterprise.
You still see your score, all four dimensions, the summary, the benchmark comparison, and three issues with three recommendations. What you cannot do on Starter is open the conversations behind any of it.
See Plans and limits.
Sharing the card
The card can be downloaded and shared from the buttons at the top, which is the usual way to get it in front of people who do not use Inquio themselves.
Next steps
- Bot Report Card — the full screen reference
- Inquio Score™ — how the score is calculated
- Diagnose what's wrong with your bot