The follow-up answers are readable but terse: each is a one-line response, while the supporting table and SQL are pushed out of the main surface into Full logs.
What was measured
Result Readability
Is the answer easy for a non-technical user to understand?
decisive for this rankingtransformation
The point is to get usable answers from non-technical users, so the result has to be understandable. (3 of 3 judges)
What was given, what came back
Test input: Best customers with unpaid-order and payment-method follow-ups · text · group: live-database-plain-english-queries
Input — what we sent
The exact prompt
Who are my best customers — the ones who order the most and spend the most? Follow-up 1: For the top 3 from that list — do any of them have unpaid orders? Follow-up 2: What payment methods do these top 3 usually use?
A conversational multi-table customer analysis with follow-up questions. It asks for the best customers by both order volume and spend, then drills into unpaid orders for the top 3 and their usual payment methods. Designed to test ranking logic, join-heavy analysis, and follow-up context retention.
Why this input is hard
- · ambiguous business term interpretation
- · multi-table joins
- · aggregation and ranking
- · follow-up context retention
- · scoping to a selected subset
- · payment behavior analysis
Output — unretouched
Also checked on this input — same tool, 4 other criteria
Ambiguity Handling✓ WorkedFaced with 'order the most and spend the most,' it did not silently choose one dimension; it built a combined equal-weight rank and explicitly surfaced the trade-off between all-around value and single-big-order spend.Business Insight✓ WorkedIt adds interpretation instead of only listing rows: Rahul is framed as the strongest all-around customer, and the tool warns that Deepak and Karan's 'usual' payment method is not yet a stable preference because each has only one order.Follow-Up Context✓ WorkedIt carried the top 3 customers through both follow-ups and, on the payment-method question, correctly did not inherit the main query's paid/non-cancelled filter.SQL Generation✓ WorkedIt generated CTE-based SQL that joins customers to paid, non-cancelled orders, aggregates order_count and total_spend, and ranks customers across both measures.
Provenance
- Observation
- ee22d062-e8e0-4ac9-8870-8cca0db9d8ec
- Evidence run
- af2abc96-3311-484b-a19d-854a2fdd2bf3
- Study
- Query Live Databases Using Plain English with AI
- Research task
- 86b9y6c99
- Tested at
- not recorded
- Source
- first-party
- Evidence state
- verified
- Proof shown
- input + output shown
- Cost / latency
- not captured
- Repeat run
- not captured
- Tester
- not captured
The last three rows are honest blanks, not placeholders — our capture has no field for them yet.
Query this
get_evidence({
tool: "dot",
scenario: "live-database-plain-english-queries"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, same check — 5 other tools
measured on Result Readability
AskYourDatabase✓ WorkedIt presents the answer as clearly labeled ranking tables and customer-level payment summaries, with visual risk cues for unpaid or shipped orders.Basedash✓ WorkedThe output stays readable for non-technical users, using compact tables and short summaries with explicit customer names and amounts.Definite✓ WorkedIt returns a ranked table with order counts, total spend, and average order value that is straightforward to read and scan.Draxlr⚠ StruggledAll three best-customer result tables keep customer_id as the lead column, so the output stays UUID-heavy and hard to scan.Querio◐ MixedThe answer is usable, but it is less readable than it could be because the output centers customer UUIDs instead of human names.
This evidence is published in
From the same study (page rebuilt from a later run)
Real inputs and real outputs, no retouching · every cell queryable via API & MCP · aidemos.com