compare runs with trend analysis.md

Compare runs with trend analysis

This article explains how to compare recent runs in Leapwork Go using Trend analysis. Use it to understand whether performance is improving, staying stable, or regressing across comparable runs.

What this article covers

Generating a trend analysis for a sequence in a timeline
Narrowing the analysis by endpoint, region, metric, and recent-run count
Setting a baseline run for comparison
Interpreting the chart, summary cards, and comparison table
Exporting the analysis to CSV

Before you start

Open the Results page for a run in the project you want to analyze.
Make sure you can view run results for the project.
You need more than one comparable run to get meaningful trend data.

Filter options at a glance

Filter	What it does
Sequence	Selects the sequence in the timeline that you want to compare across runs.
Endpoint	Lets you compare the full sequence or focus on one endpoint inside that sequence.
Metric	Lets you compare Avg response time, P90, P95, P99, or Error rate.
Number of recent runs	Controls how many recent runs are included in the analysis.
Region	Lets you compare all regions together or focus on one region.

Generate a trend analysis

Open the Results page for a run.
Open the Trend analysis tab.
In Sequence, select the sequence you want to analyze.
Optional: in Endpoint, keep All endpoints or select a single endpoint.
In Metric, select the metric you want to compare.
In Number of recent runs, choose how many recent runs to include.
Optional: in Region, keep All regions or select one region.
Select Generate analysis.

Leapwork Go then builds a view of the selected metric across the recent runs that match your filters.

Set a baseline

After the analysis is generated, you can set a baseline run directly from the comparison table.

Find the run you want to use as your reference point.
Select that row in the table.

When you select a baseline:

the selected row is highlighted
the baseline is marked in the chart and table
comparisons update to show how other runs differ from that baseline
the banner highlights how the latest included run compares with the baseline

Use a baseline when you want to compare a new build against a known-good run instead of only comparing against the average.

Interpret the results

Area	What it shows
Trend chart	The selected metric across the included runs, so you can spot upward or downward movement over time.
Go banner	A quick statement showing whether the latest included run is faster, slower, or equal compared with the selected baseline.
Average card	The average value for the selected metric across the included runs.
Baseline card	The metric value for the run you selected as the baseline.
Std. deviation card	How much the selected metric varies across the included runs. Higher variation means the run results are less consistent.
Comparison table	Per-run details including date, metric value, change versus the previous run, change versus the average, change versus the baseline, and peak virtual users.

For response-time metrics and error rate, lower values generally indicate better performance.

Export the analysis

Select Export to download the current trend analysis as a CSV file.

The export includes:

run number
date
selected metric value
comparison versus the previous run
comparison versus the average
comparison versus the baseline
peak virtual users

If you change the filters, generate or update the analysis again before exporting so the CSV matches the view you want to share.

Tips for more useful comparisons

Compare runs for the same sequence under similar test conditions.
Use the Region filter when geography may affect latency.
Use a baseline run when you want a stable reference point for release-to-release comparison.
Select a single endpoint when you want to isolate one API instead of viewing the full sequence.

Troubleshooting

I see "No data available"

Try these checks:

confirm that the selected sequence has data in the runs you included
switch Endpoint back to All endpoints
switch Region back to All regions
increase the number of recent runs if too few runs are included

The comparison is not useful

Make sure you are comparing runs that are meaningfully alike. Large differences in traffic, regions, or the selected endpoint can make trend patterns harder to interpret.

I do not see the Trend analysis tab

Availability can depend on your workspace configuration and your access to run results. If the tab is not available, contact your workspace admin or Leapwork support.

My export does not match what I expected

Update the filters, select Generate analysis or Update analysis, and then export again.