Task-Boundary Mode
How task-instance boundaries are drawn from the event stream. Applies to every Task SoP, Step SoP, and Variants view.
Triage
📝 Task SoP
Mined from 694 instances
Review Processing Queue
How 7 analysts execute this task across 269 cases — auto-generated Standard Operating Procedure with adherence scoring.
79.8
Avg Adherence Score (0-100)
Median 82.5 · Range 53.3–90.0
20.7 min
Avg Task Duration
0.1 min median
1.4
Avg Sub-Steps per Instance
21.6% have multiple sub-steps
78.4%
Reference Path Coverage
Instances following mined SoP exactly
Reference Baseline
78.4% of instances follow this exactly
Review Processing Queue — Mined SoP
The most common sub-step sequence observed across 802 task instances. Use the dropdown on the right to compare any deviation to this baseline.
1
View Processing Queue
0.0s
Phobos
99%
Variant Explorer
Reference
Review Processing Queue Reference Path
629 instances · 7 analysts
This is the mined reference path — 629 instances across 7 analysts follow this exact sequence. Treat as the proposed Standard Operating Procedure for Review Processing Queue.
Genuine Complexity
Complex-Case Execution Path
23 instances · 7 analysts
· Δ adherence +10.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 90.0 vs reference 79.8 (delta +10.2). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
19 instances · 4 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Navigate to Cases
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
7 instances · 2 analysts
· Δ adherence -11.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Start Approval Workflow
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to collaboration.merck.com
7 instances · 4 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
collaboration.merck.com
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into collaboration.merck.com during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Teams
5 instances · 4 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Teams
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Microsoft Teams during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
4 instances · 3 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
View Validation Results
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Acrobat
3 instances · 2 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Acrobat
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Acrobat during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Excel
3 instances · 2 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Excel
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Excel during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Phobos
3 instances · 1 analysts
· Δ adherence -2.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Initiate Case Voiding
extra step (reference ends at step 1)
3
Start Action
extra step (reference ends at step 1)
Cross-App Swivel — 1 analysts pivot into Phobos during this task (0% of task time in non-primary apps, 33% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Outlook
2 instances · 2 analysts
· Δ adherence -12.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Outlook
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Outlook during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
2 instances · 2 analysts
· Δ adherence +2.7
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Add New Case
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 82.5 vs reference 79.8 (delta +2.7). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
2 instances · 1 analysts
· Δ adherence -17.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Navigate to Cases
extra step (reference ends at step 1)
3
Microsoft Teams
extra step (reference ends at step 1)
Cross-App Swivel — 1 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 67% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Application Mix
Which applications analysts touch while executing this task, measured by event count and active dwell time.
| Application | Events | Event Share | Dwell Share |
|---|---|---|---|
| Phobos | 836 | 74.4% | 58.7% |
| Veeva Safety | 204 | 18.2% | 41.3% |
| Microsoft Teams | 16 | 1.4% | 0.0% |
| Microsoft Word | 16 | 1.4% | 0.0% |
| Acrobat | 10 | 0.9% | 0.0% |
| Microsoft Excel | 10 | 0.9% | 0.0% |
| Microsoft Outlook | 8 | 0.7% | 0.0% |
| collaboration.merck.com | 7 | 0.6% | 0.0% |
| Notepad | 4 | 0.4% | 0.0% |
| acrotray | 3 | 0.3% | 0.0% |
Top Cross-App Transitions (within task instance)
Consecutive events that cross application boundaries within a task instance — the signal where swivel-chair patterns live.
| From | To | Transitions |
|---|---|---|
| Phobos | Veeva Safety | 107 |
| Phobos | collaboration.merck.com | 7 |
| Phobos | Microsoft Excel | 6 |
| Veeva Safety | Microsoft Teams | 6 |
| Phobos | Microsoft Teams | 6 |
| Phobos | Microsoft Word | 5 |
| Veeva Safety | Acrobat | 5 |
| Phobos | Acrobat | 4 |
| Phobos | Microsoft Outlook | 4 |
| Microsoft Excel | Veeva Safety | 4 |
On-Case Application Journey (±15 min around anchor)
Apps the same analyst touched on the same case within ±15 minutes of the task anchor. Captures work that spans task boundaries — the Veeva/Acrobat/Outlook/Word context surrounding the anchor click.
Case counts per app are not mutually exclusive — a case that touched Veeva AND Phobos in this window is counted in both rows. Overlap reflects real cross-app work, not double-counted cases.
Veeva Safety DOM 6,203 events 259 cases 96.7 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<span>
"Complete"
938×
pyzeClick
<a>
"Show more"
260×
pyzeClick
<li>
"Narrative"
233×
pyzeClick
<li>
"Details"
183×
pyzeClick
<span>
"Narrative"
170×
pyzeClick
<span>
"Approval Complete"
169×
pyzeClick
<span>
"Adverse Events"
159×
pyzeClick
<span>
"Start Approval Workflow"
138×
Phobos DOM 2,616 events 269 cases 72.6 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<button>
"Proceed"
464×
pyzeClick
<button>
"Advance"
281×
pyzeClick
<button>
"Save Comments"
250×
pyzeClick
<button>
"Send to process"
135×
pyzeClick
<select>
"Select Compound · - sitagliptin phosphate Tablet (G"
130×
pyzeClick
<select>
"Select Report Type · Legal · Literature · Medical Recor"
119×
pyzeClick
<a>
"Cases"
115×
pyzeClick
<div>
"Please reply to “Request Information" using dropd"
109×
Microsoft Word Document 320 events 111 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
file
V-Narrative 528191 (English) (26300681_0_1_338349)
382×
file
V-Narrative 540591 (English) (26495494_0_2_345212)
294×
file
V-Narrative 531284 (English) (26355806_0_2_340584)
225×
file
V-Narrative 538518 (English) (26459147_0_1_344628)
198×
file
V-Narrative 525091 (English) (26253532_0_1_338428)
197×
file
V-Narrative 530120 (English) (26331907_0_1_341282)
186×
file
V-Narrative 531868 (English) (26362311_0_1_340534)
185×
file
V-Narrative 542172 (English) (26517690_0_2_345412)
181×
Acrobat App 185 events 64 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
2,021 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Teams App 173 events 86 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
1,920 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Outlook Message 114 events 66 cases
Message-level capture
Task Mining captures email subjects. Case numbers frequently appear in the subject line.
subj
<compensación> Case#2252065
31×
subj
<Due today><Void request><GPVCM POLAND> Case # 2326062
23×
subj
<compensación> Case#2351734
17×
subj
Due today reports on 03-DEC-2025
14×
subj
Due today reports on 01-DEC-2025
13×
subj
<GPVCM Announcement> Promotions to Safety Data Manager positions in Poland
12×
subj
RE: <compensación> Case#2309077
8×
subj
Untitled
8×
Microsoft Excel Document 110 events 50 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
collaboration.merck.com App 92 events 51 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
Notepad App 37 events 22 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
usc-excel.officeapps.live.com Document 33 events 9 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
Adherence Vectors
Each task instance is scored on four behavioral vectors. Lower values are more adherent to the mined SoP.
0.10
Swivel Rate
events in non-primary app / total
0.97
Zero-Edit Visit Rate
inspection-only clicks
1.02
Avg Revisit Depth
max repeats of any activity
0.00
Cross-App Time Share
time outside primary app
Meta Score Distribution
Instance count in each 10-point adherence bucket. Right-skewed = healthy execution; long left tail = concentrated pain points.
0
10
20
30
40
50
60
70
80
90
Top Concern Vector
Zero-Edit Visit Rate — current value 0.97
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Vector Contribution to Adherence Loss
Each vector's weighted contribution to lost adherence. The top row is your highest-leverage fix.
Zero-Edit Visit Rate
value 0.97
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Swivel Rate
value 0.10
Reduce cross-app hops by surfacing the secondary-app data inline (API integration or pre-fetched context panel).
Revisit Depth
value 1.02
High revisit depth = mid-flow rework. Trace which sub-step is being repeated; usually a data-readiness gap upstream.
Cross-App Time Share
value 0.00
Time outside primary app dominates this score. Identify which secondary app(s) and whether their function can be in-lined.
Lowest-Adherence Instances
The 5 task instances that scored lowest. These are the concrete cases worth investigating to understand the worst patterns.
| Case | Analyst | Score | Duration | Sub-Steps | Swivel | X-App | Revisit |
|---|---|---|---|---|---|---|---|
| 2225437 | Analyst 13 | 53.3 | 22.3m | 9 | 0.89 | 0.00 | 2 |
| 2219900 | Analyst 02 | 53.8 | 1.6m | 8 | 0.88 | 0.00 | 2 |
| 2335834 | Analyst 09 | 54.3 | 1.9m | 7 | 0.86 | 0.00 | 2 |
| 2304250 | Analyst 13 | 55.0 | 0.6m | 8 | 0.75 | 0.00 | 3 |
| 2294929 | Analyst 03 | 55.0 | 3.0m | 6 | 0.83 | 0.00 | 2 |
Users on This Task
7 analysts executed this task during the pilot. Scatter plots productivity (instances per active day) against adherence (meta score). Top-right = healthy high-volume analysts; bottom-right = firefighters; top-left = careful low-volume; bottom-left = struggling.
Adherence (meta score) →
Productivity (instances / active day) →
03
13
01
12
09
05
02
User-by-User Breakdown
Click any analyst to see their dominant sequence vs the reference SoP.
| Analyst | Instances | Inst / Day | Adherence | Weakest Vector | Swivel | X-App | Min | |
|---|---|---|---|---|---|---|---|---|
| ▸ | Analyst 03 | 138 | 11.5 | 80.7 | Zero-Edit Visits | 0.07 | 0.00 | 847 |
|
Reference SoP
1
View Processing Queue
Analyst 03's Dominant Sequence
— 115 of 138 instances
1
View Processing Queue
This analyst exhibits 20 distinct sequences across their 138 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 13 | 114 | 10.4 | 80.2 | Zero-Edit Visits | 0.09 | 0.00 | 876 |
|
Reference SoP
1
View Processing Queue
Analyst 13's Dominant Sequence
— 88 of 114 instances
1
View Processing Queue
This analyst exhibits 18 distinct sequences across their 114 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 01 | 165 | 11.0 | 80.1 | Zero-Edit Visits | 0.09 | 0.00 | 1,349 |
|
Reference SoP
1
View Processing Queue
Analyst 01's Dominant Sequence
— 133 of 165 instances
1
View Processing Queue
This analyst exhibits 23 distinct sequences across their 165 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 12 | 85 | 12.1 | 79.9 | Zero-Edit Visits | 0.10 | 0.00 | 29 |
|
Reference SoP
1
View Processing Queue
Analyst 12's Dominant Sequence
— 68 of 85 instances
1
View Processing Queue
This analyst exhibits 11 distinct sequences across their 85 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 09 | 106 | 7.1 | 79.8 | Zero-Edit Visits | 0.10 | 0.00 | 4,760 |
|
Reference SoP
1
View Processing Queue
Analyst 09's Dominant Sequence
— 83 of 106 instances
1
View Processing Queue
This analyst exhibits 20 distinct sequences across their 106 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 05 | 80 | 7.3 | 78.5 | Zero-Edit Visits | 0.14 | 0.00 | 5,600 |
|
Reference SoP
1
View Processing Queue
Analyst 05's Dominant Sequence
— 58 of 80 instances
1
View Processing Queue
This analyst exhibits 16 distinct sequences across their 80 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 02 | 114 | 14.2 | 78.4 | Zero-Edit Visits | 0.14 | 0.00 | 3,173 |
|
Reference SoP
1
View Processing Queue
Analyst 02's Dominant Sequence
— 84 of 114 instances
1
View Processing Queue
This analyst exhibits 23 distinct sequences across their 114 instances; the dominant one is shown.
|
||||||||
79.7
Avg Adherence Score (0-100)
Median 82.5 · Range 50.2–90.0
20.7 min
Avg Task Duration
0.1 min median
1.4
Avg Sub-Steps per Instance
21.9% have multiple sub-steps
78.1%
Reference Path Coverage
Instances following mined SoP exactly
Reference Baseline
78.1% of instances follow this exactly
Review Processing Queue — Mined SoP
The most common sub-step sequence observed across 802 task instances. Use the dropdown on the right to compare any deviation to this baseline.
1
View Processing Queue
0.0s
Phobos
100%
Variant Explorer
Reference
Review Processing Queue Reference Path
626 instances · 7 analysts
This is the mined reference path — 626 instances across 7 analysts follow this exact sequence. Treat as the proposed Standard Operating Procedure for Review Processing Queue.
Genuine Complexity
Complex-Case Execution Path
22 instances · 7 analysts
· Δ adherence +10.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 90.0 vs reference 79.7 (delta +10.3). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
20 instances · 4 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Navigate to Cases
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
7 instances · 2 analysts
· Δ adherence -11.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Start Approval Workflow
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to collaboration.merck.com
7 instances · 4 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
collaboration.merck.com
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into collaboration.merck.com during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
4 instances · 3 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
View Validation Results
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Teams
4 instances · 3 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Teams
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Microsoft Teams during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Acrobat
3 instances · 2 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Acrobat
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Acrobat during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Excel
3 instances · 2 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Excel
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Excel during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Phobos
3 instances · 1 analysts
· Δ adherence -2.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Initiate Case Voiding
extra step (reference ends at step 1)
3
Start Action
extra step (reference ends at step 1)
Cross-App Swivel — 1 analysts pivot into Phobos during this task (0% of task time in non-primary apps, 33% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
2 instances · 2 analysts
· Δ adherence +10.3
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
Select Compound
reference at this step: View Processing Queue
2
View Processing Queue
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 90.0 vs reference 79.7 (delta +10.3). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Microsoft Outlook
2 instances · 2 analysts
· Δ adherence -12.2
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Outlook
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Outlook during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
2 instances · 2 analysts
· Δ adherence +2.8
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Add New Case
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 82.5 vs reference 79.7 (delta +2.8). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Application Mix
Which applications analysts touch while executing this task, measured by event count and active dwell time.
| Application | Events | Event Share | Dwell Share |
|---|---|---|---|
| Phobos | 837 | 72.8% | 13.4% |
| Veeva Safety | 221 | 19.2% | 86.6% |
| Microsoft Word | 18 | 1.6% | 0.0% |
| Microsoft Teams | 17 | 1.5% | 0.0% |
| Microsoft Excel | 12 | 1.0% | 0.0% |
| Acrobat | 10 | 0.9% | 0.0% |
| Microsoft Outlook | 8 | 0.7% | 0.0% |
| collaboration.merck.com | 7 | 0.6% | 0.0% |
| Notepad | 4 | 0.3% | 0.0% |
| acrotray | 3 | 0.3% | 0.0% |
Top Cross-App Transitions (within task instance)
Consecutive events that cross application boundaries within a task instance — the signal where swivel-chair patterns live.
| From | To | Transitions |
|---|---|---|
| Phobos | Veeva Safety | 109 |
| Veeva Safety | Microsoft Teams | 7 |
| Phobos | collaboration.merck.com | 7 |
| Phobos | Microsoft Excel | 6 |
| Phobos | Microsoft Teams | 6 |
| Microsoft Word | Veeva Safety | 6 |
| Veeva Safety | Acrobat | 5 |
| Veeva Safety | Microsoft Word | 5 |
| Microsoft Teams | Veeva Safety | 5 |
| Microsoft Outlook | Veeva Safety | 4 |
On-Case Application Journey (±15 min around anchor)
Apps the same analyst touched on the same case within ±15 minutes of the task anchor. Captures work that spans task boundaries — the Veeva/Acrobat/Outlook/Word context surrounding the anchor click.
Case counts per app are not mutually exclusive — a case that touched Veeva AND Phobos in this window is counted in both rows. Overlap reflects real cross-app work, not double-counted cases.
Veeva Safety DOM 6,203 events 259 cases 96.7 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<span>
"Complete"
938×
pyzeClick
<a>
"Show more"
260×
pyzeClick
<li>
"Narrative"
233×
pyzeClick
<li>
"Details"
183×
pyzeClick
<span>
"Narrative"
170×
pyzeClick
<span>
"Approval Complete"
169×
pyzeClick
<span>
"Adverse Events"
159×
pyzeClick
<span>
"Start Approval Workflow"
138×
Phobos DOM 2,616 events 269 cases 72.6 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<button>
"Proceed"
464×
pyzeClick
<button>
"Advance"
281×
pyzeClick
<button>
"Save Comments"
250×
pyzeClick
<button>
"Send to process"
135×
pyzeClick
<select>
"Select Compound · - sitagliptin phosphate Tablet (G"
130×
pyzeClick
<select>
"Select Report Type · Legal · Literature · Medical Recor"
119×
pyzeClick
<a>
"Cases"
115×
pyzeClick
<div>
"Please reply to “Request Information" using dropd"
109×
Microsoft Word Document 320 events 111 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
file
V-Narrative 528191 (English) (26300681_0_1_338349)
382×
file
V-Narrative 540591 (English) (26495494_0_2_345212)
294×
file
V-Narrative 531284 (English) (26355806_0_2_340584)
225×
file
V-Narrative 538518 (English) (26459147_0_1_344628)
198×
file
V-Narrative 525091 (English) (26253532_0_1_338428)
197×
file
V-Narrative 530120 (English) (26331907_0_1_341282)
186×
file
V-Narrative 531868 (English) (26362311_0_1_340534)
185×
file
V-Narrative 542172 (English) (26517690_0_2_345412)
181×
Acrobat App 185 events 64 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
2,021 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Teams App 173 events 86 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
1,920 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Outlook Message 114 events 66 cases
Message-level capture
Task Mining captures email subjects. Case numbers frequently appear in the subject line.
subj
<compensación> Case#2252065
31×
subj
<Due today><Void request><GPVCM POLAND> Case # 2326062
23×
subj
<compensación> Case#2351734
17×
subj
Due today reports on 03-DEC-2025
14×
subj
Due today reports on 01-DEC-2025
13×
subj
<GPVCM Announcement> Promotions to Safety Data Manager positions in Poland
12×
subj
RE: <compensación> Case#2309077
8×
subj
Untitled
8×
Microsoft Excel Document 110 events 50 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
collaboration.merck.com App 92 events 51 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
Notepad App 37 events 22 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
usc-excel.officeapps.live.com Document 33 events 9 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
Adherence Vectors
Each task instance is scored on four behavioral vectors. Lower values are more adherent to the mined SoP.
0.10
Swivel Rate
events in non-primary app / total
0.97
Zero-Edit Visit Rate
inspection-only clicks
1.02
Avg Revisit Depth
max repeats of any activity
0.00
Cross-App Time Share
time outside primary app
Meta Score Distribution
Instance count in each 10-point adherence bucket. Right-skewed = healthy execution; long left tail = concentrated pain points.
0
10
20
30
40
50
60
70
80
90
Top Concern Vector
Zero-Edit Visit Rate — current value 0.97
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Vector Contribution to Adherence Loss
Each vector's weighted contribution to lost adherence. The top row is your highest-leverage fix.
Zero-Edit Visit Rate
value 0.97
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Swivel Rate
value 0.10
Reduce cross-app hops by surfacing the secondary-app data inline (API integration or pre-fetched context panel).
Revisit Depth
value 1.02
High revisit depth = mid-flow rework. Trace which sub-step is being repeated; usually a data-readiness gap upstream.
Cross-App Time Share
value 0.00
Time outside primary app dominates this score. Identify which secondary app(s) and whether their function can be in-lined.
Lowest-Adherence Instances
The 5 task instances that scored lowest. These are the concrete cases worth investigating to understand the worst patterns.
| Case | Analyst | Score | Duration | Sub-Steps | Swivel | X-App | Revisit |
|---|---|---|---|---|---|---|---|
| 1535621 | Analyst 13 | 50.2 | 13.7m | 11 | 0.91 | 0.00 | 3 |
| 2225437 | Analyst 13 | 53.3 | 22.3m | 9 | 0.89 | 0.00 | 2 |
| 2219900 | Analyst 02 | 53.8 | 1.6m | 8 | 0.88 | 0.00 | 2 |
| 2335834 | Analyst 09 | 54.3 | 1.9m | 7 | 0.86 | 0.00 | 2 |
| 2304250 | Analyst 13 | 55.0 | 0.6m | 8 | 0.75 | 0.00 | 3 |
Users on This Task
7 analysts executed this task during the pilot. Scatter plots productivity (instances per active day) against adherence (meta score). Top-right = healthy high-volume analysts; bottom-right = firefighters; top-left = careful low-volume; bottom-left = struggling.
Adherence (meta score) →
Productivity (instances / active day) →
03
13
01
12
09
05
02
User-by-User Breakdown
Click any analyst to see their dominant sequence vs the reference SoP.
| Analyst | Instances | Inst / Day | Adherence | Weakest Vector | Swivel | X-App | Min | |
|---|---|---|---|---|---|---|---|---|
| ▸ | Analyst 03 | 138 | 11.5 | 80.7 | Zero-Edit Visits | 0.07 | 0.00 | 847 |
|
Reference SoP
1
View Processing Queue
Analyst 03's Dominant Sequence
— 115 of 138 instances
1
View Processing Queue
This analyst exhibits 20 distinct sequences across their 138 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 13 | 114 | 10.4 | 80.2 | Zero-Edit Visits | 0.09 | 0.00 | 895 |
|
Reference SoP
1
View Processing Queue
Analyst 13's Dominant Sequence
— 88 of 114 instances
1
View Processing Queue
This analyst exhibits 19 distinct sequences across their 114 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 01 | 165 | 11.0 | 80.2 | Zero-Edit Visits | 0.09 | 0.00 | 1,349 |
|
Reference SoP
1
View Processing Queue
Analyst 01's Dominant Sequence
— 132 of 165 instances
1
View Processing Queue
This analyst exhibits 23 distinct sequences across their 165 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 12 | 85 | 12.1 | 79.9 | Zero-Edit Visits | 0.10 | 0.00 | 29 |
|
Reference SoP
1
View Processing Queue
Analyst 12's Dominant Sequence
— 68 of 85 instances
1
View Processing Queue
This analyst exhibits 11 distinct sequences across their 85 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 09 | 106 | 7.1 | 79.8 | Zero-Edit Visits | 0.10 | 0.00 | 4,760 |
|
Reference SoP
1
View Processing Queue
Analyst 09's Dominant Sequence
— 83 of 106 instances
1
View Processing Queue
This analyst exhibits 21 distinct sequences across their 106 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 05 | 80 | 7.3 | 78.3 | Zero-Edit Visits | 0.15 | 0.00 | 5,612 |
|
Reference SoP
1
View Processing Queue
Analyst 05's Dominant Sequence
— 57 of 80 instances
1
View Processing Queue
This analyst exhibits 16 distinct sequences across their 80 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 02 | 114 | 14.2 | 78.2 | Zero-Edit Visits | 0.15 | 0.00 | 3,128 |
|
Reference SoP
1
View Processing Queue
Analyst 02's Dominant Sequence
— 83 of 114 instances
1
View Processing Queue
This analyst exhibits 25 distinct sequences across their 114 instances; the dominant one is shown.
|
||||||||
79.0
Avg Adherence Score (0-100)
Median 82.5 · Range 53.3–90.0
23.9 min
Avg Task Duration
0.2 min median
1.6
Avg Sub-Steps per Instance
34.4% have multiple sub-steps
76.1%
Reference Path Coverage
Instances following mined SoP exactly
Reference Baseline
76.1% of instances follow this exactly
Review Processing Queue — Mined SoP
The most common sub-step sequence observed across 695 task instances. Use the dropdown on the right to compare any deviation to this baseline.
1
View Processing Queue
0.0s
Phobos
99%
Variant Explorer
Reference
Review Processing Queue Reference Path
529 instances · 7 analysts
This is the mined reference path — 529 instances across 7 analysts follow this exact sequence. Treat as the proposed Standard Operating Procedure for Review Processing Queue.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
16 instances · 4 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Navigate to Cases
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Reviewer Sweep
Reviewer Verification Pass
11 instances · 5 analysts
· Δ adherence +5.7
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
3
View Processing Queue
extra step (reference ends at step 1)
Reviewer Sweep — a viewing-heavy pattern with 3.1 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 5 analysts across 10 cases.
Cross-App Swivel
Cross-App Lookup to collaboration.merck.com
7 instances · 4 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
collaboration.merck.com
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into collaboration.merck.com during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
6 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Start Approval Workflow
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
6 instances · 3 analysts
· Δ adherence +8.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 87.5 vs reference 79.0 (delta +8.5). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Microsoft Teams
5 instances · 4 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Teams
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Microsoft Teams during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
4 instances · 3 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
View Validation Results
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Acrobat
3 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Acrobat
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Acrobat during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Excel
3 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Excel
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Excel during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Phobos
3 instances · 1 analysts
· Δ adherence -1.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Initiate Case Voiding
extra step (reference ends at step 1)
3
Start Action
extra step (reference ends at step 1)
Cross-App Swivel — 1 analysts pivot into Phobos during this task (0% of task time in non-primary apps, 33% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Outlook
2 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Outlook
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Outlook during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
2 instances · 2 analysts
· Δ adherence +3.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Add New Case
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 82.5 vs reference 79.0 (delta +3.5). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Application Mix
Which applications analysts touch while executing this task, measured by event count and active dwell time.
| Application | Events | Event Share | Dwell Share |
|---|---|---|---|
| Phobos | 836 | 74.4% | 58.7% |
| Veeva Safety | 204 | 18.2% | 41.3% |
| Microsoft Teams | 16 | 1.4% | 0.0% |
| Microsoft Word | 16 | 1.4% | 0.0% |
| Acrobat | 10 | 0.9% | 0.0% |
| Microsoft Excel | 10 | 0.9% | 0.0% |
| Microsoft Outlook | 8 | 0.7% | 0.0% |
| collaboration.merck.com | 7 | 0.6% | 0.0% |
| Notepad | 4 | 0.4% | 0.0% |
| acrotray | 3 | 0.3% | 0.0% |
Top Cross-App Transitions (within task instance)
Consecutive events that cross application boundaries within a task instance — the signal where swivel-chair patterns live.
| From | To | Transitions |
|---|---|---|
| Phobos | Veeva Safety | 107 |
| Veeva Safety | Phobos | 11 |
| Phobos | collaboration.merck.com | 7 |
| Phobos | Microsoft Teams | 6 |
| Veeva Safety | Microsoft Teams | 6 |
| Phobos | Microsoft Excel | 6 |
| Phobos | Microsoft Word | 5 |
| Veeva Safety | Acrobat | 5 |
| Microsoft Excel | Veeva Safety | 4 |
| Phobos | Acrobat | 4 |
On-Case Application Journey (±15 min around anchor)
Apps the same analyst touched on the same case within ±15 minutes of the task anchor. Captures work that spans task boundaries — the Veeva/Acrobat/Outlook/Word context surrounding the anchor click.
Case counts per app are not mutually exclusive — a case that touched Veeva AND Phobos in this window is counted in both rows. Overlap reflects real cross-app work, not double-counted cases.
Veeva Safety DOM 6,203 events 259 cases 96.7 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<span>
"Complete"
938×
pyzeClick
<a>
"Show more"
260×
pyzeClick
<li>
"Narrative"
233×
pyzeClick
<li>
"Details"
183×
pyzeClick
<span>
"Narrative"
170×
pyzeClick
<span>
"Approval Complete"
169×
pyzeClick
<span>
"Adverse Events"
159×
pyzeClick
<span>
"Start Approval Workflow"
138×
Phobos DOM 2,616 events 269 cases 72.6 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<button>
"Proceed"
464×
pyzeClick
<button>
"Advance"
281×
pyzeClick
<button>
"Save Comments"
250×
pyzeClick
<button>
"Send to process"
135×
pyzeClick
<select>
"Select Compound · - sitagliptin phosphate Tablet (G"
130×
pyzeClick
<select>
"Select Report Type · Legal · Literature · Medical Recor"
119×
pyzeClick
<a>
"Cases"
115×
pyzeClick
<div>
"Please reply to “Request Information" using dropd"
109×
Microsoft Word Document 320 events 111 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
file
V-Narrative 528191 (English) (26300681_0_1_338349)
382×
file
V-Narrative 540591 (English) (26495494_0_2_345212)
294×
file
V-Narrative 531284 (English) (26355806_0_2_340584)
225×
file
V-Narrative 538518 (English) (26459147_0_1_344628)
198×
file
V-Narrative 525091 (English) (26253532_0_1_338428)
197×
file
V-Narrative 530120 (English) (26331907_0_1_341282)
186×
file
V-Narrative 531868 (English) (26362311_0_1_340534)
185×
file
V-Narrative 542172 (English) (26517690_0_2_345412)
181×
Acrobat App 185 events 64 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
2,021 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Teams App 173 events 86 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
1,920 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Outlook Message 114 events 66 cases
Message-level capture
Task Mining captures email subjects. Case numbers frequently appear in the subject line.
subj
<compensación> Case#2252065
31×
subj
<Due today><Void request><GPVCM POLAND> Case # 2326062
23×
subj
<compensación> Case#2351734
17×
subj
Due today reports on 03-DEC-2025
14×
subj
Due today reports on 01-DEC-2025
13×
subj
<GPVCM Announcement> Promotions to Safety Data Manager positions in Poland
12×
subj
RE: <compensación> Case#2309077
8×
subj
Untitled
8×
Microsoft Excel Document 110 events 50 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
collaboration.merck.com App 92 events 51 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
Notepad App 37 events 22 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
usc-excel.officeapps.live.com Document 33 events 9 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
Adherence Vectors
Each task instance is scored on four behavioral vectors. Lower values are more adherent to the mined SoP.
0.11
Swivel Rate
events in non-primary app / total
0.98
Zero-Edit Visit Rate
inspection-only clicks
1.17
Avg Revisit Depth
max repeats of any activity
0.00
Cross-App Time Share
time outside primary app
Meta Score Distribution
Instance count in each 10-point adherence bucket. Right-skewed = healthy execution; long left tail = concentrated pain points.
0
10
20
30
40
50
60
70
80
90
Top Concern Vector
Zero-Edit Visit Rate — current value 0.98
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Vector Contribution to Adherence Loss
Each vector's weighted contribution to lost adherence. The top row is your highest-leverage fix.
Zero-Edit Visit Rate
value 0.98
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Swivel Rate
value 0.11
Reduce cross-app hops by surfacing the secondary-app data inline (API integration or pre-fetched context panel).
Revisit Depth
value 1.17
High revisit depth = mid-flow rework. Trace which sub-step is being repeated; usually a data-readiness gap upstream.
Cross-App Time Share
value 0.00
Time outside primary app dominates this score. Identify which secondary app(s) and whether their function can be in-lined.
Lowest-Adherence Instances
The 5 task instances that scored lowest. These are the concrete cases worth investigating to understand the worst patterns.
| Case | Analyst | Score | Duration | Sub-Steps | Swivel | X-App | Revisit |
|---|---|---|---|---|---|---|---|
| 2225437 | Analyst 13 | 53.3 | 22.3m | 9 | 0.89 | 0.00 | 2 |
| 2219900 | Analyst 02 | 53.8 | 1.6m | 8 | 0.88 | 0.00 | 2 |
| 2335834 | Analyst 09 | 54.3 | 1.9m | 7 | 0.86 | 0.00 | 2 |
| 2304250 | Analyst 13 | 55.0 | 0.6m | 8 | 0.75 | 0.00 | 3 |
| 2294929 | Analyst 03 | 55.0 | 3.0m | 6 | 0.83 | 0.00 | 2 |
Users on This Task
7 analysts executed this task during the pilot. Scatter plots productivity (instances per active day) against adherence (meta score). Top-right = healthy high-volume analysts; bottom-right = firefighters; top-left = careful low-volume; bottom-left = struggling.
Adherence (meta score) →
Productivity (instances / active day) →
03
13
01
09
12
02
05
User-by-User Breakdown
Click any analyst to see their dominant sequence vs the reference SoP.
| Analyst | Instances | Inst / Day | Adherence | Weakest Vector | Swivel | X-App | Min | |
|---|---|---|---|---|---|---|---|---|
| ▸ | Analyst 03 | 117 | 9.8 | 80.0 | Zero-Edit Visits | 0.08 | 0.00 | 848 |
|
Reference SoP
1
View Processing Queue
Analyst 03's Dominant Sequence
— 96 of 117 instances
1
View Processing Queue
This analyst exhibits 20 distinct sequences across their 117 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 13 | 99 | 9.0 | 79.5 | Zero-Edit Visits | 0.10 | 0.00 | 876 |
|
Reference SoP
1
View Processing Queue
Analyst 13's Dominant Sequence
— 74 of 99 instances
1
View Processing Queue
This analyst exhibits 18 distinct sequences across their 99 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 01 | 143 | 9.5 | 79.2 | Zero-Edit Visits | 0.10 | 0.00 | 1,348 |
|
Reference SoP
1
View Processing Queue
Analyst 01's Dominant Sequence
— 113 of 143 instances
1
View Processing Queue
This analyst exhibits 24 distinct sequences across their 143 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 09 | 95 | 6.3 | 79.2 | Zero-Edit Visits | 0.11 | 0.00 | 4,754 |
|
Reference SoP
1
View Processing Queue
Analyst 09's Dominant Sequence
— 72 of 95 instances
1
View Processing Queue
This analyst exhibits 21 distinct sequences across their 95 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 12 | 68 | 9.7 | 78.8 | Zero-Edit Visits | 0.11 | 0.00 | 30 |
|
Reference SoP
1
View Processing Queue
Analyst 12's Dominant Sequence
— 51 of 68 instances
1
View Processing Queue
This analyst exhibits 12 distinct sequences across their 68 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 02 | 105 | 13.1 | 77.9 | Zero-Edit Visits | 0.15 | 0.00 | 3,175 |
|
Reference SoP
1
View Processing Queue
Analyst 02's Dominant Sequence
— 75 of 105 instances
1
View Processing Queue
This analyst exhibits 24 distinct sequences across their 105 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 05 | 68 | 6.2 | 77.7 | Zero-Edit Visits | 0.15 | 0.00 | 5,602 |
|
Reference SoP
1
View Processing Queue
Analyst 05's Dominant Sequence
— 48 of 68 instances
1
View Processing Queue
This analyst exhibits 17 distinct sequences across their 68 instances; the dominant one is shown.
|
||||||||
79.0
Avg Adherence Score (0-100)
Median 82.5 · Range 51.6–90.0
23.9 min
Avg Task Duration
0.2 min median
1.7
Avg Sub-Steps per Instance
34.9% have multiple sub-steps
75.6%
Reference Path Coverage
Instances following mined SoP exactly
Reference Baseline
75.6% of instances follow this exactly
Review Processing Queue — Mined SoP
The most common sub-step sequence observed across 694 task instances. Use the dropdown on the right to compare any deviation to this baseline.
1
View Processing Queue
0.0s
Phobos
100%
Variant Explorer
Reference
Review Processing Queue Reference Path
525 instances · 7 analysts
This is the mined reference path — 525 instances across 7 analysts follow this exact sequence. Treat as the proposed Standard Operating Procedure for Review Processing Queue.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
17 instances · 4 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Navigate to Cases
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Linked Remediation Opportunities
Reviewer Sweep
Reviewer Verification Pass
12 instances · 5 analysts
· Δ adherence +5.7
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
3
View Processing Queue
extra step (reference ends at step 1)
Reviewer Sweep — a viewing-heavy pattern with 3.1 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 5 analysts across 11 cases.
Cross-App Swivel
Cross-App Lookup to collaboration.merck.com
7 instances · 4 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
collaboration.merck.com
extra step (reference ends at step 1)
Cross-App Swivel — 4 analysts pivot into collaboration.merck.com during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
6 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Start Approval Workflow
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Linked Remediation Opportunities
Genuine Complexity
Complex-Case Execution Path
6 instances · 3 analysts
· Δ adherence +8.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Select Compound
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 87.5 vs reference 79.0 (delta +8.5). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Cross-App Swivel
Cross-App Lookup to Veeva Safety
4 instances · 3 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
View Validation Results
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Veeva Safety during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Linked Remediation Opportunities
Cross-App Swivel
Cross-App Lookup to Microsoft Teams
4 instances · 3 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Teams
extra step (reference ends at step 1)
Cross-App Swivel — 3 analysts pivot into Microsoft Teams during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Acrobat
3 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Acrobat
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Acrobat during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Excel
3 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Excel
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Excel during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Phobos
3 instances · 1 analysts
· Δ adherence -1.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Initiate Case Voiding
extra step (reference ends at step 1)
3
Start Action
extra step (reference ends at step 1)
Cross-App Swivel — 1 analysts pivot into Phobos during this task (0% of task time in non-primary apps, 33% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Cross-App Swivel
Cross-App Lookup to Microsoft Outlook
2 instances · 2 analysts
· Δ adherence -11.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Microsoft Outlook
extra step (reference ends at step 1)
Cross-App Swivel — 2 analysts pivot into Microsoft Outlook during this task (0% of task time in non-primary apps, 50% of events outside the primary system). Pattern is consistent with data lookup or cross-referencing rather than a system defect — lighter-weight than a Silent Workaround.
Genuine Complexity
Complex-Case Execution Path
2 instances · 2 analysts
· Δ adherence +3.5
Observed Sequence (vs Reference SoP on left)
differs from reference
matches reference
1
View Processing Queue
2
Add New Case
extra step (reference ends at step 1)
Genuine Complexity — this sequence differs from the reference path (2 sub-steps vs 1 in reference) but shows no strong pain-point signals: adherence score 82.5 vs reference 79.0 (delta +3.5). Likely a legitimate case-complexity path — not an automation candidate, but worth cataloging.
Application Mix
Which applications analysts touch while executing this task, measured by event count and active dwell time.
| Application | Events | Event Share | Dwell Share |
|---|---|---|---|
| Phobos | 837 | 72.8% | 13.4% |
| Veeva Safety | 221 | 19.2% | 86.6% |
| Microsoft Word | 18 | 1.6% | 0.0% |
| Microsoft Teams | 17 | 1.5% | 0.0% |
| Microsoft Excel | 12 | 1.0% | 0.0% |
| Acrobat | 10 | 0.9% | 0.0% |
| Microsoft Outlook | 8 | 0.7% | 0.0% |
| collaboration.merck.com | 7 | 0.6% | 0.0% |
| Notepad | 4 | 0.3% | 0.0% |
| acrotray | 3 | 0.3% | 0.0% |
Top Cross-App Transitions (within task instance)
Consecutive events that cross application boundaries within a task instance — the signal where swivel-chair patterns live.
| From | To | Transitions |
|---|---|---|
| Phobos | Veeva Safety | 108 |
| Veeva Safety | Phobos | 12 |
| Phobos | collaboration.merck.com | 7 |
| Veeva Safety | Microsoft Teams | 7 |
| Phobos | Microsoft Excel | 6 |
| Phobos | Microsoft Teams | 6 |
| Microsoft Word | Veeva Safety | 6 |
| Veeva Safety | Microsoft Word | 5 |
| Veeva Safety | Acrobat | 5 |
| Phobos | Microsoft Word | 5 |
On-Case Application Journey (±15 min around anchor)
Apps the same analyst touched on the same case within ±15 minutes of the task anchor. Captures work that spans task boundaries — the Veeva/Acrobat/Outlook/Word context surrounding the anchor click.
Case counts per app are not mutually exclusive — a case that touched Veeva AND Phobos in this window is counted in both rows. Overlap reflects real cross-app work, not double-counted cases.
Veeva Safety DOM 6,203 events 259 cases 96.7 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<span>
"Complete"
938×
pyzeClick
<a>
"Show more"
260×
pyzeClick
<li>
"Narrative"
233×
pyzeClick
<li>
"Details"
183×
pyzeClick
<span>
"Narrative"
170×
pyzeClick
<span>
"Approval Complete"
169×
pyzeClick
<span>
"Adverse Events"
159×
pyzeClick
<span>
"Start Approval Workflow"
138×
Phobos DOM 2,616 events 269 cases 72.6 min
DOM-level capture
Pyze JS captures every element interaction — button labels, field IDs, section panels.
pyzeClick
<button>
"Proceed"
464×
pyzeClick
<button>
"Advance"
281×
pyzeClick
<button>
"Save Comments"
250×
pyzeClick
<button>
"Send to process"
135×
pyzeClick
<select>
"Select Compound · - sitagliptin phosphate Tablet (G"
130×
pyzeClick
<select>
"Select Report Type · Legal · Literature · Medical Recor"
119×
pyzeClick
<a>
"Cases"
115×
pyzeClick
<div>
"Please reply to “Request Information" using dropd"
109×
Microsoft Word Document 320 events 111 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
file
V-Narrative 528191 (English) (26300681_0_1_338349)
382×
file
V-Narrative 540591 (English) (26495494_0_2_345212)
294×
file
V-Narrative 531284 (English) (26355806_0_2_340584)
225×
file
V-Narrative 538518 (English) (26459147_0_1_344628)
198×
file
V-Narrative 525091 (English) (26253532_0_1_338428)
197×
file
V-Narrative 530120 (English) (26331907_0_1_341282)
186×
file
V-Narrative 531868 (English) (26362311_0_1_340534)
185×
file
V-Narrative 542172 (English) (26517690_0_2_345412)
181×
Acrobat App 185 events 64 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
2,021 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Teams App 173 events 86 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
1,920 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.
Microsoft Outlook Message 114 events 66 cases
Message-level capture
Task Mining captures email subjects. Case numbers frequently appear in the subject line.
subj
<compensación> Case#2252065
31×
subj
<Due today><Void request><GPVCM POLAND> Case # 2326062
23×
subj
<compensación> Case#2351734
17×
subj
Due today reports on 03-DEC-2025
14×
subj
Due today reports on 01-DEC-2025
13×
subj
<GPVCM Announcement> Promotions to Safety Data Manager positions in Poland
12×
subj
Untitled
8×
subj
RE: <compensación> Case#2309077
8×
Microsoft Excel Document 110 events 50 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
collaboration.merck.com App 92 events 51 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
Notepad App 37 events 22 cases
App-level capture
Only application name and dwell time available. Direct instrumentation would unlock deeper detail.
usc-excel.officeapps.live.com Document 33 events 9 cases
Document-level capture
Task Mining captures the specific file open during work — file names often encode case context.
Adherence Vectors
Each task instance is scored on four behavioral vectors. Lower values are more adherent to the mined SoP.
0.12
Swivel Rate
events in non-primary app / total
0.98
Zero-Edit Visit Rate
inspection-only clicks
1.18
Avg Revisit Depth
max repeats of any activity
0.00
Cross-App Time Share
time outside primary app
Meta Score Distribution
Instance count in each 10-point adherence bucket. Right-skewed = healthy execution; long left tail = concentrated pain points.
0
10
20
30
40
50
60
70
80
90
Top Concern Vector
Zero-Edit Visit Rate — current value 0.98
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Vector Contribution to Adherence Loss
Each vector's weighted contribution to lost adherence. The top row is your highest-leverage fix.
Zero-Edit Visit Rate
value 0.98
Investigate why analysts inspect-and-leave without editing — usually missing upstream data or unclear UI affordances.
Swivel Rate
value 0.12
Reduce cross-app hops by surfacing the secondary-app data inline (API integration or pre-fetched context panel).
Revisit Depth
value 1.18
High revisit depth = mid-flow rework. Trace which sub-step is being repeated; usually a data-readiness gap upstream.
Cross-App Time Share
value 0.00
Time outside primary app dominates this score. Identify which secondary app(s) and whether their function can be in-lined.
Lowest-Adherence Instances
The 5 task instances that scored lowest. These are the concrete cases worth investigating to understand the worst patterns.
| Case | Analyst | Score | Duration | Sub-Steps | Swivel | X-App | Revisit |
|---|---|---|---|---|---|---|---|
| 1535621 | Analyst 13 | 51.6 | 18.5m | 16 | 0.81 | 0.00 | 4 |
| 2225437 | Analyst 13 | 53.3 | 22.3m | 9 | 0.89 | 0.00 | 2 |
| 2219900 | Analyst 02 | 53.8 | 1.6m | 8 | 0.88 | 0.00 | 2 |
| 2335834 | Analyst 09 | 54.3 | 1.9m | 7 | 0.86 | 0.00 | 2 |
| 2304250 | Analyst 13 | 55.0 | 0.6m | 8 | 0.75 | 0.00 | 3 |
Users on This Task
7 analysts executed this task during the pilot. Scatter plots productivity (instances per active day) against adherence (meta score). Top-right = healthy high-volume analysts; bottom-right = firefighters; top-left = careful low-volume; bottom-left = struggling.
Adherence (meta score) →
Productivity (instances / active day) →
03
13
09
01
12
02
05
User-by-User Breakdown
Click any analyst to see their dominant sequence vs the reference SoP.
| Analyst | Instances | Inst / Day | Adherence | Weakest Vector | Swivel | X-App | Min | |
|---|---|---|---|---|---|---|---|---|
| ▸ | Analyst 03 | 117 | 9.8 | 80.0 | Zero-Edit Visits | 0.08 | 0.00 | 848 |
|
Reference SoP
1
View Processing Queue
Analyst 03's Dominant Sequence
— 96 of 117 instances
1
View Processing Queue
This analyst exhibits 20 distinct sequences across their 117 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 13 | 99 | 9.0 | 79.6 | Zero-Edit Visits | 0.10 | 0.00 | 896 |
|
Reference SoP
1
View Processing Queue
Analyst 13's Dominant Sequence
— 74 of 99 instances
1
View Processing Queue
This analyst exhibits 18 distinct sequences across their 99 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 09 | 95 | 6.3 | 79.2 | Zero-Edit Visits | 0.11 | 0.00 | 4,754 |
|
Reference SoP
1
View Processing Queue
Analyst 09's Dominant Sequence
— 72 of 95 instances
1
View Processing Queue
This analyst exhibits 21 distinct sequences across their 95 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 01 | 142 | 9.5 | 79.2 | Zero-Edit Visits | 0.10 | 0.00 | 1,348 |
|
Reference SoP
1
View Processing Queue
Analyst 01's Dominant Sequence
— 111 of 142 instances
1
View Processing Queue
This analyst exhibits 24 distinct sequences across their 142 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 12 | 68 | 9.7 | 78.8 | Zero-Edit Visits | 0.11 | 0.00 | 30 |
|
Reference SoP
1
View Processing Queue
Analyst 12's Dominant Sequence
— 51 of 68 instances
1
View Processing Queue
This analyst exhibits 12 distinct sequences across their 68 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 02 | 105 | 13.1 | 77.6 | Zero-Edit Visits | 0.16 | 0.00 | 3,130 |
|
Reference SoP
1
View Processing Queue
Analyst 02's Dominant Sequence
— 74 of 105 instances
1
View Processing Queue
This analyst exhibits 25 distinct sequences across their 105 instances; the dominant one is shown.
|
||||||||
| ▸ | Analyst 05 | 68 | 6.2 | 77.4 | Zero-Edit Visits | 0.16 | 0.00 | 5,614 |
|
Reference SoP
1
View Processing Queue
Analyst 05's Dominant Sequence
— 47 of 68 instances
1
View Processing Queue
This analyst exhibits 17 distinct sequences across their 68 instances; the dominant one is shown.
|
||||||||