Shopware Migration

Appearance

Shopware Migration Program Analysis — Week of 2026-08-12 ​

Executive Summary ​

  • SXSW velocity crisis deepens: Advances just +1 story to 68% (605/890) despite 42 commits from 9 contributors—0.14 stories/day is program's lowest sustained rate, now 6+ weeks behind mid-July projections with 285 stories outstanding and no acceleration visible.
  • CE Shopware board chaos: Drops 37 completions to 91% (587/642) after -170 scope reclassification—105 commits from Blake/Emil/Scott prove work continues but massive denominator swing (likely tickets moved to W1DA/infrastructure per prior week's orphan-fix pattern) renders progress illegible.
  • ARROW Frontend/Graph both flatline: Frontend stalls at 94% (463/495, zero movement) despite Pablo's 194 commits; Graph holds 83% (52/63, zero movement) despite 15 commits from 6 people—highest commit volumes in program yield zero board progress, indicating systemic board-vs-reality gap.
  • Baker/ECMD reallocation pays off (finally): Baker hits 35% (+4, 109/312) and ECMD 14% (+5, 54/374) after Sebastian/Gabriel's second week—predicted 1–2 week lag now clearing, but Baker's 89 commits vs ECMD's 24 suggests decomposition still mismatched.
  • Review pipeline stabilizes but oldest tickets calcify: 145 tickets in review (up from 139) but stale >30 days holds at 10—SXSW-1255 now at 116 days and SXSW-1261 at 70 days with no resolution path, suggesting fundamental ownership/priority gap on aging items.

Progress Overview ​

SXSW  [████████████████░░░░░░░░] 68% (605/890)   ↑1  ↑14
ARROW [███████████████████████░] 94% (463/495)   →0  →0
Graph [███████████████████░░░░░] 83% (52/63)     →0  →0
Infra [████████████████████░░░░] 85% (90/106)    →0  →0
ARWUI [█████████████░░░░░░░░░░░] 54% (36/67)     ↑1  →0
CESW  [██████████████████████░░] 91% (587/642)   ↓37 ↓170
W1DA  [██████████████░░░░░░░░░░] 58% (135/231)   →0  →0
ECMDSW[███░░░░░░░░░░░░░░░░░░░░░] 14% (54/374)    ↑5  ↑10
GSHOP [░░░░░░░░░░░░░░░░░░░░░░░░] 0%  (0/74)      →0  ↑2
BSWM  [████████░░░░░░░░░░░░░░░░] 35% (109/312)   ↑4  ↑5

1. Shopware Platform - SXSW Initiatives ​

Current: 605/890 done (68%) — up +1 story with +14 scope growth.

Week-over-week: +25 done, +27 scope (from 2026-08-05: 580/863). Two-week trend shows recovery from prior regression but velocity remains critically low at 3.6 stories/week (0.5/day).

Repository activity: 42 commits from 9 contributors (Paul: 15, Adrian: 15, Rob: 7, others: 5 or fewer). Work concentrates in ecommerce-supergraph (30 commits), sw-plugin-offers (12), scattered across Login-as-Customer, Offer filtering, Fee/Supersede resolvers—all Domain API layer, no Shopware plugin commits this week.

Epics status: 10 of 11 initiatives remain incomplete. Infrastructure (46%, 104/227) and Cart/Checkout (70%, 89/128) lag furthest behind. All others 77%+ but none at 100%—classic long-tail stall where final 20–30% refuses to close.

Risk: Commit activity proves work continues but board movement decoupled—0.14 stories/day (1/week) is unsustainable for 285 remaining stories. Timeline now projects late 2026 Q4 at current pace, 6+ weeks behind original mid-July targets. No visible intervention (resourcing, rescoping, or blocker removal) addresses this.

3. Arrow - Domain API / Graph ​

Current: 52/63 done (83%) — zero movement this week, zero last week.

Repository activity: 15 commits from 6 contributors (Ilir: 11, Cristian: 10, others scattered). Work lands in ecommerce-supergraph but no board tickets transition to Done—same pattern as ARROW Frontend.

Epics status: 5 of 24 epics incomplete (Inventory 1/3, Quotes 2/5, Permissions 0/5 not started). Remaining 11 stories concentrate in feature gaps (Permissions, Inventory detail) and cross-cutting concerns (Persisted Queries 7/12).

Trend: Board flatlined for 3 weeks despite sustained commit activity. Work either landing in wrong repo, tickets not transitioned, or decomposition mismatch where commits close sub-tasks not reflected in epic rollups. Identical symptom to ARROW Frontend's 194-commit / zero-progress week—suggests systemic issue, not isolated to one board.

4. Arrow - Frontend / Open Components ​

Current: 463/495 done (94%) — zero movement despite highest commit week in program.

Repository activity: 194 commits from Pablo alone (Next.js 16.3 upgrade, ESLint 9 migration, Quick Order features, Store Locator enhancements), plus 36 from German, 25 from Gabriel, 16 from Robert. Work spans watsco-frontend-boilerplate, peirce-frontend, baker-headless, homans-next-frontend—all correct repos per board scope.

Epics status: 28 of 43 epics show "Done ✅" or "nearly complete" (≥97%). Remaining 32 stories scatter across 15 epics: Unclassified (78/89), Store Locator (14/19), Quick Order (1/5), Manage Cart (3/4), Open Components (71/73). All active-development epics per Jira.

Analysis: This is the clearest board-vs-reality failure in the program. 194 commits from one person with zero board movement is impossible unless: (a) commits close sub-tasks not tracked in epic rollups, (b) tickets stuck in review/QA and not transitioned, or (c) board denominator still polluted with stale/won't-ship scope despite July's ARROWUI reclassification.

Recommendation: Audit epic-to-ticket parent links for Pablo's commits this week. If commits reference tickets outside tracked epics, either fix rollup or reclassify tickets. If tickets in review, escalate review bottleneck. If denominator is junk, aggressively close won't-ship items.

5. Arrow - Infrastructure / Cross-cutting ​

Current: 90/106 done (85%) — zero movement, zero last week.

Repository activity: 11 commits (Nathan: 14 across multiple boards, subset here). Work in ecommerce-supergraph and sw-plugin-conduit.

Epics status: 3 of 13 epics incomplete: Demo Environment Bugs (23/28), Performance Tuning (12/18), Homans to ARROW+Shopware (10/13). All unblocked per Jira, no stale tickets, but none closing.

Trend: Classic "last 15%" stall—no blockers, commits landing, but tickets not transitioning. Likely overly-granular decomposition (28-ticket epic for "bugs") or test/verification gates not being hit. Less critical than SXSW or Frontend stalls (only 16 stories outstanding) but pattern repeats across all ARROW boards.

11. Arrow UI (ARROWUI) ​

Current: 36/67 done (54%) — up +1 story, zero scope change.

Week-over-week: +5 done (from 31/67 on 2026-08-05). Sustained progress after reclassification—this board now cleanest signal of "actual shipping frontend work" in program.

Epics status: 5 of 13 epics in active development (Checkout 5/6, Frontend Bug Fixes 11/12, Nearby/Branch Filters 2/3, Production Bug 1/2, Playwright E2E 0/7 not started). Rest either Done or planning phase.

Repository activity: Inherits Pablo/German/Gabriel's frontend commits from above—work shows up here instead of ARROW Frontend now that reclassification complete.

Analysis: This board proves the ARROW consolidation strategy works when executed—progress visible, velocity measurable, scope reflects reality. Contrast with ARROW Frontend's 194-commit flatline shows the cost of not finishing cleanup: same work, one board moves, one doesn't.

6. CE Shopware (CESW) ​

Current: 587/642 done (91%) — down -37 completions with -170 scope reduction.

Week-over-week: Massive swing after prior week's surge to 77% (624/812). Board drops 37 stories and sheds 170 tickets—largest single-week denominator change since tracking began.

Repository activity: 105 commits from Blake (62), Emil (105), Scott (34). Sustained high velocity in cesw repo (all Shopware plugins). Work includes Docebo PATH emails, OEEPI reporting, DMI/FBT cleanup, address/promotions migrations, order approval imports—all real, measurable work per commit messages.

Analysis (referencing program notes): This mirrors the 2026-08-04 "orphan ticket discovery" pattern on W1DA/BSWM/GSHOP when tracking was fixed. CE likely reclassified 170 tickets out of CESW board (into W1DA for Domain API work, or Infrastructure for cross-cutting concerns) to clean up denominator pollution. The -37 "completions" are an artifact of tickets being removed from the board's rollup, not work being undone.

Do NOT flag this as unexplained regression—program notes from 2026-08-04 establish this is a deliberate, one-time tracking fix. Next week should show whether the remaining 642 denominator is clean and progress resumes, or if further cleanup needed.

7. Domain API (W1DA) ​

Current: 135/231 done (58%) — zero movement, zero commits this week.

Week-over-week: +2 done (from 133/232 on 2026-08-05), but flatlined vs last week's snapshot.

Epics status: 23 of 36 epics "Not started" (light gray). Active work concentrates in Cart (8/10), Persisted Queries (7/12), Offers (2/11). Remaining 96 stories scatter across future work (Checkout Payment 0/3, Product Detail 0/3, Pricing 0/1).

Analysis: Board became visible after 2026-08-04 orphan-ticket fix (+36 scope that week) but work hasn't staffed yet. Zero commits this week confirms no active development—backlog exists, prioritization unclear.

Risk: 96 outstanding stories with zero velocity = undefined timeline. If W1DA is critical path for CE or other BUs, this is a resourcing gap. If it's parking-lot work, should be deprioritized or closed to avoid inflating program metrics.

8. ECMD Shopware (ECMDSW) ​

Current: 54/374 done (14%) — up +5 completions with +10 scope growth.

Week-over-week: +12 done, +20 scope (from 42/354 on 2026-08-05). Second week of progress after Sebastian/Gabriel reallocation.

Repository activity: 24 commits this week (Sebastian: 24 in ecommerce-supergraph and domain-api-ecmd, Gabriel: 25 across multiple frontends). Damir contributes 5 commits in domain-api-ecmd.

Epics status: Infrastructure surges to 61% (27/44), Auth to 55% (6/11), Catalog to 34% (16/47). Remaining 9 epics untouched (Integrations 1/21, Account Mgmt 3/103 total).

Analysis (referencing program notes): Week 2 of reallocation now clearing predicted 1–2 week startup lag. Board movement matches commit activity—foundational layers (Infra, Auth) completing first per vertical-slice pattern. 320 stories outstanding (86%) signals multi-month runway but velocity now measurable at 8.5 stories/week (1.2/day).

Contrast with Baker: ECMD's 24 commits → +5 board movement vs Baker's 89 commits → +4 movement suggests ECMD's decomposition cleaner (commits map to stories 1:1 roughly) while Baker's still mismatched (commits close sub-tasks not visible in epic rollups).

9. Gemaire Shopware (GSHOP) ​

Current: 0/74 done (0%) — zero movement with +2 scope growth.

Repository activity: 14 commits from Yovany (7), Damian (7), Nikolay (6). Work in gem-headless frontend.

Blocker analysis: 8 active blockers cascade from 4 root causes, impacting 21 tickets total:

  • GSHOP-148 (Search fallback) blocks 5 test tickets across all phases
  • GSHOP-147 (Nav dropdown) blocks 4 test tickets
  • GSHOP-156 (Branch finder) blocks 2 test tickets
  • GSHOP-118 (Google Maps) blocks 2 test tickets

All 4 root causes assigned to Damian/Mike D'Ottavio, all in "Not started" status per Jira. Tickets affected are testing/QA phase—suggests foundational work incomplete, can't validate until blockers clear.

Analysis: Board shows zero completions because nothing can close until foundational blockers resolve. 14 commits land but affect infrastructure (AHRI, warranty, address book per commit messages)—not the Search/Nav/Maps blockers holding tests. Aggressive "Definition of Done" (requires E2E validation) prevents premature closure, which is healthy, but means velocity invisible until blockers clear.

Timeline risk: If blockers don't resolve this week, 21 blocked tickets can't advance, and 0% persists. Escalate Damian's workload—4 root causes on one person is single-point-of-failure.

10. Baker Shopware (BSWM) ​

Current: 109/312 done (35%) — up +4 completions with +5 scope growth.

Week-over-week: +44 done, +45 scope (from 65/267 on 2026-08-05). Sustained momentum after reallocation.

Repository activity: 89 commits (Tyler: 89, Jerrold: 9, Nick: 11, Pablo: 41, others scattered). Work spans baker-headless (Tyler's frontend work), baker-shopware (Nick's order sync), watsco-frontend-boilerplate (shared components).

Epics status: 6 epics advance: Header 79% (23/29), Shopware Backend 89% (24/27), Baker Headless 84% (21/25), Branch Mgmt 92% (11/12), Pricing 78% (7/9), Job Accounts 17% (1/6). Remaining 26 epics "Not started" including all feature work (Order Approvals 0/9, StockPro 0/20, Bids 0/8, Warranty 0/9, AHRI 0/9, Parts Lookup 0/6).

Analysis: Progress concentrates in infrastructure/foundation layer (Header, Backend, Headless shell) per vertical-slice pattern—same as ECMD. 203 stories outstanding (65%) in feature epics that haven't started yet.

Decomposition mismatch vs ECMD: Baker's 89 commits this week yield +4 board movement (22:1 commit-to-story ratio) while ECMD's 24 commits yield +5 movement (5:1 ratio). This suggests Baker's commits close sub-tasks not visible in epic rollups, or Tyler's frontend work doesn't map cleanly to Jira stories (common when refactoring/infrastructure work spans multiple epics).

Not a failure—infrastructure work naturally has worse commit-to-story ratio than feature work—but worth monitoring. If ratio doesn't improve as feature epics start (Order Approvals, StockPro next per roadmap), indicates board decomposition needs adjustment.

Review Pipeline Aging ​

Current state: 145 tickets in review/QA (up from 139 last week), 10 stale >30 days (same as last week).

Stale ticket analysis:

  • SXSW-1255 (116 days, "In Review"): CE Product List price events—Gabriel assigned, no movement in 4 weeks
  • SXSW-1261 (70 days, "In Review"): RudderStack page view props—Gabriel assigned, paired with 1255
  • SXSW-1125 (56 days, "Active Development"): EKS/K8s hosting—Chris Nitchie, infrastructure work known to be blocked on Webscale per prior program notes
  • GSHOP-93 (47 days, QA): Customer export install—girish assudani, Gemaire migration work
  • SXSW-567 (46 days, "Development Complete"): Express Pickup resolver—Mitchell Cowie, waiting unknown gate
  • CESW-483 (42 days, "Waiting On External"): Auth0 config—Blake, external dependency per status
  • SXSW-390 (41 days, "Development Complete"): Peirce schema approval—Lucian, waiting sign-off
  • CESW-915/914 (34 days, "Deploying"): Conduit cache-bypass features—Blake, deployment pipeline delay

Pattern: Two clusters emerge:

  1. Gabriel's analytics pair (SXSW-1255/1261): 116 and 70 days with no resolution path visible. Gabriel contributed 25 commits this week to checkout/ECMD work—suggests deprioritization or review bottleneck, not capacity issue.
  2. Blake's deployment queue (CESW-915/914/483): All "Waiting On External" or "Deploying" status—external dependencies or slow deployment pipeline, not code-review bottleneck.

Recommendation: Escalate SXSW-1255/1261 to Gabriel + reviewer (likely Paul or Adrian per team map). If analytics work deprioritized, close tickets as won't-fix to clean pipeline. If waiting review, assign explicit reviewer with deadline. 116 days in "In Review" without resolution is ownership gap, not technical blocker.

Repository Activity & Team Velocity ​

Top contributors this week (by commit count):

  1. Pablo Romero (194 commits): Next.js 16.3 upgrade, ESLint 9 migration, Quick Order CSV upload, Store Locator enhancements, breadcrumb fixes—all watsco-frontend-boilerplate and BU frontends. Zero board movement despite highest output.
  2. Emil Stewart (105 commits): FBT cleanup, address fixes, delivery express migration, promotions import—all cesw Shopware plugins. Board shows -37 but work is real per commit log.
  3. Tyler Hawkins (89 commits): Baker frontend infrastructure, drawer tokenization, tabs, input components—baker-headless primarily. +4 board movement from 203 outstanding.
  4. Blake Saunders (62 commits): Docebo PATH, OEEPI, DMI, address validation—cesw Shopware. Board regression artifacts, not velocity issue.
  5. Adrian Pescar (42 commits): Offer filtering, Login-as-Customer, branch context—ecommerce-supergraph and account structure. SXSW +1 movement, work visible.

Velocity analysis by layer:

  • Frontend (Pablo, Tyler, Gabriel, German): 343 commits, minimal board movement (ARROW +0, Baker +4, ARROWUI +1). Highest output, lowest visible return.
  • Graph (Adrian, Rob, Ilir, Cristian): 96 commits, SXSW +1, ARROW Graph +0. Domain API work landing but not closing stories.
  • Shopware (Blake, Emil, Scott, Nick): 212 commits, CESW -37 (artifact), ECMD +5, Baker +4. Real work but board metrics polluted by reclassification.

Team-level observations:

  • Human Element (Paul, Rob, Adrian, Ilir, Joshua): 108 commits across SXSW/Graph/Infra. Steady output but SXSW velocity crisis (0.14 stories/day) persists despite their work.
  • Baker/ECMD (Sebastian, Gabriel, Cleo, Tyler): 149 commits. Reallocation paying off—ECMD +5, Baker +4, both inside predicted effect window.
  • CE (Blake, Emil, Scott): 201 commits. Sustains program's highest per-person velocity but board regression this week masks it.
  • Gemaire (Yovany, Damian, Nikolay): 20 commits. Work lands but blockers prevent closure—board stuck at 0% until foundational issues resolve.

Cross-layer observation: Commit activity proves work continues across all boards, but board movement decouples systematically:

  • ARROW Frontend: 194 commits → 0 stories (infinite ratio)
  • ARROW Graph: 15 commits → 0 stories
  • SXSW: 42 commits → 1 story (42:1)
  • Baker: 89 commits → 4 stories (22:1)
  • ECMD: 24 commits → 5 stories (5:1, healthiest ratio)

This isn't a velocity problem—it's a board tracking problem. Either epic rollups are broken (commits close sub-tasks not visible in parent epics), or "Definition of Done" is systematically misaligned across boards (work lands but tickets stay open waiting for review/test/deploy gates).

mermaid
gantt
    title Projected Completion (based on current velocity)
    dateFormat YYYY-MM-DD
    todayMarker stroke-width:2px,stroke:#ef4444
    
    section Core Platform
    SXSW (likely range)        :active, 2026-08-12, 2027-01-15
    SXSW (optimistic)          :crit, 2026-08-12, 2026-11-01
    
    section Business Units
    CE Shopware                :done, 2026-08-12, 2026-09-01
    ECMD Shopware              :2026-08-12, 2027-02-15
    Baker Shopware             :2026-08-12, 2027-01-01
    Gemaire (blocked)          :crit, 2026-08-12, 2026-09-15
    
    section Foundation
    ARROW Frontend             :done, 2026-08-12, 2026-09-01
    ARROW Graph                :2026-08-12, 2026-10-01
    ARROW Infra                :2026-08-12, 2026-10-01
    ARROWUI                    :2026-08-12, 2026-10-15

Timeline notes:

  • SXSW: 285 stories at 0.14/day = 2,036 days (5.6 years) at current pace—chart shows realistic 5-month window assuming velocity recovers to 2 stories/day baseline
  • CE Shopware: Marked "done" as 91% with clean-up phase remaining—board regression this week is tracking artifact, not real backslide
  • ECMD/Baker: 6–7 month projections based on current 1.2/day and 0.8/day velocities respectively, assuming no further reallocation
  • Gemaire: Blocked until foundational issues resolve—2-week estimate assumes blockers clear this week, otherwise indeterminate
  • ARROW boards: Frontend/Graph marked near-complete (94%/83%) but flatlined—1-month buffer assumes board cleanup, not new work