tests/expected/comprehensive-portfolio/report.md

# FY Outside Counsel Panel Spend Review
**Internal Legal Ops Review — Full Detail**

Data reviewed: 22 matters, 5 firms, 7 matter types, 44 timekeeper line items. Prepared from `sample_comprehensive_portfolio.csv`. A companion Excel workbook (`portfolio_review.xlsx`) and findings CSV (`portfolio_review_findings.csv`) accompany this report for sorting, filtering, and sharing — see note at the end.

**Fixture status:** Entirely synthetic test data; no real client, matter, firm, or billing data is included.

---

## 1. Executive Summary

The panel closed the year **$134,360 over budget (+24.9%)** on $674,360 billed against $540,000 budgeted — but that headline number is misleading on its own. Almost the entire overrun is explained by one pattern: **all four Commercial Litigation matters, handled by four different firms, were staffed 84–100% by partner hours with little to no associate leverage**, and every one of them blew through budget by 74% to 154%. Strip Commercial Litigation out of the portfolio and the remaining 18 matters actually ran **5.8% under budget** ($372,030 billed vs. $395,000 budgeted). Two further outliers stand out against otherwise consistent staffing patterns: one Employment matter staffed entirely by partners (Reyes Chen Kaplan, M-2018), and one M&A matter staffed with zero partner hours at all (Castellan Vega, M-2017). Everything else in the dataset — Real Estate, IP/Licensing, Tax Advisory, Regulatory, and the remaining three Employment matters — is staffed consistently and lands at or under budget.

---

## 2. Recommended Actions

Ranked by leverage. These pull together the detailed findings below; several individual findings share the same root cause and are addressed by one action rather than restated separately.

### Priority 1 — Audit and fix Commercial Litigation staffing panel-wide
**What to do:** Audit staffing on all four Commercial Litigation matters (M-2013 Harmon & Slate, M-2014 Whitfield Bosch, M-2015 Dunmore Ellery Pratt, M-2016 Castellan Vega). Require each firm to justify partner-only or >70%-partner staffing on the work billed to date, and write an associate-staffing requirement for discovery and motion-practice phases into every Commercial Litigation engagement letter going forward.
**Why it matters:** This single pattern drove $157,330 of overrun — more than the entire portfolio's net $134,360 overage. Every other matter type in the panel is effectively running under budget once litigation is set aside.
**Timeframe:** Immediately for the four matters in question; written into engagement letters as a standing term going forward.

### Priority 2 — Rebuild the litigation intake/budgeting process
**What to do:** Replace the current flat budget-setting approach for Commercial Litigation ($25k–$45k estimates that missed actuals by 74–154% in all four instances this year) with a complexity questionnaire at matter intake that flags likely trial posture, motion volume, and expected partner-hour share before a budget number is set.
**Why it matters:** A 4-for-4 miss rate across four unrelated firms means the estimating process itself is broken for this matter type — it isn't isolated firm underperformance. Current litigation budgets can't be trusted as a benchmark until this is fixed.
**Timeframe:** Before setting the budget on the next Commercial Litigation matter or renewal.

### Priority 3 — Address M-2018 directly with Reyes Chen Kaplan
**What to do:** Raise M-2018 (Employment Matter) directly with Reyes Chen Kaplan: two partners billed all 35 hours with zero associate or paralegal involvement, unlike every other Employment matter in the portfolio. Request a staffing rationale and require associate-level staffing on any future employment work assigned to this firm.
**Why it matters:** This is the single clearest staffing outlier in the dataset and a 75% budget overrun ($13,500 excess) on work that should be routine and inexpensive.
**Timeframe:** Immediately, before the matter closes.

### Priority 4 — Require partner review on under-staffed M&A work
**What to do:** Require a documented partner review/sign-off step on M&A matters with no partner hours logged, using M-2017 (Castellan Vega, 120 total hours, zero partner hours) as the trigger case. Build this into the standard M&A engagement letter template panel-wide.
**Why it matters:** This is a deal-risk and quality issue, not just a cost one — reps/warranties, closing mechanics, and deal terms typically warrant partner-level review regardless of cost. The matter also modestly overran budget (+6.9%).
**Timeframe:** Immediately, before this matter closes; standard engagement-letter term going forward.

### Priority 5 — Revisit firm-to-matter-type fit
**What to do:** Reassess whether Reyes Chen Kaplan (an IP/Licensing specialist for 3 of its 4 matters here) and Castellan Vega LLP (spread across all five other matter types in the panel) are the right fits for the Employment and M&A work currently routed to them, versus firms with a demonstrated bench in that specific matter type.
**Why it matters:** Both of the worst outliers in the dataset (M-2018 and M-2017) sit at firms handling work outside their apparent core specialty — allocation looks driven by convenience or history rather than fit.
**Timeframe:** Ongoing practice change, next matter assignment cycle.

### Priority 6 — Confirm intent behind the systemic underrun pattern
**What to do:** Revisit estimate-setting for Tax Advisory, Real Estate, IP/Licensing, and standard Employment matters, which underran budget by 10–27% in nearly every one of the 15 matters of those types in this dataset. Either tighten the estimating process or explicitly confirm the cushion is intentional so it isn't later mistaken for a performance signal.
**Why it matters:** A pattern this consistent is itself a scoping-accuracy question, even though it isn't a cost overrun. The data doesn't tell us *why* budgets are generous — that needs a direct check with whoever sets these estimates before acting further.
**Timeframe:** Ongoing practice change, next budget-setting cycle.

---

## 3. Staffing Mix Findings

**Finding 1 — Commercial Litigation cluster: partner-heavy staffing at every firm that touched it (High)**
Matter type: Commercial Litigation. Firms: Harmon & Slate LLP (M-2013), Whitfield Bosch (M-2014), Dunmore Ellery Pratt (M-2015), Castellan Vega LLP (M-2016).

| Matter | Firm | Partner hrs | Associate hrs | Partner share |
|---|---|---|---|---|
| M-2013 | Harmon & Slate LLP | 80 | 15 | 84.2% |
| M-2014 | Whitfield Bosch | 65 | 0 | 100% |
| M-2015 | Dunmore Ellery Pratt | 92 (two partner line items) | 0 | 100% |
| M-2016 | Castellan Vega LLP | 58 | 6 | 90.6% |

*Driver:* Every other matter type in this dataset is staffed predominantly by associates and paralegals, with partners contributing 0–55% of hours in an oversight capacity. Commercial Litigation is the exception across all four firms that handled it — two matters (M-2014, M-2015) used **zero associate hours at all**. Because this shows up consistently across four otherwise differently-staffed firms, it looks like either a genuinely trial-heavy posture across all four matters, or a panel-wide habit of not building associate leverage into litigation staffing plans. It does not look like one firm behaving badly.
*Recommendation:* Require associate-level staffing for discovery/motion-practice phases in future litigation engagement letters; ask each of the four firms for a staffing-plan rationale on the matters above.

**Finding 2 — M-2018 Employment Matter (Reyes Chen Kaplan): 100% partner staffing on routine work (High)**
Two partner line items (25 hrs and 10 hrs, both at $900/hr) total 35 hours with zero associate or paralegal touch. Every other Employment matter in the portfolio — M-2001 (Harmon & Slate), M-2010 (Dunmore Ellery Pratt), M-2020 (Whitfield Bosch) — was staffed entirely by associates and paralegals.
*Driver:* This is the single cleanest staffing outlier in the dataset. It's plausible the firm simply doesn't have an associate assigned to employment matters, or the matter was treated as a client-relationship engagement rather than routine work.
*Recommendation:* Raise directly with Reyes Chen Kaplan; require associate-level staffing on future employment assignments to this firm.

**Finding 3 — M-2017 M&A Matter (Castellan Vega): zero partner hours on a complex deal (Medium)**
90 associate hours + 30 paralegal hours = 120 total hours, with no partner hours at all. Compare to the panel's other M&A matter, M-2006 (Whitfield Bosch), which used 20% partner hours across 100 total hours and came in under budget.
*Driver:* This is an under-staffing risk rather than overstaffing — M&A work typically needs partner-level review of deal terms and closing mechanics regardless of cost. Given Castellan Vega handles five different matter types in this panel, this may reflect a thin M&A bench rather than a deliberate choice.
*Recommendation:* Require a documented partner sign-off step on M&A matters with no partner hours logged before the matter closes.

**Finding 4 — Regulatory and Tax Advisory: consistent, no outliers (Low)**
Tax Advisory (M-2002, M-2011, M-2022) shows partner shares of 30%, 33%, and 38% respectively across three different firms. Regulatory (M-2007, M-2012) shows 56% and 55% partner share across two firms. Both are internally consistent and look purposeful.
*Recommendation:* No staffing action needed; use these ratios as the internal benchmark when reviewing future engagement letters for these matter types.

**Finding 5 — Real Estate and IP/Licensing: junior-driven, no outliers (Low)**
Real Estate (M-2008, M-2009, M-2021) and IP/Licensing (M-2003, M-2004, M-2005, M-2019) are staffed almost entirely by associates and paralegals, with at most one light partner touch (M-2004, 12.5% partner share) across seven matters and three firms.
*Recommendation:* No action needed; this is the staffing model to hold up as the standard for these matter types.

---

## 4. Budget Variance Findings

**Finding 1 — Commercial Litigation cluster: $302,330 billed vs. $145,000 budgeted, +108.5% (High)**

| Matter | Firm | Billed | Budget | Variance |
|---|---|---|---|---|
| M-2014 | Whitfield Bosch | $63,375 | $25,000 | +153.5% |
| M-2015 | Dunmore Ellery Pratt | $94,300 | $40,000 | +135.8% |
| M-2013 | Harmon & Slate LLP | $83,875 | $45,000 | +86.4% |
| M-2016 | Castellan Vega LLP | $60,780 | $35,000 | +73.7% |

*Driver:* Directly tied to the partner-heavy staffing pattern in Staffing Mix Finding 1. A 4-for-4 miss across four different firms is a strong signal the litigation estimating process itself is broken, not that any one firm is padding hours.
*Recommendation:* Rebuild the litigation intake/budgeting process (see Recommended Action #2) before setting the next budget for this matter type.

**Finding 2 — M-2018 Employment Matter (Reyes Chen Kaplan): $31,500 vs. $18,000, +75% (High)**
*Driver:* Direct consequence of the partner-only staffing flagged above — partner rates ($900/hr) applied to 35 hours of what should be associate-level employment work.
*Recommendation:* Address as part of the direct firm conversation in Recommended Action #3; consider whether the firm should retain this work type at all without an associate on the file.

**Finding 3 — M-2007 Regulatory Matter (Whitfield Bosch): $14,070 vs. $12,000, +17.3% (Medium)**
*Driver:* A modest overrun consistent with the matter type's normal partner-heavy staffing model (56% partner share is typical for Regulatory in this dataset) — not clearly a staffing problem, but the magnitude is worth a scope check.
*Recommendation:* Confirm with Whitfield Bosch whether scope expanded mid-matter before renewing similar Regulatory estimates.

**Finding 4 — M-2017 M&A Matter (Castellan Vega): $58,800 vs. $55,000, +6.9% (Low)**
*Driver:* A small overrun; the more significant issue on this matter is the missing partner hours (Staffing Mix Finding 3), not the cost variance itself.
*Recommendation:* Handled via the partner-review-step action; no separate budget action needed.

**Finding 5 — Systemic underrun: 15 of 22 matters came in 10–27% under budget (Low)**
Every Tax Advisory, Real Estate, IP/Licensing, and standard Employment matter (all except M-2018) landed under budget, most by double digits (e.g., M-2002 at -27.5%, M-2003 at -26.9%, M-2012 at -20.5%).
*Driver:* Consistent underrun across matter types and firms suggests budgets/estimates are set with a built-in cushion rather than tightly calibrated to actual work. This isn't a cost problem, but it is a scoping-accuracy question — the data can't tell us whether this cushion is intentional (e.g., contingency built in on purpose) or just a stale estimating habit.
*Recommendation:* Confirm with whoever sets these estimates whether the cushion is deliberate; if not, tighten estimating guidance for these matter types.

**Portfolio net variance:** $674,360 billed vs. $540,000 budgeted = **+$134,360 (+24.9%)**. Excluding the four Commercial Litigation matters: $372,030 billed vs. $395,000 budgeted = **-$22,970 (-5.8%)**. The litigation cluster alone accounts for $157,330 of overrun — more than the total net overage — meaning the rest of the portfolio is actually running under budget in aggregate.

---

## 5. Allocation Findings

**Finding 1 — Commercial Litigation staffing pattern is panel-wide, not firm-specific (High)**
The same overstaffing/overrun pattern shows up at all four firms that handled a Commercial Litigation matter this year — Harmon & Slate, Whitfield Bosch, Dunmore Ellery Pratt, and Castellan Vega — even though those same four firms staff their other matter types very differently from one another (e.g., Dunmore Ellery Pratt's Real Estate work is entirely associate/paralegal-driven, but its one litigation matter is 100% partner hours).
*Driver:* This consistency across otherwise-different firms points to a panel-wide intake or staffing convention specific to litigation matters, not any one firm's practice.
*Recommendation:* Treat this as a process fix (litigation intake and staffing-plan requirements applied panel-wide) rather than a single-firm renegotiation.

**Finding 2 — Reyes Chen Kaplan: IP specialist handling one Employment matter, with the worst result in the dataset (Medium)**
Reyes Chen Kaplan handled 3 IP/Licensing matters (M-2003, M-2004, M-2005) efficiently — standard associate/paralegal staffing, all under budget by 15–27%. Its one Employment matter (M-2018) is the single worst staffing and budget outlier in the entire dataset.
*Driver:* Consistent with being assigned work outside the firm's demonstrated specialty without an appropriate associate bench for that practice area.
*Recommendation:* Route future Employment matters to a firm with a demonstrated employment-practice bench rather than this firm's IP team.

**Finding 3 — Castellan Vega LLP: broadest matter-type spread in the panel, sitting at both ends of the allocation-imbalance spectrum (Medium)**
Castellan Vega is the only firm in the panel handling five different matter types (Tax Advisory, Regulatory, Commercial Litigation, M&A, Real Estate). It is also the firm behind the zero-partner-hours M&A matter (M-2017) and one of the four overstaffed litigation matters (M-2016).
*Driver:* Breadth of assignment may be outrunning depth of bench in at least two of the five matter types this firm is asked to cover.
*Recommendation:* Review whether Castellan Vega's M&A and litigation staffing plans reflect genuine specialty depth before routing more of that work to them.

**Finding 4 — Whitfield Bosch's M&A matter (M-2006) is a useful positive benchmark (Low)**
Partner/associate/paralegal split of 20%/55%/25% across 100 hours, and the matter came in under budget (-5.9%).
*Driver:* Not a problem — this is what appropriately leveraged M&A staffing looks like in this dataset.
*Recommendation:* Reference this staffing ratio when negotiating M&A staffing plans with other panel firms, including Castellan Vega on M-2017.

**Finding 5 — Real Estate work concentration (Dunmore Ellery Pratt x2, Castellan Vega x1) (Low)**
All three Real Estate matters are staffed consistently (associate + paralegal only, no partner hours) and all landed at or under budget.
*Driver:* Appropriate specialization; no allocation concern.
*Recommendation:* No action needed.

---

## 6. AI Leverage Opportunities

AI leverage was not assessed from this fixture because it contains matter names and role-level totals but no task codes, activity codes, or line-item descriptions. Obtain task-level detail before identifying work as suitable for AI-assisted review, research, or drafting.

## Note on the accompanying spreadsheet

A companion Excel workbook (`portfolio_review.xlsx`) and a flat findings CSV (`portfolio_review_findings.csv`) are included alongside this report in the same folder. The workbook has five tabs — Summary, Action Plan, Findings (color-coded by severity), AI Leverage (which records that task-level evidence was unavailable), and Normalized Data (matter-level rollup of all 22 matters with hours by role, billed amount, budget, and variance) — for further sorting, filtering, or sharing with others who need the underlying numbers rather than the narrative.