After $30M, Your Bottleneck Is Decision Rights

8:02 a.m., Monday. Revenue is at $34M run rate, PE board meeting in eight days. The CEO wants answers on margin drift. Sales has a seven-figure deal stuck at 18% gross margin if they don’t get a discount approved by Friday. Ops needs to commit $420k to a packaging line upgrade before the vendor pulls terms. Marketing can’t finalize the Q3 campaign without pricing direction. Four different Slack threads, three Google Docs, two “quick” huddles, and everyone is waiting for “one more approval.”

By 5 p.m., nothing moves. The CFO’s calendar gets hammered with “just a sanity check.” The CEO asks for a cross-functional alignment meeting. A new pre-read deck is born. Decision age: one more day older. The company is still acting like it’s $12M, but the surface area of decisions has tripled.

The Misconception

Most teams blame decision slowdown on scale, complexity, or “we don’t have enough data.” They add dashboards, OKRs, and meetings. They hire a COO. They declare “empowerment” and “alignment” goals. They ask the board for patience.

Then cycle times get worse.

If you ask who is actually allowed to say “yes” to the critical, repeatable decisions—spend, pricing, people moves, risk—the answers are vague. Or there are too many of them. Or the board sits in the middle of the line. The organization is solving a rights problem with tools and ceremonies.

The Reality

Decision velocity collapses after $30M because the decision rights model that worked at $10–$20M fails under a larger decision surface. Tacit founder vetoes, inherited caps from the scrappy years, and consensus culture collide with higher stakes and more interdependencies.

• Thresholds are too low. Capex under $100k still needs CEO sign-off. Discounts over 10% still roll to Finance. Vendor changes above $50k route through four functions “just to be safe.”

• Approvals stack vertically. Legal, Finance, Ops, and Sales all need to say yes. Each additional “yes” multiplies delay risk.

• P&L ownership is muddled. Everyone feels accountable; no one is the decider. Cross-functional work defaults to unanimity.

• The board leaks into operations. “Reserved matters” and shadow approvals creep into everyday decisions, creating defensive escalation.

Here’s the math you already feel: if a decision requires four approvals and each approver has a 75% probability of responding on time with a yes, the probability of a on-time yes is 0.75^4 ≈ 32%. Add one more approver and it drops to 24%. That’s your pipeline of “almost there” decisions.

Data accuracy won’t overcome that geometry. Better meetings won’t either. Only a rights redesign does.

The Pattern

I’ve watched this movie across PE-backed software, industrial services, and specialty manufacturing. The plot points rhyme.

• $5–$20M: Heroics and proximity. Founder knows every major customer. “Come to my office” is the decision system. Tacit rules, tight loops, high speed.

• $20–$35M: More SKUs/segments/geographies. More managers. Initial PE oversight. The founder/CEO tries to keep personal veto power on spend, pricing exceptions, and key hires. Finance starts policing risk with well-meaning gatekeeping.

• $35–$60M: Matrix emerges. Sales ops, RevOps, S&OP, PMO. Decision surface expands. Pre-reads proliferate. Email CC lists lengthen. “We’ll decide next week” becomes a reflex. Morale drifts as A-players stop bringing bold proposals because aged decisions get them punished by time.

Symptoms you can measure:

• Decision age. Exceptions and approvals sit 7–21 days before a yes/no.

• Approval count. 3–6 functional approvals for non-strategic moves.

• Threshold misfit. CEO or board in the loop for sub-$250k, sub-2% margin-impact items.

• Shadow re-litigation. Decisions revisited within 60 days because the true decider wasn’t defined.

Real-world stakes:

• A $38M industrial services company lost $900k EBITDA in 12 months because a packaging line upgrade sat in approval limbo for 11 weeks, forcing overtime and rush freight. After moving capex rights to the Ops VP for ≤$250k and reserving only >$1M to the board, they cut approval time from 41 days median to 6, recovered 220 bps of gross margin, and freed 18% of overtime spend.

• A $45M SaaS provider’s average discount exception cycle was 12 days. Win rate on deals >$250k dropped 8 points, and they burned two quarters of forecast accuracy. By setting role-based discount guardrails tied to gross margin and prepay, they moved 64% of exceptions to same-day approval and clawed back $1.4M ARR within two quarters.

The message: if your decision rights are misaligned, your P&L pays for it in overtime, discount leakage, rush premiums, slippage, and attrition of your best people.

The Framework

Primary pillar: Decision Rights. You don’t need a task force. You need four moves executed in four weeks. This is architecture, not theater.

1. Build a Decision Catalog with SLAs

• In a 90-minute working session per function, list the 20–30 recurring decisions that move money, market, people, and risk. Examples: discount exceptions; vendor selection by spend tier; capex by band; headcount adds and backfills; price changes; payment terms; customer credits; roadmap prioritization gates.

• For each, define: the DRI (single name), approver(s) if any, required inputs (not a 20-page deck—two or three fields of evidence), decision SLA (e.g., 48 hours for discounts, 5 business days for $250k–$1M capex), and an appeal path.

• Publish as a one-page per function plus a cross-functional index. Put it where work happens: the CRM, ERP, and Slack/Teams wiki. This is the new “who says yes” map.

Why it works: ambiguity is the hidden tax. A catalog collapses ambiguity into explicit rights and time boxes.

2. Raise Thresholds, Compress Approvals

• Raise capex and spend thresholds to match materiality. As CFO, tie them to unit economics and cash conversion, not fear. A common pattern:

  • Managers: approve up to $50k non-recurring, within budget, if ROI > 18 months.

  • Directors: up to $150k; VPs: up to $250k; COO/CFO: up to $500k; CEO only for $500k–$1M, board reserved >$1M (or >2% of TTM EBITDA impact).

  • Remove the CEO from all spend < $250k; remove the board from all but reserved matters.

• For pricing/discounts, set role-based guardrails anchored in gross margin and CAC payback:

  • AEs can approve up to X% discount if deal GM ≥ target and prepay ≥ Y months.

  • Sales leaders approve up to target+Z% if GM within 100 bps of target and term ≥ N months.

  • Finance notified, not an approver, unless below GM floor or non-standard terms.

• For cross-functional moves (pricing changes, S&OP tradeoffs), name a single DRI with tie-break authority. Kill the unanimity trap. Set a clear appeal path to one executive, not a committee.

Why it works: every extra approval is a coin flip on time. One decider with clear thresholds beats five “inputs” that function as vetoes.

3. Move Boards Out of the Line

• Re-articulate reserved matters with the chair/operating partner: define what is actually strategic (e.g., M&A, capital structure, executive hires/fires, >$1M capex, new site commitments, long-term contracts >$2M TCV, compensation plans). Everything else lives inside management’s rights.

• Create a board “pre-brief” rhythm for items near reserved lines, not an approval gate. The board gets visibility, management keeps velocity.

• Prohibit back-channel “just check with me” agreements. If an op partner wants a say, it must live in the reserved list. Put this in writing.

Why it works: shadow approvals destroy trust and time. Clean lines speed both.

4. Manage Decision Cycle Time as a P&L Metric

• Instrument decision age in your systems. Add “submitted,” “first response,” and “final decision” timestamps to workflows in CRM/ERP/ITSM. Track by decision type, decider, function.

• Publish a weekly Decision Review: 30 minutes, Fridays. Agenda: aged items violating SLA; blocked-by-approver count; top three cycle-time wins/losses; one closed-loop post-mortem. Escalate on time, not title.

• Tie manager performance to cycle-time adherence on their decision types. Reward speed within guardrails and clean documentation. Penalize hoarding and ghosting.

Why it works: what you track moves. Decision velocity is no different from DSO or on-time delivery.

Implementation Notes (because you’ll ask)

• Start with money decisions. You’ll get the fastest payback. Capex, vendor selection, pricing exceptions, credits/chargebacks.

• Don’t over-document. If your catalog takes more than two weeks to draft and publish, you’re writing policy, not designing rights.

• Expect edge cases. Keep the appeal path short. Document one paragraph of rationale on exceptions and move on.

• Revisit thresholds quarterly for two quarters, then semi-annually.

Executive Takeaways

• Complexity isn’t the enemy at $30M. Fuzzy decision rights are.

• The multiplier that kills you is approver count, not data quality.

• Raise thresholds. Remove the CEO and board from sub-material approvals.

• Name a single DRI for cross-functional decisions. End unanimity.

• Put SLAs on decisions. Escalate on time, not title.

• Track decision age like DSO. Publish it. Manage it.

• You’ll see margin, forecast accuracy, and morale improve within a quarter.

Closing

Speed comes back the moment one person can say yes within a known guardrail and time box.