I have sat in the meeting where the CFO points to a single contract cycle time number and asks the obvious question. If cycle time is down, why does it still feel like deals are stuck. Sales says legal is slow. Procurement says approvals are slow. Finance says signatures are slow. Everyone is right, and the metric still looks “better.”

The hidden reason is simple. A blended cycle time KPI is not a performance metric. It is a composition metric. It moves when your contract mix moves, not just when your process improves, and it compresses wildly different work into one average that hides the tail.

This is more of a measurement problem than a tooling problem. Here’s why.

The flaw is not in your contracts, it is in your math

Contract cycle times behave like most operational cycle times. The distribution is not tidy. It is skewed. A small percentage of contracts take a very long time because they trigger negotiation, escalations, exceptions, or counterparty delay.

That is exactly the trap described in the HBR flaw of averages framing, where plans based on “average conditions” fail because averages erase variability. In contracting, that variability is not random. It is structural.

If leadership is using one number to decide whether to hire, whether to change policy, or whether to accept more risk, the average is the wrong instrument. The business does not experience the average. The business experiences the deal that is in front of it, and the deals that matter most often live in the tail.

So when legal ops reports “12 days average cycle time,” sales hears “my strategic deal will be 12 days.” Finance hears “revenue timing is predictable.” The board hears “control is stable.” None of those are safe inferences.

Your blended KPI is mostly measuring contract mix

Here is the part teams miss because the metric looks objective.

If 70% of your volume is NDAs and low-risk templates, and 30% is revenue, vendor, and strategic paper, then the blended average mostly tells you how fast you move the high-volume, low-complexity work. That can be good work to optimize, but it is not the work leadership is anchoring decisions on.

Now change the mix for one quarter. A new product launch drives a flood of partner NDAs. Procurement runs a vendor consolidation and reduces new supplier paper. The blended cycle time drops, even if nothing improved for complex deals. Everyone celebrates. The next quarter, complex paper returns and the metric “regresses,” and now leadership assumes the team backslid.

This is the same pattern the Harvard CLP contracting complexity report is warning about in operational terms. Contracting is portfolio work at scale, with different stakeholders and different risk tradeoffs per work type. A single blended KPI collapses the portfolio into a fiction.

If you want one executive number, you can have one. It just cannot be a raw average across everything.

Approval architecture is usually the real bottleneck

Most “cycle time” debates are actually debates about approvals, not legal review.

The ACC guidance on streamlining approvals is blunt about what drives delay. You need a written Delegation of Authority, you need thresholds, and you need contract-type specific approval flows. Without that, contracts queue in inboxes, get re-litigated late, and ping-pong between functions with no single owner.

This matters for metrics because approval time is structurally different from negotiation time.

• Negotiation time can be reduced with playbooks, fallback positions, template quality, and better counterparty engagement.
• Approval time is a governance design problem. It is about who has authority, when escalation is required, and whether approvals add signal or just add steps.

When leadership is angry about cycle time, they usually want two things that conflict. They want speed and they want control. A DOA policy is how you make that trade visible. It makes explicit which contracts can move fast and which contracts should move slower because the organization is making a deliberate control choice.

If you do not separate “legal drafting and negotiation” from “internal approvals,” your metrics will keep pointing at the wrong root cause.

Value erosion is the cost of getting this wrong

Cycle time debates become political because they are really about money.

The Deloitte ROI of Contract Excellence summary cites average contract value erosion of 8.6%, with best performers much lower and worst performers far higher. That erosion is not only about post-signature compliance. It is also about fragmented workflows, weak data, and lack of insight into where value is being won or lost.

WorldCC has been consistent about the structural drivers of complexity and variability across portfolios. The WorldCC Keep it simple report describes how differing levels of contract complexity and the inability to access reliable data drive operational drag and executive blind spots.

When you report a blended cycle time, leadership may think they are managing efficiency. What they are really managing is value leakage. They just do not know where it is coming from.

The HBR article on AI and contracts makes the same point in a different register by describing how inefficient contracting can drive material value loss depending on circumstances. The key phrase there is “depending on circumstances.” That is segmentation. The cost is not uniform across your portfolio.

Segmentation changes the decisions leadership can make

Once you segment cycle time by contract type, leadership stops arguing about anecdotes and starts choosing levers.

Here is what changes in real decision-making.

1. You stop funding the wrong fixes.
   If NDAs are already fast and your pain is revenue paper, investing in NDA automation will improve the blended KPI and do almost nothing for bookings.
2. You stop blaming legal for control choices the business made.
   If a vendor agreement is slow because information security, privacy, and finance must sign off, that is not a legal speed problem. It is a risk gating design choice.
3. You can defend headcount requests with specificity.
   “Cycle time is high” is weak. “MSAs over $250K have a 90th percentile approval delay of 18 days in finance queue” is actionable.
4. You can surface where negotiation is actually happening.
   Contract type segmentation paired with stage timing shows whether delay is counterparty redlines, internal review, or approvals.

This is also where I see tools matter. In day-to-day practice, I use Concord to keep the conversation grounded in actual slices of the portfolio, not vibes, and the Concord AI Reporting approach is built around structured, repeatable reporting rather than one-off dashboard debates. When leadership asks “how many contracts are stuck in approvals,” you need deterministic reporting, not a best-effort estimate.

A measurement model that holds up in the exec room

If you want cycle time metrics that leadership can trust and act on, do six things.

1) Define a contract taxonomy that reflects how the business works

Do not start with your CLM’s default categories. Start with how work is staffed and where risk lives. A practical starting point is 6 to 10 contract types: NDA, order form, customer MSA, partner, vendor, employment, real estate, and any regulated outliers.

The portfolio lens in the Harvard CLP report supports this, because contracting complexity is not one thing. It varies by purpose, counterparty power, and stakeholder load.

2) Add a risk tier that controls approvals

Contract type alone is not enough. A low-dollar vendor renewal and a new strategic outsourcing agreement are both “vendor.” They should not have the same path.

This is where the ACC DOA guidance is operationally useful, because it ties contract categories to thresholds and assigns authority to roles.

3) Split cycle time into stages that map to root causes

At minimum, track four clocks.

• Intake to first legal touch
• Legal negotiation time
• Internal approvals time
• Signature to full execution

This is how you keep “legal speed” separate from “organizational gating.”

4) Report percentiles, not just averages

Executives understand a service-level framing. They will accept “half of NDAs finish in two days” and “10% take over ten days” because that matches reality. It also aligns with the measurement logic from the HBR flaw of averages without turning the conversation into a stats lecture.

5) Add a mix-adjusted KPI for the one-number crowd

If your CEO insists on one number, give them a mix-adjusted metric that holds contract mix constant quarter to quarter. That way improvements are real process improvements, not shifts in volume.

The WorldCC Benchmark report 2023 is useful context here because benchmarking is only meaningful when you compare like with like.

6) Pair speed with a quality signal

Speed without quality is how you buy future pain. The “quality” proxy can be simple: exception rate from playbook, number of escalations, or post-signature dispute volume.

This keeps cycle time from becoming a vanity metric.

What I tell leadership when the blended KPI is already out there

I do not try to “fix the number” first. I fix the decision the number is being used to justify.

I will typically say: our blended cycle time moved this quarter because the portfolio shifted, and here is what happened in the segments that drive revenue and risk. Then I show a segmented view with stage timing and percentiles. After that, we agree on one or two interventions, usually approvals architecture and one negotiation bottleneck.

This is also where a disciplined explore-then-measure workflow matters. The Concord AI Reporting trust model describes the difference between exploratory search and deterministic reporting in plain operational terms. Leadership does not want exploration when money is on the line. They want accountable numbers tied to defined criteria.

Blended cycle time metrics are not malicious. They are just the wrong abstraction for a heterogeneous portfolio. Once you segment by contract type and risk tier, the politics drop. The levers show up. The work becomes manageable again.