The 4-Quadrant Framework: What to automate, what to delegate, what to centralize — and what to kill
Not every task deserves automation. Some deserve termination. And that's one of the statements that most unsettles legal departments when said for the first time.
The previous edition introduced the concept of a productivity layer. But before adopting any layer, there's a prerequisite exercise that many departments skip — and skipping it is expensive later: deciding, with clarity, what each category of work deserves.
Without this exercise, automation only accelerates chaos. You save ten minutes per request but keep processing requests that never needed to exist.
The mistake of 'automate everything' is that automation accelerates what exists. If what exists is waste, you end up wasting faster.
The two axes: value and frequency
Axis 1: value added
Does this task require genuine legal expertise, generate significant business impact, or both? High-value tasks are those where the attorney's presence makes a real difference. Low-value tasks are those where any reasonably trained person would do equally well.
Axis 2: frequency
Does this task happen many times (weekly or more), or sporadically (monthly or less)? Frequency is what makes automation viable: tasks that appear five times a year rarely justify the investment of automating them.
Crossing these two axes creates four quadrants. Each requires a different strategy — and treating all tasks with the same strategy is the most common mistake in legal transformation projects.
Quadrant 1 — High value + high frequency: centralize and create visibility
This is where the department's most important work lives. Recurring contract review for strategic clients or vendors. Structural litigation portfolio management. Decisions on recurring regulatory matters. These tasks need senior counsel, and they appear frequently.
The temptation is to automate — after all, it's high frequency. That's a mistake. Generic automation removes the human judgment that justifies the high value. The right strategy is to centralize and give visibility: all instances flow through a single workflow, with defined criteria, consistent records, and clear owners.
Quadrant 2 — High value + low frequency: protect (and protect well)
Strategic cases. Critical company decisions. Complex opinions on rare topics. High-stakes negotiations. This quadrant defines legal's contribution to the business — and it's where the team spends the least time, precisely because the other three quadrants consume the calendar.
The test question: how much of your most senior professionals' time is spent on Q2 decisions? If the answer is less than 30%, something is wrong with allocation.
Quadrant 3 — Low value + high frequency: automate aggressively
Status updates to business units. Standard NDAs. Internal system data entry. Answering FAQs about standard clauses. This is the natural quadrant for automation. Bloomberg Law found that attorneys spend two of every eight hours on administrative duties. The InView report estimated three hours per day on manual tasks. In smaller teams, the percentage reaches 40%.
Automate aggressively means three things: identify all recurring tasks in this quadrant (not just the obvious ones), build self-service wherever possible (internal FAQs, self-serve templates, intake portals), and use technology for what remains (workflow automation, generative AI for drafting, system integration).
Quadrant 4 — Low value + low frequency: kill (yes, kill)
This is the uncomfortable part. There's a set of tasks in every legal department that survives by inertia, not by necessity. Reports nobody reads that keep being produced because they always have been. Ceremonial meetings that generate no decisions. Multi-level approval processes for trivial matters.
The test question: if this task stopped being done tomorrow, how long until someone noticed, and what would the real impact be? If the answer is nobody would notice, or would notice late, or would notice without practical consequence, it's in Q4. Courage to kill is the most underused management skill in legal departments.
How to apply the framework in a 90-minute meeting
- First 30 minutes: brainstorm all categories of work the department does. Not individual tasks — categories. Aim for 25 to 40 categories.
- 30-60 minutes: classify each category on both axes (high/low value, high/low frequency). Discuss disagreements — they reveal more about the department than the agreements.
- 60-90 minutes: for each quadrant, choose one to three categories for immediate action. Q1: how to centralize? Q2: how to protect time? Q3: what to automate first? Q4: what will we stop doing this month?
The result is a concrete action plan, with up to twelve items, based on self-diagnosis — not on a technology vendor's assessment. This plan becomes the basis for any subsequent conversation about tool adoption, including the productivity layer. And it becomes the basis for the business case you need to present to leadership.
What's coming in the next edition
You've done the diagnosis. Applied the framework. Know what needs to be automated, centralized, protected, and eliminated. The next edition is about the hardest part: presenting this to the CFO, the COO, the CEO. How to translate what makes sense in your head into language that makes sense to whoever approves budget. Because productivity, on its own, doesn't convince the person who signs the check.
