Repetitive Informational Work Audit
Disrupt automatable human workflows before AI competitors do
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 7
- Confidence
- 97%
The audit treats dependence on repetitive informational labor as a strategic risk signal. A company inventories workflows and marks tasks where people repeatedly retrieve facts, answer similar questions, transfer information between systems, or act as gatekeepers to knowledge. Each exposed workflow is evaluated for repetition, information intensity, customer value, automation feasibility, and the consequences of error. The company then redesigns the strongest candidate before an AI-native competitor does, assigning machines the repeatable information work while retaining humans for judgment, empathy, creative differentiation, and accountability. A small controlled implementation establishes whether the new process improves speed, quality, cost, and customer outcomes. The audit repeats as AI capabilities advance, turning self-disruption into an ongoing operating discipline rather than a one-time technology project.
Origin
Extracted from Marketing Against The Grain as Kipp Bodnar articulated an “AI first principle” for finding vulnerable work across marketing, sales, and customer support.
Core principles
- 01Repetitive informational work is highly exposed to AI substitution
- 02Human gatekeeping is not a durable source of value
- 03Companies should disrupt vulnerable workflows before competitors do
- 04Human effort should move toward judgment, relationships, creativity, and accountability
- 05Early adaptation compounds into operating advantage
How to run it
- 1
Inventory Workflows
List recurring workflows across major functions, including their inputs, actions, outputs, owners, frequency, and customers. Describe what people actually do rather than relying on job titles.
Pro tip Start with high-volume customer-facing workflows because their repetition is easier to observe.
- 2
Flag Informational Repetition
Mark tasks dominated by retrieving, summarizing, explaining, routing, or reformatting information. Note where employees repeatedly answer equivalent questions.
Pro tip Review tickets, call notes, templates, and standard operating procedures for repeated patterns.
Watch out Do not label a whole role automatable when only some tasks are repetitive.
- 3
Find Human Gatekeeping
Identify cases where customers or colleagues must wait for a person primarily to access existing information. Distinguish true judgment from unnecessary intermediation.
Pro tip Ask whether the user could proceed safely if the same information were immediate and conversational.
- 4
Score Exposure and Value
Rank each workflow by repetition, information intensity, automation feasibility, customer impact, and error risk. Select a high-value workflow with manageable consequences for the first redesign.
Pro tip Favor bounded workflows with clear success metrics and reversible implementation.
Watch out Do not begin with a consequential process merely because it offers the largest theoretical savings.
- 5
Redesign the Division of Labor
Assign repeatable informational actions to AI while preserving human ownership of judgment, exceptions, relationships, and accountability. Define escalation conditions explicitly.
Pro tip Require the system to expose uncertainty rather than hiding it behind fluent output.
Watch out Automating an unclear or broken process can amplify its defects.
- 6
Pilot and Measure
Run the redesigned workflow on a controlled sample and compare it with the current process. Measure speed, quality, cost, customer outcomes, and escalation rates.
Pro tip Review failures qualitatively as well as tracking aggregate metrics.
- 7
Repeat the Audit
Reassess workflows as AI capabilities, customer expectations, and business conditions change. Expand only after the pilot demonstrates reliable value.
Watch out A one-time audit quickly becomes obsolete in a fast-changing capability environment.
In the wild
A SaaS company discovers that agents repeatedly explain the same configuration steps. It builds an AI assistant grounded in approved documentation, routes uncertain cases to people, and makes support leaders responsible for reviewing failures and updating source material.
→ Routine answers arrive faster while human agents spend more time on complex diagnoses and sensitive customer situations.
A sales team manually assembles standard product, pricing, and implementation answers for every prospect. The company creates a governed conversational resource for approved information while requiring representatives to handle negotiation, fit assessment, and commitments.
→ Prospects receive routine information sooner without removing human accountability from consequential decisions.
Common mistakes
Automating Roles Instead of Tasks
A role usually combines repetitive work with judgment, relationships, and accountability. Audit at the workflow and task level to avoid destructive overgeneralization.
Ignoring Error Consequences
A workflow can be technically automatable but inappropriate for autonomous execution when mistakes create legal, financial, safety, or trust risks.
Preserving the Old Process Around AI
Adding an AI tool without redesigning handoffs, ownership, and escalation rules often creates extra work instead of eliminating repetition.
Is it for you?
Best for
It is best for organizations with employees repeatedly retrieving, summarizing, explaining, routing, or reformatting information.
Not ideal for
It is not ideal for work whose primary value comes from physical execution, trusted relationships, accountability, or nuanced high-stakes judgment.
From the transcript
“One of my AI first principles is, anything, any business or any part of the business I am working on that has a dependence on…”
“Humans has gatekeepers, humans has access, humans doing the same thing over and over again, that AI is going to do that work and I…”
“Disrupt yourself before the robots come get you.”
From the episode
Google’s Master Plan To Destroy ChatGPT (#119)