How a Consulting Firm Saved 15 Hours Per Week Using AI Prompts

Meridian Advisory Group is a 45-person management consulting firm specializing in operational transformation for mid-market manufacturing companies. Like most consulting firms, their biggest constraint wasn't client demand — it was consultant time. Their senior consultants were spending 35-40% of their time on research synthesis, secondary data analysis, and report writing rather than client-facing strategy work.

In Q1 2024, they implemented a structured AI prompt library across three core workflow areas. Nine months later, the results are measurable and significant.

The Problem: Time Leaking Into Low-Leverage Work

A time audit across 12 senior consultants revealed the breakdown: 8 hours/week on research and secondary data synthesis, 5 hours/week on first drafts of deliverables and proposals, 4 hours/week on formatting and structuring client-ready documents, and 3 hours/week on meeting summaries and action item documentation. That's 20 hours per consultant per week spent on tasks that, with the right prompts, AI could handle a significant portion of.

The Implementation: Three Prompt Libraries

Library 1: Research Synthesis

They created 12 prompts specifically for taking raw research inputs (industry reports, interview notes, data tables) and synthesizing them into structured executive-ready summaries. Each prompt was templatized so consultants could drop in client-specific content and get a structured synthesis in minutes instead of hours.

You are a senior management consultant specializing in [industry]. I will provide you with [X] pieces of research. Synthesize these into a structured 400-word summary that: (1) identifies the top 3 industry trends most relevant to a mid-market manufacturing company, (2) highlights key data points with source attribution, (3) flags any conflicting findings, (4) suggests 3-5 strategic implications. Format as a structured summary with clear sections.

Library 2: Proposal Drafting

Proposal writing had been one of the most time-consuming activities, with senior consultants spending 4-6 hours on each new proposal document. They built a 6-prompt sequence that walked through each section of their standard proposal format: situation analysis, problem framing, proposed approach, team and credentials, timeline, and investment.

Library 3: Client Report Structure

Their standard client deliverable was a 20-30 page findings and recommendations report. Building the structure, writing executive summaries, and turning interview quotes into narrative had been the most time-consuming part. AI now handles the structural scaffolding and first drafts of all non-analytical sections.

The Results: 9 Months Later

• 15.2 hours saved per consultant per week on average (across all 12 senior consultants)
• Proposal turnaround time reduced from 4.5 days to 2 days
• Client satisfaction scores increased 12% (consultants now have more time for client contact)
• The firm took on 22% more client engagements with the same team
• ROI on their AI tooling: estimated 18x return in first year

What Made It Work

The firm's managing partner credits three things: (1) they documented prompts as shared assets rather than individual discoveries, (2) they assigned a "prompt champion" in each practice area to own prompt quality and iteration, and (3) they set a rule — any consultant who found a prompt that consistently saved 30+ minutes had to add it to the library within the week.

The prompts aren't replacing our consultants' thinking. They're eliminating the mechanical work that got in the way of the thinking. That's the entire point. — Managing Partner, Meridian Advisory Group

The lesson here isn't specific to consulting. Any knowledge work organization where skilled professionals spend significant time on first-draft production work can replicate this model. The investment is a few weeks of prompt development. The return is permanent.