For Construction Cost Estimators ·
What you'll accomplish
You'll upload your past bid summaries, job cost reports, and unit cost records to Claude Pro and be able to ask questions like "What did we pay for concrete slab-on-grade on our last 5 healthcare projects?" instead of hunting through old files. This turns years of institutional cost knowledge into a queryable resource that junior estimators can access, one that doesn't walk out the door when a senior estimator retires.
What you'll need
Before uploading, collect what you have. You don't need everything. Start with what's accessible:
Organize files by project type in a folder (Healthcare, Office, Warehouse, Education).
What you should see: A folder with 10-30 files covering past bids and job costs.
Claude Projects let you maintain persistent context across multiple conversations, which makes them perfect for a reference database you'll query repeatedly.
What you should see: A new project workspace with an option to add files and set project instructions.
Click on Project Instructions (or similar settings area) and add:
You are a cost reference assistant for a commercial construction estimating team. You have access to our historical bid summaries and job cost reports. When asked about costs, always:
1. State which project(s) the data comes from
2. Give a range (low/high) not just an average
3. Note whether the cost is from a bid estimate or actual job cost
4. Flag if the data is older than 3 years (may need escalation adjustment)
5. Ask for project type and size context if not provided
What you should see: Files appear in the project's knowledge base. Claude can now reference these in all conversations within this project.
Start a new conversation inside your cost history project and ask a test question:
What are our historical costs for concrete slab-on-grade, including forming, placing, and finishing? Give me a range and note which projects this comes from.
What you should see: Claude references specific projects from your uploaded files and gives you a cost range. If it can't find data, it will tell you, which helps you identify gaps in your historical records.
Ask a higher-level question to test the analytical capability:
Based on the projects I've uploaded, what's our average cost per SF by project type (medical, office, warehouse)? Include only completed projects with actual costs where available.
Unit cost lookup:
What are our historical unit costs for [work item, e.g., exterior masonry, structural steel, MEP rough-in]? Give a range, note project types, and flag data older than 3 years.
Project type benchmark:
What's our average total cost per SF for [project type] projects? Break it down by CSI division if the data allows.
Bid vs. actual comparison:
For completed projects with both bid estimates and actual job costs, where do we consistently under- or over-estimate? Which trades have the biggest variance?
Labor productivity check:
Do we have any historical data on labor productivity rates for [task]? What crew size and production rate have we used in past estimates?
Escalation query:
What year did we estimate [project]? If I need to escalate those costs to today's pricing, what escalation factor would you suggest?