What Are the Four Layers That Keep Prices and Quotes From Drifting?
I settled on four layers meant to keep my services, prices, quotes and job records from drifting apart: Owner Development and Strategy, the Commercial Master, Generators and Software Applications, and Client and Job Records. This post walks through how those layers connect.
What's the difference between business management and service operations?
In How Did I Build the Business Behind Ross Core IT?, I wrote about how I arrived at this structure. This post stays with the distinction that made the structure necessary, and with how each layer works.
Eventually, I came across the term "service operations." That helped me put a name to the distinction I had been trying to understand.
The way I came to understand it, business management sets the direction, rules, and boundaries of the company. Service operations takes a customer's job through those boundaries without reinventing the process every time.
When someone hires me, the operations side asks: What information do I collect? What am I quoting? What needs approval? What do I check before handing over the work?
That was when organizing the business started to mean more than organizing folders. The four layers are how I gave that distinction a structure I could actually use. The first two layers hold the direction and the approved information. The last two are what I use when a customer's job comes in.
What belongs in Owner Development and Strategy?
This is where I keep the operating standards, playbooks, business concepts, and rules I use to make decisions. It helps me work out how I want the business to run.
My generators never read from this layer directly. That is intentional. This layer is for thinking through how I want the business to run, not for feeding a quote. If I change a playbook or a rule here, that change should not show up in a proposal until I have approved it into the Commercial Master.
What does the Commercial Master hold?
This holds the approved service descriptions, prices, scope boundaries, and editable document sources.
I wanted one authoritative place for that information. If I change a price or update a service, I need to know which version is current. I also need to know which version was used for an older quote.
Each approved version is tracked by edition, and the files themselves are identified by a SHA256 hash. That gives me a current edition to generate from, and a way to recognize an older edition later.
How do the generators and software applications use that information?
These are the tools that use the approved information. My proposal builder and web manager belong here.
I didn't want each application maintaining its own separate copy of my prices and business wording. That would give me more places to update and more opportunities for things to drift apart.
The applications use service IDs, editions and hashes to locate approved sources through the shared catalog. I do not want each tool maintaining a competing price list or its own authoritative business wording. Historical quote snapshots are retained as evidence of that job, rather than treated as the current source. The applications read the catalog; they do not write to it.
What do the client and job records keep?
This is the record of what happened on a particular job. Client, in this layer name, means the customer record. Which service was quoted? Which edition was used? What did the customer accept? What was delivered?
Updating my service information later should not silently change an agreement someone already accepted. Each quote revision captures its source and pricing snapshot when created. Acceptance locks the specific accepted revision and its supporting evidence; it does not replace the original snapshot.
How the system works: the catalog, the hashes, and the path from audit to promote
Because I was building Client Manager alongside Proposal Builder, I connected them through a shared catalog rather than making each application maintain its own commercial sources.
The shared catalog is a file called catalog.manifest.json. JSON is a plain-text format for structured data. The catalog identifies service IDs, approved editions, source references and hashes. It points to approved Services sources; it does not maintain a second price list. Pricing stays in those approved sources, and any custom job amount is recorded separately on the quote.
When approved sources change, I update the handoff information so the applications know what they should consume. The catalog is that handoff.
The hash. A hash is a fingerprint of a file's contents that helps identify the exact version. I use SHA256. If even one character in a file changes, the hash changes. Matching hashes are a reliable way to verify that two files have the same contents. Microsoft's Get-FileHash documentation covers this, and SHA256 is its default. NIST's FIPS 180-4, the Secure Hash Standard, says these digests are used to detect whether messages have been changed since the digests were generated.
The job snapshot and source review. When a quote revision is created, it captures the approved source and pricing evidence used for that revision. Acceptance locks the specific revision and its supporting evidence; it does not replace the creation snapshot. Changes before acceptance create a new draft revision. Changes afterward require an approved amendment or change order. If an approved source changes after a quote snapshot, the system can flag it as "Services source changed - review required". I review the difference and decide whether it applies. Reviewing the source, creating a new revision and generating its document are separate steps. A new source edition does not rewrite an accepted agreement.
A consistent starting point for AI. I use AI to inspect what exists, update the handoff information, and generate something new into a candidate area, so I started standardizing how I ask it to work on the system, with three prompts:
- Audit and diff. Before changing commercial documents, inspect what exists and explain the differences.
- Handoff manifest. When approved sources change, update the handoff information so the applications know what they should consume.
- Isolated staging. When generating something new, put it into a separate candidate area and check it before making it current.
The order I follow is audit, review, approve, generate, verify, promote. New work is checked in that candidate area before it becomes current. I wanted future sessions to have a clear starting point instead of making a fresh set of assumptions about how my business works.
Where does that leave this structure now?
I verified the connected Services, Proposal Builder and Client Manager workflow end to end in an isolated test. That run covered a source change, review, a new quote revision, real PDF generation, acceptance locking, a blocked edit and a linked amendment. I am extending that tested foundation into Ross Core IT’s operating workflow; the test does not mean every service or ordinary client job is connected yet.
The four layers are the structure I use behind the website: a consistent way to define the work, quote it, deliver it, and keep a record of what was agreed, for people and businesses in Holbrook, Ronkonkoma, Bohemia and across Suffolk County.
The story of how I got here is in How Did I Build the Business Behind Ross Core IT?.
Frequently asked questions
Why don't the generators read Owner Development and Strategy directly?
That layer is where I work out how I want the business to run. Generators never read it directly. They use only the approved information in the Commercial Master.
What happens when I update a service after a quote has already been created?
Each quote revision captures its source and pricing snapshot when created. Acceptance locks that revision and its supporting evidence without replacing the original snapshot. If an approved source changes after a quote snapshot, the system can flag it as "Services source changed - review required". I review the difference and decide whether it applies. Reviewing the source, creating a new revision and generating its document are separate steps. A new source edition does not rewrite an accepted agreement. Changes before acceptance create a new draft revision; changes afterward require an approved amendment or change order.
What is catalog.manifest.json?
It identifies service IDs, approved editions, source references and hashes. Applications use it to find approved Services information. The manifest itself is price-free.
What is the audit, review, approve, generate, verify, promote order?
It is the order I follow when AI or a generator works on the system: inspect what exists, review and approve the change, generate into a separate candidate area, verify it, and then make it current.
What's the next step?
Call (631) 229-4446 or use the contact page at rosscoreit.com/contact. By appointment.
Sources
Most of this post is my own experience building my own business. For the technical terms I mention:
- Microsoft, Get-FileHash (Microsoft.PowerShell.Utility)
- NIST, FIPS 180-4, Secure Hash Standard (SHS)
- IETF, RFC 8259: The JavaScript Object Notation (JSON) Data Interchange Format
