Legacy SaaS dossier · salesforce

Salesforce

CategoryCRMSales Cloud Enterprise, 25 seatsSite ↗

A contacts list, a pipeline, and a notes field, sold back to you with eighteen modules, four certifications, and a consultancy tier.

01 · Data

Data the SaaS sits on

Gmail / OutlookYours

Google CalendarYours

LinkedIn (contacts)API

Stripe / billingAPI

Internal Build Score

0.72

Feasibility of a useful internal substitute for the actually-used 20% of this SaaS, built with Claude and your own data layer.

IBS = 0.25·U + 0.25·C + 0.20·D + 0.15·I + 0.15·Mthis record · 72%

U
User adoption gap
weight 0.250.88
How much licensed surface area sits idle.
C
Core simplicity
weight 0.250.85
How small the actually-used 20% is when re-implemented honestly.
D
Data portability
weight 0.200.55
How clean the exit is, exports, formats, audit lock.
I
Integration burden (inverse)
weight 0.150.40
How few external systems must keep working through the SaaS.
M
Moat erosion
weight 0.150.80
How much the original differentiator has been commoditised by LLMs + cheap infra.

Editorial heuristic. Not investment advice, not a procurement recommendation.

03 · Feature usage

Commonly used vs. rarely touched

Commonly used

Accounts + contacts list
Deal pipeline with stages
Activity notes per record
Forecast roll-up by rep
Email + calendar capture

Rarely touched

Lightning App Builder pages no one ships
Process Builder + Flow + Apex, three half-overlapping automation engines
Einstein scoring nobody trusts
Tableau CRM bolted on after MuleSoft
Knowledge / Communities / Experience Cloud
Salesforce Inbox + Sales Engagement + Sales Dialer SKUs
Validation rules a $400/hr admin maintains
Sandbox / change-set release process

Build it yourself→

Internal build

Build it yourself

01 · Stack & flow

Postgres

Next.js

Claude API

Resend

Clerk

Internal build map

Data in

Inbound email

Calendar

Accounts table

Agent layer

Note extractor

Weekly digest

Logic

route to deal

schema validate

upsert row

risk flag → Slack

Outputs

Updated deal row

Weekly digest

Forecast view

Build time: 6 weeks
Run cost: $8/seat·mo
vs. SaaS: $180K / yr

Build outline

Postgres schema: accounts, contacts, deals, notes
Email + calendar inbox sync (Nylas / Postmark MX)
Claude classifies + extracts to deal fields
Weekly pipeline digest via Resend
Forecast roll-up SQL view, no AI required

02 · Claude prompt

First, map what you actually use (optional)

Step 1 · usage audit prompt

// Salesforce usage audit, scope the rebuild Input: the object + field population report (rows per object, % of fields populated), the installed managed-package list, active Flows / Apex triggers, and 90 days of feature-adoption stats by profile. ## Do 1. Mark every object, field, and automation as used, rare (< 5% of records or sessions), or dead. 2. Output the minimum schema (objects + fields) that covers 95% of writes. 3. Separate automations with real downstream effects (billing, outbound email) from ones that only touch Salesforce-internal fields. ## Output JSON: { keep_schema[], keep_automations[], drop[], evidence }.

Step 2 · rebuild the core

// CRM note → field extractor, internal sandbox You are a sales-ops assistant inside a Postgres-backed CRM. You receive one account or deal note (plain text, may include forwarded email) and the deal's current row. ## What you must do 1. Extract structured updates only, do not invent facts. 2. Set fields that the note clearly supports: - next_step (≤ 8 words, action verb first) - stage_change ("none" or one of the project's stage enum) - amount_signal (USD number or null) - close_date (YYYY-MM-DD or null) - risk_flag (boolean) and one-line reason if true 3. Quote the supporting sentence from the note in `source`. ## What you do not do - Do not "fill in" missing values from prior context. - Do not score, qualify, or recommend, just extract. - Do not write the reply email; the rep does that. ## Output Strict JSON only. No prose, no markdown fences.

03 · Result

Spoke w/ Acme CFO, needs $400k by EOQ, board meeting Sept 22.

deal:acme-q3

next_step: send EOQ-ready quote · amount_signal: 400000 · close_date: 2025-09-22

← Legacy SaaS index AI wrappers index →