Mind Maps — Current vs Future State

Draft

Updated 22 May 2026

Two pictures of the same business. The first is how things work today, based on our meeting. The second is what we're proposing instead. Read them side by side, then jump to the delta at the bottom to see what changes and what stays.

As-is · 2026

How the business runs today

A templated EPOS from a third-party vendor, sitting on top of incomplete data, with manual workarounds layered everywhere people have given up.

mindmap
  root((Current State))
    Hardware
      2 touchscreen tills (Windows PCs)
      Old thermal receipt printers
      "Slow — receipts take up to ~1 min"
      Barcode scanner (old)
      Cash drawers
      All owned outright (bought from Nadeem)
    EPOS Software
      Magnum EPOS Ltd (Bolton)
      Developers in India
      Templated — reused across many clients
      Per-license fee
      No developer documentation
      "£80k quote just to fix shortcuts"
    Sales Process
      Shortcut buttons removed
      "Shortcuts didn't decrement stock"
      Staff must type product name on every sale
      Slower per transaction
      Trade customers waiting longer
    Stock Data
      11500 products
      Most have no stock level
      Most have no cost price
      Most have no supplier link
      No date stamps on prices
      No barcode for timber & loose items
      "Staff have given up maintaining it"
    Credit Accounts
      All balances zeroed (data was unusable)
      No part-payment flow
      No split tender on repayment
      No credit-limit enforcement
      "Staff skip updates — too slow"
    Payments
      DNA Payments (no EPOS integration)
      Amounts typed manually on card machine
      Risk of typos & fraction errors
    Staff
      3 on counter
      Shared till logins
      No transaction-level accountability
      No upsell incentives
    Reporting
      Z & X readings per till
      Basic VAT and cash totals
      Most reports unused or broken
      Partial exports only
    Customer Reach
      In-shop only
      No website
      No online presence
      Brand recognised locally

What this picture tells us

The EPOS isn't really the problem — the data underneath it is. Stock, credit, prices, supplier links — all broken at the data layer. The software is just the surface where the symptoms show up.
Staff workarounds compound the breakage. When one thing doesn't work, people stop maintaining the rest. Shortcuts were ripped out because they didn't sync to stock. Credit balances were ignored because updating them was too slow. The system trained the team to give up.
There's no accountability trail. No per-user login means no way to attribute a problem (or a great sale) to a person.
The business is one customer-facing layer away from being invisible online. No website, no social, no click-and-collect — pure walk-in trade for a shop on a busy road.
Cost ceiling already set: Magnum quoted £80,000 to fix the shortcuts→stock link. That's now the implicit budget yardstick for any alternative — anything materially below this with materially more capability is a clear win for the owners.

To-be · Proposed

How the business should run

A clean database at the centre. A browser-based EPOS on the existing tills. A customer web app on top. Multi-branch ready from day one. Built specifically for this shop, owned outright.

mindmap
  root((Future State))
    Hardware
      Keep existing 2 tills - Windows PCs
      Add 3rd till if needed
      New fast thermal printers
      New barcode scanner
      Runs in browser on any OS
    Custom EPOS Web App
      Built specifically for this business
      Owned outright - no license
      Full developer documentation
      Cross-platform browser interface
      Easy to extend and integrate later
    Clean Data Layer
      One source of truth - modern database
      Products by barcode, weight or length
      Cost and retail price history with dates
      Suppliers properly linked
      Min and max stock levels per product
      Multi-branch ready from day one
    Smart Sales Process
      Configurable shortcut buttons
      Shortcuts now decrement stock correctly
      Weight and length entry inline at till
      Fast common-product flows
    Credit Account System
      One account view per customer
      Part-payments supported
      Split tender with cash and card on repayment
      Credit limit enforced at point of sale
      Customers in credit report
    Payment Integration
      Dojo Payments - amount auto-pushed
      No manual typing on card machine
      Receipts unified across till and terminal
    Staff and Accountability
      Per-user PIN or card login
      Every transaction tagged to a staff member
      Sales-based incentives become possible
      Discrepancies traceable to one person
    Rich Reporting
      Per-till Z and X readings
      Per-staff sales and performance
      Customers in credit with balances
      Stale-price filter for old prices
      Stock reorder suggestions
    Customer Web App - Phase 2
      Product-first search-first
      Phone-number OTP login with no passwords
      Click and collect
      Optional delivery later
      Trade-customer credit view online
    Future Growth
      Multi-branch central management
      Branch-level stock and ordering
      Cross-branch reporting
      Plug-in integrations with bookkeeping

Why this shape, not another

Database first, EPOS second, web app third. Everything else fails if the data layer isn't clean. We build that once, and then both the till and the customer web app are just different views onto the same data.
Reuse the hardware that already works. The existing tills are just Windows PCs — there's no reason to throw them out. A browser-based EPOS runs on what's already there and saves ~£1,200–£1,500 per replaced till.
Multi-branch from day one. The owners' 5-year goal is to scale and sell. If the second branch is plausible within 2 years, "branch" must be a concept in the data model from the start — bolting it on later is expensive.
Integration-friendly by design. Dojo for payments now, bookkeeping software later, supplier feeds eventually. Each is a small adapter on top of the clean data, not a structural change.
Owned, not licensed. Full documentation, full keys. No £80k quotes from a third party to change a button.

Delta

What changes, what stays, what's new

A quick three-column view so the owners can see at a glance what's kept (no spend), what's gone (no longer paying for), and what's net new.

Kept (no spend)

The 2 touchscreen tills (Windows)
Cash drawers
The brand & physical store
The supplier relationships
The credit-customer relationships (once cleaned)
Z & X reading concept (just done properly)

Gone (no longer needed)

Magnum EPOS per-license fees
Dependency on third-party developers in India
The £80k "restructure" quote
DNA Payments (replaced by Dojo)
Old slow thermal printers
Manual typing of card amounts
Manual reconciliation of credit accounts on paper
Staff give-up cycle on bad data

New (net additions)

Clean product database (single source of truth)
Weight & length product types as first-class
Price-change history with dates
Real credit-account flow (part payments, split tender, limit enforcement)
Per-user staff login & accountability
Per-staff sales reporting & incentive base
Dojo Payments integration (auto-amount)
Faster receipt printers
Customer web app — search, phone OTP login, click & collect
Multi-branch architecture baked in
Bookkeeping & supplier-feed integration points
Full ownership of the codebase + documentation

Working draft · share feedback with Ilya before the next call · back to portal