Structural View discovery intermediate

Application Landscape Diagram

Every application in the estate placed against the business areas it serves, so duplication and orphans become visible.

Merger & Acquisition Integration Legacy Migration Cost Optimisation Cloud-Agnostic Cross-Industry CxO Architecture Review Board Finance

flowchart TB
  subgraph sell["Sell"]
    direction LR
    s1["CRM<br/><i>SaaS · Sales</i>"]
    s2["Quoting<br/><i>bespoke · Sales</i>"]
    s3["Pricing A<br/><i>legacy · Sales</i>"]
    s4["Pricing B<br/><i>legacy · Regional</i>"]
  end
  subgraph fulfil["Fulfil"]
    direction LR
    f1["Order Mgmt<br/><i>bespoke · Ops</i>"]
    f2["WMS<br/><i>COTS · Ops</i>"]
    f3["Carrier Portal<br/><i>SaaS · Ops</i>"]
  end
  subgraph support["Support"]
    direction LR
    b1["ERP<br/><i>COTS · Finance</i>"]
    b2["Billing<br/><i>mainframe · Finance</i>"]
    b3["HRIS<br/><i>SaaS · HR</i>"]
    b4["Expenses<br/><i>SaaS · unowned</i>"]
  end
  s3 -.->|"same capability"| s4

What it is

The estate laid out against the business, one box per application, annotated with type and owner. It is deliberately a map rather than an integration diagram — the arrows are kept out so that the picture stays readable at fifty or five hundred applications.

The dotted line in the example marks the finding: two systems doing the same job for different regions. That, and unowned systems, are what the artifact exists to surface.

When you produce it

At the start of a rationalisation, a merger integration or a cost programme, and whenever somebody asks "how many systems do we actually have?" and nobody can answer within a factor of two.

Who reads it

Executives and finance, who see duplication and spend. Enterprise architects, who use it as the inventory everything else references. Application owners, who usually discover a neighbour they did not know existed.

What good looks like

Bands are business areas, not departments, so the map survives reorganisation.
Each box carries type (COTS, SaaS, bespoke, legacy) and a named owner.
Duplication is annotated, not left for the reader to spot.
Unowned applications are marked. They are the cheapest wins and the biggest risks at the same time.
Sourced from a maintained inventory, with a date on the page.

Common mistakes

Adding integration arrows until it is unreadable — that is a different artifact.
No ownership column, which removes the ability to act on anything.
Counting environments as applications, inflating the number and the conclusions drawn from it.