Line-of-business survey · August 2026

The work nobody should be doing.

We surveyed the ten desktop applications our clients depend on most, and traced every repetitive task back to a documented integration point. Fifty opportunities. Forty-five of them high impact. Seven we would scope first.

0
Client installs
0
Endpoints
0
Opportunities
0
Highest confidence
01

Where the leverage is

Every opportunity plotted by how hard it is to build against how much it returns. The cell in the top left is where we start — small builds, high return, and the integration already exists. Click any square to read the full case.

Opportunity
Small effort, high impact
Filter by vertical to see where each client type sits.
02

The ten applications

Ranked by how many client organisations run them. Each one carries five opportunities. Open a row to see what we found.

03

Start here

Seven builds that are small, high return, and rest on integration surfaces the vendor already documents. No new licensing, no platform migration, no change to how anyone works.

04

How we built this

Every opportunity is traced to a vendor-documented integration point. Where one does not exist, we say so rather than inventing it.

Installed base first

Ranked by number of client organisations running each application, drawn from our own managed client base rather than market research. That base is not exclusive to Nova Scotia — read these as a profile of the software businesses of this size run generally.

Documented surfaces only

Six applications expose a public API. Three are partner-gated. One offers database or export access only. Where the honest mechanism is desktop automation, we labelled it that way.

Sized before promised

Each build carries an effort and impact rating and a written risk note covering what could go wrong, what needs human approval, and what must never run unattended.

Human in the loop

Nothing posts to the books, files a return, or sends a client document without a person approving it. Automation prepares and validates; people decide.