Replace the Spreadsheet Ops Stack: Governed Operations Intelligence for NZ SMEs

Most New Zealand service businesses already know the feeling. The monthly ops pack is a spreadsheet. Someone exports CSVs from the system of record, pastes them into a workbook, fixes the formulas, and sends a PDF. Two weeks later a client asks why the numbers differ from last month, and nobody can say which version is right.

That spreadsheet stack is not a failure of discipline. It is what happens when the system of record is good at recording work and weak at answering questions that span modules. The fix is rarely another dashboard bolted onto the same exports. It is governed operations intelligence: a layer where every number has a definition, an owner, a freshness stamp and a visible source.

Why the Spreadsheet Ops Stack Breaks

Spreadsheets are a reasonable first answer. They are fast, flexible and familiar. The trouble starts when they become the reporting infrastructure for a business with real clients, real margins and a team larger than the one person who built the workbook.

• The same metric is calculated three ways by three people, so meetings start with a debate about the number instead of the decision.
• Reporting means exporting data page by page, so the people with the most context spend their time on production instead of analysis.
• Views that matter span modules the source system never joins (for example jobs, finance and service in one picture), so the join lives in a fragile formula.
• Nobody can say how fresh a figure is, so decisions get made on stale data without anyone knowing.
• Client-ready reporting is a manual job, so it happens late, inconsistently or not at all.
• The person who built the workbook becomes a single point of failure.

None of these are exotic. They are the ordinary cost of running operations on exports. The cost is usually hidden because it is spread across salaries, delayed decisions and awkward client conversations rather than appearing as a line item.

Why Generic BI on Exports Rarely Fixes It

The common upgrade is a BI tool pointed at the same exports. Charts get nicer. The underlying problem stays. If the data underneath has no governed definitions, no lineage and no freshness, a better chart simply presents an ungoverned number with more confidence.

There is a second trap. Many operational systems were not designed to serve aggregate reporting at scale through their APIs. Teams end up calling the source repeatedly and reshaping raw responses into dashboards, which works until volume, permissions or a field change breaks it. Durable reporting needs a trusted copy of the data with meaning attached, not a live scrape dressed up as a dashboard.

What a Governed Ops Dashboard Actually Contains

Governance is an overused word, so here is the practical version. A governed operations layer has three properties. If a vendor cannot describe all three in plain English, you are buying a chart library.

1. Canonical data

Trusted reporting reads from governed copies of operational data, not raw calls to the source on every page load. That is what makes numbers stable, auditable and fast enough to use in a meeting.

2. Semantic meaning

Labels, definitions and allowed operations are governed separately from raw fields. "Active project", "billable hours" and "overdue" each mean one thing, defined once, and everyone sees the same answer.

3. Provenance

Every figure can say where it came from, how fresh it is and what state it is in. When a client or a board member challenges a number, the answer is a click, not a search through old workbooks.

Around those three properties sit the things teams actually touch: dashboards, widgets they can build themselves, permissions, and a shared artifact model so a report card is the same object whether it is viewed internally or shared with a client.

A Practical Replacement Sequence for NZ SMEs

You do not need a big-bang migration. The point of a governed layer is that the system of record stays the system of record. A sensible sequence looks like this.

1. List the five to ten numbers your leadership team actually uses to run operations. Ignore the other forty columns.
2. Write one plain-English definition and one named owner for each number.
3. Record where each number comes from today, how often it is refreshed, and who last changed the formula.
4. Pick one cross-module view the source system does not give you, and treat it as the first governed report.
5. Move that report onto a canonical layer with visible provenance, and retire the spreadsheet version only when the team trusts the replacement.
6. Add client-facing visibility last, on the same governed artifact, so you are not maintaining a second reporting product.

Steps one to three cost nothing and are worth doing even if you never buy anything. They are also the fastest way to find out whether your real problem is tooling or a lack of agreed definitions.

Rent, Build or Stay on Spreadsheets

Mid-sized NZ businesses usually face three options, and each is right in some situations.

• Stay on spreadsheets when reporting is small, the owner of the workbook is stable, and no client depends on the output.
• Build in-house when you have engineering capacity, a unique reporting need and the appetite to maintain a data layer indefinitely. Budget for ownership, not just the first version.
• Rent a productised layer when the problem is common across businesses running the same system of record, you want a faster path to governed reporting, and you would rather not carry the maintenance of a custom data platform.

Productised means the vendor has already solved the canonical, semantic and provenance layers for a specific operational system, and you configure it to your business instead of commissioning it from scratch. The trade-off is fit: a vertical product is excellent when it matches your stack and wrong when it does not. Be wary of anyone who claims otherwise.

Where Sightra Fits

Sightra is ShiftCurve's governed operational intelligence layer for project-based service businesses running Workbench. It sits above Workbench, reads and elevates the data, and does not replace your system of record. It is built around governed report-card definitions with canonical data, semantic meaning and visible provenance, so every number can say where it came from and how fresh it is.

It is not a generic BI tool, and it is not for teams looking to replace Workbench or shop for BI with no operational system of record. If you run Workbench and your reporting still means exporting CSVs, rebuilding the same pack each month and arguing about whose number is right, it is worth a conversation. Early access centres on Core, the governed ops picture. Portal, Connect, Field and Automate extend the same product later.

How This Fits with Foundations, Integration and Growth Systems

This essay is about the operations side of the business. It sits beside, not on top of, the rest of our thinking. Foundations Before Models is the sequencing rule for the lead system. Data Integration for NZ B2B Firms covers the website-to-CRM loop. Enterprise AI Integration NZ covers bounded joins when models touch enterprise systems. Governed ops intelligence answers a different question: once work is being delivered, can the people running it see one reliable picture?

The common thread is the same in every case. Name the owner, define the number, show the source, and review it on a weekly rhythm. If your growth side needs the same discipline, Compound applies it to website conversion, search, content and ads. If you need the digital foundation first, Services productises it.

Who This Is For

Owners and ops leaders at established NZ service businesses, particularly project-based firms, whose reporting is inconsistent, manual or dependent on one or two people. Account teams who rebuild the same client pack every month. Leaders who want visibility without another platform migration.

It is not for businesses whose reporting needs are genuinely small, or for teams who have not yet agreed what their key numbers mean. In that case, start with the definitions exercise above. Tools are the second step.

FAQ: Governed Operations Dashboards for NZ SMEs

What is a governed operations dashboard?

A reporting layer where metric definitions are agreed and owned, data comes from a trusted canonical copy, and every figure shows its source and freshness. It differs from a normal dashboard by making trust visible instead of assumed.

Is this just business intelligence under a new name?

No. Generic BI is a tool for charting data. Governed operations intelligence is workflow-specific: it carries definitions, lineage and permissions for a particular operational system. You can use BI tools on top of governed data, but BI alone does not supply the governance.

Do we have to replace our current system?

No. The system of record stays the system of record. A governed layer reads from it and makes the data legible. A migration is a separate decision and usually a much larger one.

When should we keep our spreadsheets?

When reporting is simple, stable and internal, and no client or board decision depends on it. Spreadsheets become a risk when the numbers carry commercial weight and only one person understands how they are built.