Almost every business now holds digitised data — in the UK’s 2026 business data survey it was 86% — but only a quarter actually analyse it to make decisions, and for small firms it’s about one in three. The gap between “we have data” and “we use data” is where most of the value sits. You don’t need more data or a data warehouse: connect the systems you already run (Xero, your CRM, your job or stock system), clean the few fields that matter, and put them in a live dashboard built around one decision you make often.
Ask most business owners if they’re “data-driven” and you’ll get a rueful smile. The truth is nearly every New Zealand business generates useful data every single day — sales, jobs, stock, customer interactions, finances. The problem isn’t a lack of data. It’s that the data sits unused in spreadsheets, accounting software and inboxes, never brought together into anything you can actually act on.
The numbers bear this out. The UK Government’s Business Data Survey 2026 — the closest thing to a full national measurement, and a fair proxy for New Zealand’s similarly SME-heavy economy — found 86% of businesses handle digitised data, but only 25% analyse it to generate new insight. For small businesses the figure is 35%; for micro businesses, 26%. Locally, Datacom’s 2025 State of AI Index found data quality and integration was the second-biggest barrier (22%) NZ organisations faced in getting value from AI — behind only skills. And the stakes are real: Xero’s first NZ Small Business Productivity Report (June 2026) puts small-business output at just $74 an hour worked, consistently behind Australia and the UK. Better decisions, made faster, are one of the few levers on that number an owner controls directly.
You don’t need more data — you need it connected
The instinct is to start collecting more. Usually the bigger win is connecting what you already have. A typical NZ business we work with runs Xero or MYOB for money, a CRM or an inbox for customers, a job-management, POS or stock system for operations, and a payroll tool for people — four or five systems that each hold a slice of the truth and never talk to each other. When those feed one consistent source, patterns appear that were invisible when the numbers lived in separate places: which job types actually make margin, which customers pay late, which products are quietly dead stock. Pulling those together is the foundation of any data and business intelligence project, and it is usually the first, most valuable step in any AI project too — in the UK survey, businesses using AI were far more likely to be collecting and analysing data properly (41% and 37%) than those that weren’t.
Where the data already is
- Accounting (Xero, MYOB): revenue by customer and product, margin by job, debtor days, cash-flow timing. Most owners look at the P&L once a month; the same data can tell you daily which jobs are drifting over budget.
- CRM and inbox: enquiry volume, response time, win rate, where leads come from. If enquiries are answered in a shared inbox, that data exists — it’s just never counted.
- Operations (job management, POS, stock, dispatch): throughput, utilisation, lead times, stock turns, rework. This is the data that explains why the accounting numbers look the way they do.
- Documents and email: quotes, invoices, forms, supplier correspondence. Historically locked in PDFs — now readable with AI document processing, which turns a folder of invoices into a table you can analyse.
Three questions good data answers
- What’s actually happening? Live dashboards replace month-old spreadsheets, so you see the business as it is now — not as it was when someone last exported the report.
- What’s about to happen? Trends and simple forecasting help you plan stock, staffing and cash instead of reacting. Even a 13-week cash forecast built from Xero and your order book removes most of the surprises.
- Where’s the money leaking? Bringing data together surfaces the jobs, products or processes quietly costing you — the customer segment with 40% of your invoices and 5% of your margin, the product line that ties up stock for six months.
From spreadsheets to a live view
The practical destination for most businesses is a clear, live dashboard — the numbers that matter, updated automatically, accessible to the people who need them. That’s what our BI dashboards solution delivers: one source of truth, no more conflicting reports, and the hours currently spent every week assembling the Monday report given back. The mechanics are straightforward: integrate the source systems through their APIs, clean and reconcile the handful of fields that matter (customer names, job codes, GST treatment), model the metrics once, and present them in Power BI, Looker Studio or a custom web dashboard your team opens on a phone.
Increasingly the dashboard is only half of it. Once the data is connected and clean, an AI agent can sit on top of it — answering “which jobs are over budget this week?” in plain English, flagging anomalies before month-end, or drafting the follow-up to the customers who are 30 days overdue.
The goal isn’t a data project for its own sake. It’s making the next decision with evidence instead of a gut feeling.
A real example
A nationwide trades business we worked with — 60 staff, multiple regions — ran jobs from spreadsheets and WhatsApp groups. The data existed: every job, every variation, every invoice was recorded somewhere. None of it was connected, so managers spent a large part of each week chasing status, and variations went un-invoiced. Once jobs, crews and Xero were brought into one live system, each manager got about 12 hours a week back and revenue rose 23% — much of it from work that had always been done but never billed. No new data was collected; the existing data was connected. Read the case study.
Start with one decision
You don’t need a data warehouse to begin. Pick one decision you make often and wish you made better — pricing a job type, ordering stock, chasing debtors, rostering — and build the smallest dashboard that informs it. From there it compounds. We scope this fixed-price, using the data you already have, so the first step is low-risk and quick to value. If you’re not sure where the value is, a short consulting engagement to map your systems and the decisions that depend on them is the cheapest way to find out.
Sources
- UK Department for Science, Innovation and Technology, UK Business Data Survey 2026, 18 June 2026 — share of businesses handling and analysing digitised data, by size; data practices of AI users
- Xero, NZ Small Business Productivity Report, 18 June 2026 — output per hour worked, international comparison
- Datacom, 2025 State of AI Index — reported by IT Brief NZ — data quality and integration as a barrier for NZ organisations
The UK survey is used as a proxy where no equivalent NZ measurement exists; Stats NZ’s Survey of Business Operations 2026 will provide the first official NZ figures later this year. Last updated 15 September 2026. Written by Ashok Poshamalla, Founder, Avtrix Software Solutions, Taupō.