What does a Power BI consultant actually do?
A Power BI consultant designs how your business data is collected, joined and calculated, then builds dashboards on top of that model and keeps them refreshing. The visuals are the last ten percent; the model underneath decides whether anyone trusts the numbers.
Most small businesses that call a Power BI consultant already have Power BI Desktop installed and a few charts somebody built. The problem is usually one of these: numbers that don't match Xero, a report that only one person knows how to refresh, measures copied between files with slightly different logic, or dashboards so crowded nobody reads them.
The real work splits into four jobs. First, connection: getting data out of Xero, Shopify, ServiceM8 and spreadsheets reliably. Second, modelling: arranging that data into fact and dimension tables, with a proper date table, so one measure works everywhere. Third, measures: writing DAX calculations such as gross margin, average job value or repeat-customer rate once, documented, so every page uses the same definition. Fourth, delivery: dashboard design, scheduled refresh, access control and handover. A good Power BI consultant spends more time on the first three than the fourth, and shows you why.
Do you need a Power BI consultant, or can you build it yourself?
Build it yourself if you have one clean data source, a person who enjoys learning DAX and a few hours a week. Bring in a Power BI consultant when data comes from several systems, numbers must reconcile with Xero, or the reports need to refresh without anyone touching them.
Power BI Desktop is free to download and there are good Microsoft Learn tutorials, so a curious owner can get a sales chart from a single Shopify export in an afternoon. The hard parts arrive later: joining Shopify orders to Xero invoices when customer names differ, handling refunds and credit notes, building a date table that runs July to June, and keeping refresh running when a password changes.
A practical decision rule. If your reporting question can be answered from one system's built-in reports, use those; Xero and Shopify both have decent native reporting. If you need to combine two or three systems, a lightweight model in Looker Studio or Power BI makes sense, and a consultant saves you weeks of trial and error. If several people need different views of the same data with access controls, that is squarely Power BI consultant territory.
- One source, one person, simple charts: DIY or native reports
- Two or three sources that must reconcile: consultant-built model
- Several roles needing different views and access: Power BI with row-level security
- External viewers without licences: a custom reporting portal
Getting Xero data into Power BI for Australian reporting
Xero data reaches Power BI either through the Xero API, pulled on a schedule into a small database or data file, or through regular report exports. The API route is more reliable for daily dashboards; exports can be enough for a monthly board pack.
Xero holds the numbers your accountant signs off on, so it is usually the anchor for finance dashboards: invoices, bills, payments, contacts, tracking categories and the chart of accounts. Pulling this data through the API means we write a small scheduled job that fetches new and changed records and writes them to storage Power BI can refresh from. That job lives in your cloud account, not ours.
A few details matter for Australian businesses. Tracking categories often hold location or department, which becomes the main slicer on every page. GST-inclusive and GST-exclusive amounts must be handled consistently, or margins drift. Credit notes need to reduce revenue in the right period. And the profit and loss in Power BI should match Xero's to the dollar for any closed month; we build a reconciliation page that shows exactly that, so your bookkeeper can check it. If your Xero integration itself needs work, such as syncing jobs or orders into Xero, our Xero integration page covers it.
Shopify, ServiceM8 and other operational systems
Operational systems explain why the financial numbers look the way they do. Shopify shows orders, products, discounts and customers; ServiceM8 shows jobs, quotes, staff time and materials. A Power BI consultant links them to Xero so you can see margin by product, by job type or by technician.
For Shopify, we pull orders, line items, refunds, products and customers through the Shopify Admin API on a schedule. The model separates gross sales, discounts, returns and shipping so the numbers line up with Shopify's own analytics and with Xero after fees. Repeat-purchase rate, average order value by channel and sell-through by product are common pages.
ServiceM8 offers a REST API that returns JSON, according to its developer documentation, covering jobs, job activities such as scheduled bookings and recorded time, clients (called companies in the API) and materials. That lets a trades or field-service business see quoted versus actual hours, jobs completed per technician, callback rates and revenue per job type, joined to Xero invoices for the real margin. Other tools such as Tradify, Simpro or booking systems can usually be connected the same way if they provide an API or scheduled export; we confirm this during scoping rather than assuming.
What happens to the spreadsheets you already rely on?
They stay, but they stop being the reporting engine. Useful spreadsheets, such as targets, budgets or a price list, become tidy input tables that Power BI reads; the copy-paste monthly workbook gets retired.
Nearly every small business runs on a few critical spreadsheets: sales targets by month, a budget from the accountant, a staff roster, a list of product costs that Shopify does not know. These hold information no system has, so a Power BI consultant should not throw them away. Instead we move each one into Google Sheets, SharePoint or OneDrive, fix the layout so every column has one type of data and every row is one record, and connect it so changes show up at the next refresh.
The monthly workbook that someone spends a day rebuilding is different. That is the job the dashboard replaces. Before retiring it, we document every formula and check the new model produces the same figures for at least one past month, so nobody loses confidence in the change. You keep the old workbook archived in case a question comes up later.
Data modelling: why a Power BI consultant starts with a star schema
A star schema puts the things you measure (sales lines, job records, costs) in fact tables and the things you slice by (date, customer, product, staff, location) in dimension tables around them. It makes measures simpler, reports faster and numbers consistent.
The alternative, one wide flat table per report, works for a single chart and falls apart by the third dashboard. Joins get duplicated, a customer renamed in Xero appears twice, and each new page needs new workarounds. With a star schema, the customer dimension is built once from Xero and Shopify contacts, matched on email or ABN where possible, and every fact table points to it.
Measures are written in DAX against this model and stored in one place with plain-English descriptions. "Gross margin %" means one thing everywhere. Time intelligence, such as same period last financial year or year to date from 1 July, uses the shared date table. Row-level security, where a regional manager only sees their region, is applied to dimensions and flows through automatically. A Power BI consultant who skips modelling to get to visuals faster is borrowing time from your future.
Setting up July–June financial-year and BAS-period reporting
Australian dashboards need a date table where the financial year runs from 1 July to 30 June, quarters line up with the quarterly BAS periods (July–September, October–December, January–March and April–June), and labels read "FY26" rather than a calendar year.
Power BI's default date handling assumes a calendar year, so year-to-date measures reset in January unless told otherwise. We build a dedicated date table with columns for financial year, financial quarter, financial month number (July as month 1), BAS quarter label, week-starting-Monday, and flags for national and relevant state public holidays if your trading depends on them. Measures such as FYTD revenue, same period last FY and rolling twelve months then work correctly across every page.
For BAS-period views, the dashboard can show GST collected and GST paid by quarter as drawn from Xero, giving your bookkeeper a quick cross-check before lodgement. To be clear, this is a reporting convenience, not a tax calculation: Xero and your accountant remain the source for what you lodge with the ATO. Businesses on monthly GST reporting get monthly views instead. Seasonal businesses, such as tourism in the north or retail around Christmas and end-of-financial-year sales, often want comparisons by financial-year week, which the same table supports.
How scheduled refresh works, and what limits apply
Scheduled refresh reloads your data model from its sources at set times each day. Microsoft's documentation says semantic models on shared capacity, which covers Power BI Pro workspaces, are limited to eight scheduled refreshes a day, while Premium, Premium Per User and Fabric capacities allow up to 48.
For most small businesses, eight refreshes is plenty: early morning before staff arrive, late morning, mid-afternoon and end of day covers most needs. Microsoft notes that refresh times follow the time zone chosen in the model settings, so we set yours to the right Australian zone rather than leaving the default. Manual "Refresh now" clicks in the service are not counted against the eight.
Where a data source is something Power BI cannot reach directly over the internet, such as a file on an office PC or a local database, Microsoft's documentation says you need a data gateway before a refresh can run. We usually avoid that dependency by moving spreadsheets to OneDrive, SharePoint or Google Sheets and landing API data in cloud storage, so nothing breaks because an office computer was switched off. Refresh failures send an email alert; under a care plan, that alert comes to us as well and we fix the cause.
Microsoft's article on data refresh in Power BI documents the limits and gateway rules.
Which KPIs should a Power BI dashboard show? Sets by industry
Start with the five to eight numbers the owner already asks about every week, then add drill-downs. A dashboard with forty tiles gets opened once; one with six trusted numbers gets opened every morning.
KPI sets differ by business model, so we start from a shortlist for your industry and cut it down with you. A trades business cares about jobs completed, quoted versus actual hours, first-visit completion and revenue per technician. An online retailer watches revenue, average order value, repeat-customer rate, returns and stock cover. Hospitality groups track sales per labour hour, food and beverage cost percentages and covers by session. Professional services firms look at billable utilisation, work in progress, debtor days and revenue per fee earner. NDIS providers often need hours delivered against plans, claim status and staff utilisation, handled with extra care because the data is sensitive.
Each KPI gets a written definition, a target where one exists, and a clear source. If a number cannot be sourced reliably yet, it is left off rather than estimated. The table further down this page lists starter sets you can take into your first scoping call.
How much do Power BI licences cost in Australia?
Microsoft's Australian pricing page listed Power BI Pro at 21 Australian dollars per user per month and Premium Per User at 35.90 Australian dollars per user per month, both paid yearly and excluding GST, when we checked in September 2026. A free account is also available. Check Microsoft's page before budgeting, as prices change.
The licence question that matters is who needs one. In a typical SMB setup, anyone publishing reports and anyone viewing shared reports in the Power BI service on shared capacity needs a Pro licence. A business with five managers viewing dashboards is therefore looking at five Pro licences. Premium Per User adds larger model sizes, more frequent refresh and some advanced features, which most small businesses do not need at the start.
If your business already pays for a Microsoft 365 plan, check whether Power BI Pro is included in your subscription before buying separately; your Microsoft 365 administrator or reseller can confirm. Licences are always bought by you, in your own tenant, and paid to Microsoft or your reseller. We never resell licences, so there is no incentive for us to recommend more seats than you need. When licence costs outweigh the benefit, we will say so and suggest Looker Studio.
Current figures are on Microsoft's Australian Power BI pricing page.
Power BI or Looker Studio: which should a small business choose?
Choose Looker Studio when your data already lives in Google Sheets or Google tools, the model is simple and you want to share dashboards widely at no licence cost. Choose Power BI when you need a proper data model, complex calculations, row-level security or you already work in Microsoft 365.
Google's documentation describes its reporting tool, known to most people as Looker Studio and labelled Data Studio in recent Google documentation, as a no-cost tool for building dashboards and reports. It connects naturally to Google Sheets, Google Analytics, Google Ads and BigQuery, and dashboards can be shared by link. For a café owner who wants weekly sales from a Google Sheet next to website traffic, it is often the right answer.
Its limits show as needs grow. Joining several sources is clumsier, heavier calculations slow down, and row-level permissions take more work. Power BI's modelling engine, DAX measures and security features are stronger, which matters once finance data from Xero has to reconcile exactly. Some businesses start in Looker Studio and move later; if we design the underlying Google Sheets or data store well, that move is straightforward. As a Power BI consultant we are happy to recommend the cheaper tool when it fits.
How much does a Power BI consultant cost in Australia?
Quotes vary widely, and the difference mostly comes from scope: number of data sources, spreadsheet clean-up, how many dashboard pages and whether automated data pipelines are needed. With BtechWaleTech, automated reporting starts from US$600 and a custom reporting portal from US$900.
A single-source dashboard from a clean Shopify or Xero export is at the small end. A model combining Xero, Shopify and two spreadsheets, with a financial-year calendar, four or five dashboard pages and scheduled refresh through a small cloud data pipeline, sits in the middle. Multi-entity businesses with intercompany eliminations, several locations and role-based security are larger projects.
Be cautious about hourly engagements with no defined outcome, where the model grows as a pile of fixes. Ask for an itemised quote that names the data sources, the pages, the measures and the refresh setup. Ours does, and nothing is billed until you approve it in writing. Ongoing care from US$120/mo covers monitoring and small changes; you can also run everything yourselves after handover.
Keeping business and customer data safe in Power BI
Keep data in your own Microsoft tenant and cloud accounts, give each person the least access they need, and bring in only the fields a dashboard actually uses. Customer names, phone numbers and health details rarely need to be in a sales dashboard.
A Power BI consultant should design with data minimisation in mind. For most KPI work, aggregated figures or customer IDs are enough, and personal details can stay in the source system. Where personal information is needed, such as NDIS participant hours or clinic appointments, we limit which pages show it, apply row-level security so staff see only their own clients or region, and keep workspace membership tight.
The OAIC explains that the Privacy Act 1988 covers organisations with annual turnover above AUD 3 million and some smaller businesses, including health service providers. If that includes you, how reporting handles personal information is part of your wider privacy obligations; that assessment belongs with your own adviser, not with us. What we deliver technically: data held in your tenant, API credentials stored securely, access reviewed at handover, and a written list of where each dataset comes from and who can see it.
From scattered data to first dashboard: timeline and process
A typical SMB project reaches a first working dashboard page within the first two weeks and finishes in two to four weeks, depending on how many sources and how much spreadsheet clean-up is involved.
Week one is discovery and connection. We agree the KPI list with you, get read access to Xero, Shopify, ServiceM8 or whatever systems apply, and pull a first extract. Data problems surface immediately: duplicated customers, inconsistent product codes, missing tracking categories. We list them and agree which to fix in the source and which to handle in the model.
Week two builds the model and the date table, writes the core measures and delivers the owner's summary page for you to check against numbers you already know. Weeks three and four add the remaining pages, row-level security, scheduled refresh, and the reconciliation page against Xero. Handover includes a recorded walkthrough and a short document explaining each measure. Review calls happen in your afternoon, with screen sharing so you can click through the report together with us.
Who owns the dashboards, data model and workspace?
You do. Reports and semantic models are published to a workspace in your Microsoft tenant, the .pbix files and any pipeline code are handed over, and all API connections are authorised under your accounts.
During the project we work as guest users or with accounts you create for us, so we can publish and test in your workspace. Nothing is built in a tenant we control and then "transferred". API credentials for Xero, Shopify and ServiceM8 are created under your accounts, and the small scheduled jobs that fetch data run in cloud accounts registered to you, with costs billed to you directly by the provider.
At handover you receive the Power BI files, the pipeline code in a repository you own, a data dictionary listing every table and measure, and instructions for changing refresh times or adding a user. If you later hire an in-house analyst or another Power BI consultant, they start from documented work rather than a mystery file. When the project ends, you remove our access; there is nothing left for us to release.
Working with a remote Power BI consultant in India from Australia
Work runs through screen-sharing calls in your afternoon, a shared KPI document and WhatsApp for quick questions. Quotes and invoices come from India in USD, paid by Wise or bank wire, and nothing is billed before your written approval.
India Standard Time sits four and a half hours behind AEST and five and a half behind AEDT, so a 1:30 pm review in Sydney during standard time is 9 am in India. Brisbane has no daylight saving, which keeps the overlap steady all year, and Perth, two and a half hours ahead of India, shares most of our working day. Data work suits this rhythm well: you send questions and sample reports during your day, and fixes are usually ready by your next morning.
The first two weeks look like this. Day one: a scoping call to agree the KPIs and list the systems. Days two and three: you grant access; we pull first extracts and send a data-quality note. By the end of week one, the model structure is agreed. In week two you receive the first dashboard page, check it against figures you trust, and tell us what looks wrong. We do not visit offices or install anything on your computers; everything runs in your cloud accounts.
Example: a Perth electrical contractor combining ServiceM8 and Xero
A hypothetical scenario. Say a Perth electrical contracting business with eight technicians runs jobs in ServiceM8, invoices through Xero and tracks targets in a Google Sheet. The owner wants to know which job types make money and which technicians need support.
The build pulls ServiceM8 jobs, time entries and materials through its API, Xero invoices and bills through the Xero API, and targets from the Google Sheet. The model joins jobs to invoices, so each job shows quoted hours, actual hours, materials cost and invoiced revenue. A financial-year date table from 1 July lets the owner compare this FY to last FY by month.
Four pages come out of it: an owner summary (revenue, gross margin, jobs completed, debtor days), a job-type page (margin for switchboard upgrades versus maintenance callouts versus new builds), a technician page (hours, first-visit completion, callbacks) and a finance reconciliation page matching Xero's profit and loss for closed months. Refresh runs four times a day in Perth time. With a handful of managers viewing, Power BI Pro licences are bought in the business's own tenant. If the owner only needed the summary page, a Looker Studio version from the same Google Sheet and exports would also be on the table. This is an illustration, not a client story.