The spreadsheet is rarely the problem on its own. The problem is that it exists in four places at once — a laptop, a shared folder, an email attachment, and a version only one person can open — and each copy drifts a little further from the truth every week. By the time the operations meeting starts, the first fifteen minutes go on reconciling numbers instead of deciding anything.
What a stale spreadsheet actually costs
The visible cost is the copying. The expensive cost is the gaps. A spreadsheet shows you the totals it holds, not what is missing from its rows. If a job-tracking sheet has no line for a cancelled order, the cancellation does not exist in the report, and nobody notices until a customer chases it. Imagine a small team that closes each month by comparing a project sheet and a headcount sheet that were never joined on the same key: the utilisation figure it hands to management is an estimate wearing a suit.
There is a decision cost too. When the numbers are a week old, the meeting reviews the past instead of committing to the next three weeks. A live dashboard does not fix a broken process, but it moves the state of the business out of month-end paperwork and into the room where decisions get made.
What Power BI Pro does with those sources
Power BI Pro connects to Excel, SQL, Microsoft 365 and hundreds of other data sources, builds interactive reports and dashboards from what it finds, and embeds those reports in Teams and SharePoint. The source files stay where the team already keeps them; the dashboard reads them rather than asking someone to compile them.
Refresh is the part that decides whether the thing actually gets used. Power BI Pro refreshes data up to 8 times a day, which is comfortable for operational reporting — daily job status, open tickets, month-to-date billings — and not enough for anything that has to move minute by minute. A dashboard refreshed overnight is a working tool; one refreshed on request is a presentation.
Licensing is where small teams trip. In a standard Pro workspace, everyone who creates or views shared reports needs a Power BI Pro licence, including the manager who only ever opens the dashboard in a Teams tab. Scope the audience before you build, because the answer changes how much of the thing you build. Power BI Pro on sakal.shop is S$238.71 per user per year including 9% GST (S$219 before GST).
A worked example: one number, three sources
Imagine a small services company that wants one number visible to its managers: billable utilisation this month. Today it is assembled by hand from a project tracker in Excel, a timesheet export, and a client list held in Microsoft 365.
- Name the decision, not the chart. “Do we have capacity for another contract in the next three weeks?” is a decision. “Utilisation dashboard” is a wish. The decision tells you the granularity you need.
- Write down the sources and the join key. A client ID or a job number, usually. If the tracker spells one client three ways and the timesheet uses a code, fix it at the source — cleaning it inside the report only hides the mess from the next person.
- Build the smallest version. One joined table and two or three visuals answer the capacity question. Fourteen charts do not, and they will not survive their first contact with a manager who wants a different cut.
- Publish, embed, schedule. Publish the report, embed it in the Teams channel the managers already use, and set the refresh to match how often the underlying data actually changes.
Governance basics, and when to involve a managed IT partner
Three things carry most of the weight. Ownership: every shared dataset gets a named owner who is allowed to change its definitions. One definition: utilisation is defined once, in the report, and the manual version is retired — two dashboards with two definitions is a worse outcome than one spreadsheet. Access: grant it through Microsoft 365 groups rather than individual accounts, keep refresh credentials in a service account rather than one person’s mailbox, and keep the source files out of personal OneDrive.
Know the limits. If a number is needed by one person once a week, a spreadsheet is cheaper and better. If the source data is inconsistent, fix that first; a dashboard renders ambiguity clearly and quickly. And when the report has to reach an on-premises SQL server, a shared service account, or a security review, that is the point to bring in help. Credentials and access boundaries are ordinary work for a support team and a distraction for yours — managed IT support plans exist for exactly this kind of plumbing.
The first version does not need to be clever. Pick the one number your team argues about, trace it back to its sources, join it on a key that is already clean, and publish it with a refresh schedule attached. Then leave it alone for a month and see which questions it actually answers — the second version should be shaped by those, not by the features you have not tried yet.
Disclosure: Code n Rob is written by the founder of Sakal Network, which runs sakal.shop.

