Why modern BI hasn't killed the spreadsheet
BI vendors have predicted the end of the spreadsheet for twenty years, and Excel is still everywhere. What spreadsheets genuinely do better, what BI tools do better, and a realistic way for a small team to use both.
BI vendors have been announcing the death of the spreadsheet for about twenty years. It has not happened. Excel and Google Sheets are still where most business numbers are first typed, checked and argued over. That is not because people are stubborn. Spreadsheets are very good at some jobs that BI tools are not designed for, and a BI rollout that ignores this ends up as a set of dashboards nobody opens.
This post looks at what each tool is genuinely better at, and how a small team can split the work sensibly.
What spreadsheets still do better
1. Throwaway analysis
Add a column, sort it, glance at the result, delete the file. A spreadsheet is a scratch pad, and the cost of starting is zero. In most BI tools the unit of work is a saved dashboard or query, which carries more weight than "open a blank sheet and paste". For a question you will ask once, the scratch pad wins.
2. Fixing one bad row
The export from your accounting tool has one duplicated invoice. In a spreadsheet you delete the row and carry on. A BI tool treats the source as authoritative, which is right in principle (the fix belongs in the source system) but slow when you need a number by five o'clock.
3. What-if modelling
You are considering a price change. You build a small table, change the price, and watch revenue, margin and break-even volume update. That instant loop between an input and everything downstream of it is what spreadsheets were invented for. BI tools are built to describe what happened; most are clumsy at "what if we changed this".
4. Complete control of layout and logic
A spreadsheet does exactly what its author told it to, cell by cell. For a board pack with specific formatting, a commission calculation with odd rules or a one-page budget, that control matters more than automation.
What BI tools do better
1. Data that refreshes from the source
A dashboard connected to Stripe or Google Analytics refreshes from the source on a schedule. A spreadsheet shows the data as of the last time someone pasted an export into it, and nothing on the page says when that was.
2. Several sources at once
Revenue from billing, spend from two ad platforms and deals from the CRM do not live in one place. In a spreadsheet you export four files and join them by hand every week. A BI tool connects to each once and lines them up by date or a shared column. Our post on multi-source dashboards covers how that works.
3. One version of the number
Two people looking at the same dashboard see the same figure. Two people with copies of the same spreadsheet may not, because one sorted, one filtered and one pasted last week's data. Copies are where competing numbers come from.
4. Access control and sharing
Giving someone view-only access to a dashboard is safer than emailing a workbook with every tab, hidden sheet and formula inside it. You can also take access away later, which you cannot do with an attachment.
5. Repetition
Anything you rebuild every week, such as the Monday revenue summary or the monthly marketing report, is better done once in a BI tool and refreshed or scheduled from then on.
The realistic split
For most small teams the answer is both, divided by the kind of work:
| Work | Better in |
|---|---|
| Recurring numbers people check weekly | BI dashboards |
| Anything combining two or more systems | BI |
| Reports that go to the same people on a schedule | BI |
| One-off checks and quick sorting | Spreadsheet |
| Pricing, budget and scenario models | Spreadsheet |
| Data that exists nowhere else (targets, manual logs) | Spreadsheet, connected to BI |
The last row matters most. Plenty of useful data only exists in a sheet: sales targets, a staffing plan, a list of promotions and their dates. Do not force it into another system. Keep it in the sheet and connect the sheet to your BI tool so it can sit next to everything else.
The mistake is picking a side on principle. Moving everything into dashboards makes the quick jobs slow. Keeping everything in spreadsheets makes the recurring jobs error-prone.
How clariBI works with spreadsheets
clariBI is built on the view that the spreadsheet stays. You can:
- Upload CSV and Excel files as data sources. Each one gets an automatic dashboard, and you can ask questions about it in plain English. See the CSV and Excel page.
- Connect Google Sheets, so a sheet your team keeps up to date becomes a source that syncs on a schedule and on demand. The Google Sheets and Drive article explains the setup.
- Combine a sheet with app data, for example monthly targets from a sheet next to revenue from Stripe, and ask "How is revenue tracking against target this quarter?"
- Take data back out: dashboard data exports as CSV or JSON, and reports download as PDF. The export formats article lists what each contains.
What clariBI does not try to be is a spreadsheet. It has no cell-level editing or what-if modelling grid; keep those in Excel or Sheets.
If your team is further along and wants to move recurring reporting out of spreadsheets, the guide to migrating from Excel to a BI tool covers the order to do it in.
A first step
Pick the one report you rebuild in a spreadsheet every week by pasting in fresh exports. Connect the sources it is built from, or the sheet itself, and ask the question the report answers. If the answer matches your spreadsheet, you have one less paste job each week, and the spreadsheet is still there for everything else.