Business intelligence in plain words
Every company already has the data it needs to run better. It sits in the ERP, the CRM, the invoicing system, project tools and a long list of Excel files. Business intelligence (BI) is the work of collecting that data, organising it and putting it in front of the people who decide — as a number, a table, a chart or a report.
BI answers questions like these:
- How are sales this month against target and against last year?
- Which customers owe us money, and for how long?
- Which products make money, and which only make revenue?
- Which items are about to run out, and which have not moved in six months?
If you have ever exported a report from the ERP to Excel, added a pivot table and sent it to the management team, you have done BI by hand. A BI system is meant to make that work faster, repeatable and less dependent on one person.
The classic BI stack
Most BI projects in larger companies are built from the same layers:
- Data sources. The ERP, CRM, accounting, e-commerce, spreadsheets.
- Data pipelines (ETL). Extract, transform, load: jobs that copy data out of each system every night or every hour, clean it and reshape it.
- A data warehouse. One central database built for analysis, where data from all systems sits side by side.
- A data model. Definitions everyone agrees on — what counts as a sale, how gross margin is calculated, which customers belong to which region.
- Dashboards and reports. Built in a BI tool such as Power BI, Tableau or Looker, and shared with managers.
- A BI team. Analysts and data engineers who build all of the above and keep it running when a system changes or a new question comes up.
Done well, this gives a company one trusted version of its numbers. It also takes time and people: a warehouse and a set of dashboards are a project, not a setting, and they need someone to own them after go-live.
Where classic BI struggles in small and mid-sized companies
For a company with 20 to 300 employees, the classic stack often stalls for familiar reasons:
- There is no BI team. The project depends on an outside consultant or on one person in finance who also has a day job.
- Dashboards answer yesterday's questions. They show what was planned when they were built. The question the CEO asks on Sunday morning usually is not on any of them.
- Managers are not at a desk. The sales manager is on the road, the operations manager is on the floor. They are in WhatsApp, not in a BI portal.
- Every change is a ticket. A new filter, a new breakdown, a new customer group — each one waits in line.
The result is common: the dashboards exist, few people open them, and the real questions still go to the person who knows how to pull the data.
The conversational approach to BI
The newer approach starts from the question instead of the dashboard. A manager asks in plain language, in the chat they already use, and gets the answer from the company's own data — as a number, a chart, a table or an Excel file.
This is how WizBI works. It builds a company brain: one layer over the company's systems, documents and spreadsheets that knows the business, its terms and its permissions. It connects to Priority, SAP, monday.com, SQL databases, Excel and Google Sheets and more, and AI workers on top answer the team in WhatsApp, Microsoft Teams, Slack, Telegram, e-mail or the web. One question can draw on several systems at once.
It also works when nobody asks: a sales summary every Sunday at 08:00, or an alert when a figure crosses a line you set.
| Classic BI | Conversational BI | |
|---|---|---|
| Starts from | A data model and dashboards | The questions people actually ask |
| Who builds it | A BI team or consultant | Connect one system; no BI team needed |
| New question | Change request to the BI team | Ask it in the chat |
| Where answers appear | A BI portal or app | WhatsApp, Teams, Slack, e-mail, web |
| Strongest at | Fixed, carefully designed dashboards | Ad-hoc questions, reports and alerts in the chat |
When classic BI is the right choice
Conversational BI does not replace every BI project. Classic BI is the better fit when:
- You have analysts who build and maintain data models as their main job.
- The board or a regulator needs the same fixed set of reports, in the same format, every month.
- You need rich interactive dashboards on a wall screen or in a monthly review meeting.
- Your data volumes and number of sources call for a dedicated warehouse anyway.
Many companies end up with both: a small set of fixed dashboards for the monthly review, and a chat for everything in between. For a side-by-side look at dashboards and asking in the chat, see Power BI alternative: dashboards vs asking in the chat.
How to start with BI in a small business
- Write down 10–20 real questions. Ask managers what they asked this month and how long the answer took. This list is worth more than any tool comparison.
- Find where each answer lives. Usually most questions come from one system — often the ERP — and the rest from a spreadsheet or the CRM.
- Start with one system. Connect the source that answers most of the list, and only that one.
- Decide who sees what. Finance, sales and management do not need the same view. Settle permissions before anyone asks the first question.
- Start with answers. Get the numbers right first; automate reports and processes once you see what people use.
- Check use after a few weeks. Are people asking? Which questions came back wrong or incomplete? Add the next source only after that.
With WizBI that first step is one system and your team's real questions — first answers within days. If you run Priority, start with AI for Priority ERP; if the numbers live in spreadsheets, see Excel.