BC2Fab - AI Insights

Sales Report

A ready-made sales report on your sales semantic model, opened with one click from the role center. One page shows how your sales are doing: nine KPIs against the prior-year period, a calculation scheme from orders to result, cumulative sales with an outlook, a free breakdown under every KPI, AI interpretation on request, and Ask your data anchored to any figure. Every number is computed live by the Power BI engine under the signed-in user's own identity — nothing is copied, nothing is recalculated in the browser. The report is part of the Sales Analysis plan; switching it on takes three steps described on Sales Report setup.

What the report shows

The page reads from top to bottom like a management report. Everything below the slicer bar follows the period, comparison and filters set there.

KPI tiles

Nine sales KPIs, each with its value, the change against the prior-year period and the prior-year value.

Net Salesperiod, vs. prior year
Order Incomeperiod, vs. prior year
Profit Marginperiod, vs. prior year
Margin %ratio, change in percentage points
New Customer's Net Salesperiod, vs. prior year
Active Customersperiod, vs. prior year
Discount %ratio, lower is better
Order Backlogkey-date value, no date window
Expected Net Salesoutlook, no date window

Order backlog and expected net sales are stock figures: they show the state as of the key date and carry no prior-year comparison. All other KPIs are period figures with the prior-year period alongside.

At a glance (AI)

On request, the AI writes what moved and why — for the report as a whole, from the report's own figures. See AI cards.

KPIs at a glance — the calculation scheme

The full KPI schema, current period against the prior-year period, switchable between Consolidated and Per company.

1
Orders — order income, order backlog, expected net sales.
2
From sales to result — gross sales, less sales discount and credit amount, gives net sales (split into new and existing customers), less direct costs, gives profit margin.
3
Further KPIs — discount %, credit ratio %, margin, active customers, average net sales per customer.

Net Sales cumulative (YTD) — actual & expected

Month by month: the cumulative actual net sales up to the cut-off month, then the expected net sales from the order backlog. The prior year runs alongside for comparison.

KPIs in detail

One section per KPI, in the order of the tiles. Each section has two parts.

1
Overview — the key figures of the KPI, a trend chart over the months of the period against the prior year (a projection chart for the key-date KPIs), and the AI interpretation of this KPI on request.
2
Free analysis — the KPI broken down by a dimension of your choice, by company by default. Pick the grouping from the model's dimension catalog (categories, customers, salespeople, dimensions), add up to four levels, expand a row to drill into the next level, and load more rows in steps of 500. Columns: current period, prior year, change in %.

Summary & assessment (AI)

The closing card: an AI-written summary and assessment of the whole report, on request, marked as AI generated.

Period, comparison and filters

  • Report year and cut-off month. The default is year to date up to the last closed month; alternatives are the current year, the previous year, or any year with a cut-off month of your choice. A cut-off after the last closed month is marked provisional.
  • Comparison. Always the same window one year earlier — the prior-year period. Ratios compare in percentage points, everything else in absolute terms and in %.
  • Company. Pick one or more companies; the default is all companies. The per-company view of the calculation scheme and the default breakdown of every free analysis follow the same list.
  • Filter line. Under the slicer bar the report states its period, the comparison, the selected companies, the base filter of the sales model (prepayment documents excluded) and Data as of — the latest posting date in the model. What you see is what every figure on the page was computed with.
  • Remembered per user. Period, companies and filters are stored for you and the model, and are there again next time. Reset returns to the defaults.

From the report onward

Advanced analysis

Hands a KPI section over to the analysis mode as a real analysis tab — same model, same filters, same breakdown — and you keep drilling with the full pivot grid. Each KPI keeps its own tab per user; re-opening updates that tab instead of creating another.

Excel (live)

Downloads a workbook whose table holds the query of the section with a live connection to the semantic model. Refreshing in Excel re-runs the query under your identity.

Print

Opens the browser's print dialog; choose Save as PDF. The print shows the report as it is on screen — expanded levels, sorting, loaded rows, the active view of the calculation scheme and any AI cards — with a header carrying period, comparison, companies, filters, data as of and the print time on every page.

Ask your data

A ✦ on every tile, row, card and chart starts a conversation about exactly that figure. Answers refer to the data of this report, come from live DAX queries, and show every query that ran. Everything on the Ask your data page applies here.

AI cards

  • On request, never unasked. Add AI interpretation writes the card for the report, for a KPI or the closing summary. Cards are marked AI generated.
  • Written from the report's own figures. The AI receives the KPI values with their prior-year comparison, the calculation scheme, and per KPI the largest increases and decreases by company, customer, product hierarchy and salesperson — aggregated figures, no detail records.
  • Shared where the numbers are the same. A stored card is shown to every reader whose data produce the same figures — users with a different row-level-security view never see a card written from figures they cannot see. An administrator can switch to AI Texts per User on the setup page instead.
  • Deletable. Delete AI Texts on the models page removes the stored cards of a model; the next request writes them anew.

Your data

The AI features run against an Azure OpenAI resource in your own subscription, called server-side from Business Central. No navida service is involved. Everything else on the page — tiles, scheme, charts, breakdowns — is plain DAX against your semantic model and involves no AI at all.

Sent to your Azure OpenAI resource

  • The structure of the semantic model: table, field and measure names with their descriptions
  • For AI cards: the aggregated report figures — KPI values, prior year, deltas, the per-company scheme and the largest movers per driver dimension
  • For the chat: your questions, the anchored figure, and the results of the queries it runs

The boundaries that hold

  • Every query runs as the signed-in user — row-level security limits what the report and the AI can ever read
  • Data travels only between Business Central, your Power BI tenant and your own Azure resource
  • Read-only: the report queries the model and never writes anything back

Every number from the engine

Prior-year values, changes, year-to-date states and the outlook are either the comparison measures your sales model already carries or measures defined inside the query itself — never arithmetic in the browser. Ratios, distinct counts and semi-additive figures such as active customers are therefore correct at every level, in every breakdown, exactly as in Power BI. Architecture explains the mechanism.

Set-up in short

Three things on top of the base installation: a sales semantic model built on navida's sales template, the Report Code SALES on that model's row, and the Sales Analysis licence for each report user. The step-by-step guide, the list of what the model must contain and the troubleshooting table are on Sales Report setup.

Contact us

Need support? Use this contact link to send us an email.