The verdict in three sentences
A 32-store retail group in Toronto whose CEO receives the monthly report 10 days after close is steering a month behind. The fix is a data warehouse fed nightly from 5 sources (POS, ERP, payroll, bank, e-commerce) and an executive dashboard, for example in Power BI Pro at about 14 USD per user per month. The custom part (connectors, data model, calculation rules) sits within the Kolonell application grid, from 3,810 to 22,870+ EUR depending on complexity, with the rest set in the quote.
Why the report takes 10 days
Two financial analysts export store sales from the POS every month, purchases and stock from the ERP, labour cost from payroll, bank balances, and web orders from the online store. They then reconcile store codes that differ between systems and fix returns, credit notes and buy-online-pick-up-in-store orders. The table describes the 5 sources of a group this size (2026 order of magnitude).
| Source | Useful data | Typical volume | Extraction method | Difficulty |
|---|---|---|---|---|
| POS (32 stores) | Transactions, baskets, discounts, shrink | 18,000 transactions per day | API or nightly export | Medium |
| ERP | Purchases, stock, margins, suppliers | 25,000 SKUs | Database or API | High |
| Payroll | Labour cost, hours, absences per store | 480 employees | Secure monthly export | Medium |
| Bank | Balances, card settlements, fees | 6 accounts | Standard statements or aggregator | Low |
| E-commerce | Orders, returns, in-store pickups | 1,200 orders per day | API | Low |
The real work is not drawing charts. It is the store mapping table, the rule that assigns an online sale picked up in Mississauga, and a shared definition of margin. Once those rules are written, the warehouse applies them every night without human intervention.
KPIs that deserve a place on the CEO's screen
An executive dashboard showing 60 indicators will not be read. Leadership teams usually keep 10 to 15 KPIs, broken down by region, store and department.
| KPI | Calculation | Frequency | Indicative alert threshold |
|---|---|---|---|
| Sales vs last year | Like-for-like sales | Daily | Below -3% over 7 days |
| Gross margin | (Net sales - cost of goods) / net sales | Daily | Drop of more than 1 point |
| Average basket and traffic | Sales / transactions, transaction count | Daily | -5% in one store |
| Known shrink | Damage, reported theft, expiry / sales | Weekly | Above 1.2% |
| Labour cost to sales | Loaded payroll / net sales | Monthly | Above 14% |
| Inventory turnover | Cost of sales / average stock | Weekly | Below 6 per year |
| E-commerce and pickup share | Web sales / total sales | Daily | Gap above 20% vs budget |
| Available cash | Consolidated bank balances | Daily | Below the CFO's floor |
Each KPI carries a written definition, an owner and a source. That is what ends arguments about numbers in leadership meetings.
Cost: licences, warehouse and custom build
Costs split between visualisation licences, warehouse hosting and connector development. Market prices below are 2026 orders of magnitude.
| Item | Option | Indicative cost |
|---|---|---|
| Visualisation | Power BI Pro, 25 users | 14 USD x 25 = 350 USD per month |
| Data warehouse | Managed cloud database in Canada | 100 to 500 USD per month depending on volume (market) |
| Traditional BI integrator | Consultant day rate | 900 to 1,500 USD per day (market) |
| Kolonell custom, simple | 2 to 3 sources, standard KPIs | 3,810 to 6,100 EUR |
| Kolonell custom, medium | 5 sources, reconciliation rules | 6,100 to 10,670 EUR |
| Kolonell custom, complex | 5+ sources, forecasting, alerts, store portal | 10,670 to 22,870+ EUR |
A 32-store group with 5 sources usually sits in the medium or complex band. Maintenance runs at 53, 114 or 183 EUR per month depending on the service level; warehouse hosting is detailed in the quote.
Mini case study
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Michael, CEO of a 32-store group in Toronto, reports 130,000,000 USD of annual sales. His two analysts each spend 6 days a month on the report: 12 days at 550 USD loaded, or 6,600 USD a month and 79,200 USD a year, redirected to analysis.
The decisive effect is elsewhere: with daily numbers, a margin or shrink drift is corrected in one week instead of six. Gaining just 0.2 margin points on 130,000,000 USD is worth 260,000 USD a year. Licences cost about 4,200 USD a year and the warehouse about 3,600 USD. Even at the top of the complex band, 22,870 EUR, the project pays back in a little over a month of margin gains.
FAQ
Is Power BI mandatory?
No, the warehouse is independent of the visualisation tool. Looker Studio, Metabase or a custom web dashboard can connect to it; Power BI is common because a Pro licence costs about 14 USD per user.
How long until the first version?
Allow 8 to 12 weeks for 5 sources and around ten KPIs. Sales and margin come first, within 4 weeks, then payroll and bank.
Can store managers see other stores' numbers?
That is a governance choice. Row-level security lets each of the 32 managers see only their store while leadership sees the consolidated view.
Is payroll data protected?
Yes, only per-store aggregates are loaded, never individual salaries. Access is logged in line with PIPEDA and Ontario employment record practices.
What if a source fails overnight?
The load flags the missing source and the dashboard shows the time of the last successful refresh. An alert reaches the data owner before 7 am.
Let's scope your project. An executive dashboard connected to your 5 sources is scoped in two workshops and delivers its first daily KPIs within 4 weeks. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
