Somebody spends Sunday evening exporting three files and building a spreadsheet so that Monday's meeting has numbers. It works, everyone relies on it, and it is one person, one laptop and one set of assumptions nobody else can see.
Getting stock and sales in one view is rarely a dashboard problem. It is a connection problem and a definitions problem, and the chart at the end is the small part. Knowing that before you buy changes what you buy.
Why stock and sales are never in one view
In most distribution and wholesale businesses the numbers live in three places that were each chosen sensibly and never designed to talk.
- Stock, in a warehouse or inventory system, which knows quantities and locations but not what anything is worth to you.
- Orders, in a sales system or a shop platform, which knows what customers have asked for but not whether you can supply it.
- Cash, in accounting, which knows what has been invoiced and paid but not what is on a shelf.
Every useful question crosses at least two of them. What is our margin by product this month needs orders and accounting. What can we promise a customer this week needs stock and orders. Which lines are tying up money needs all three. That crossing is the actual project, and it is why a quote that is mostly about charts is quoting the wrong thing. Our systems do not talk to each other covers the mechanics, and the connecting work is data pipelines.
The argument about what in stock means
The delay nobody plans for, and it is not technical.
Ask four people in a wholesale business how much of something you have and you will get four honest answers. The warehouse counts what is on the shelf. Sales counts what is on the shelf minus what is already promised. Purchasing counts that plus what is on order from the supplier. Finance counts what you have paid for, including the container that has not landed.
All four are correct for their purpose, and a single screen showing one number to everyone forces a decision. The dashboard cannot make it and should not try. Someone with authority decides, it is written down, and the screen labels which definition it is showing. Skip that and you will build something people argue with rather than use, which is the same outcome as not building it. A dashboard cannot fix messy data covers the general version.
- Available to promise is usually the one sales needs, and it is the hardest to compute because it depends on allocations.
- On hand is what the warehouse means and it is the easiest to trust, because somebody can go and count it.
- On order and inbound matters for purchasing and needs supplier lead times that are frequently optimistic.
- Stock value is finance's number and depends on a costing method that somebody chose years ago and may not have revisited.
Different screens for different people
The other reason single dashboards disappoint: a warehouse manager and a finance director want genuinely different things, and a screen that serves both usually serves neither.
| Who | Wants to know | Freshness |
|---|---|---|
| Sales | What can I promise, and when | Close to live, because they are on the phone |
| Warehouse | What is coming in, what is going out today | Live |
| Purchasing | What is running out, what is overstocked | Daily is plenty |
| Finance | Margin, stock value, what is tying up cash | Daily or weekly, and accuracy matters more than speed |
| The owner | Three numbers, and whether they are better than last month | Weekly, and it should fit on a phone |
The freshness column has a large cost attached. Live is a different architecture from daily, and almost nobody who asks for live needs it beyond the first two rows.
The number that is always wrong
Worth its own section because it derails more of these projects than anything technical: margin.
Everyone wants margin by product on the dashboard, and it is the figure most likely to be wrong, because it depends on a cost that nobody looks at closely. If your costing has not been reviewed in a while, the dashboard will confidently report profitability that is not real, and it will be believed for months.
- Which cost are you using? The last price you paid, an average, or the standard cost somebody set two years ago. On imported goods with a moving exchange rate these can differ substantially.
- Is carriage in the cost? Freight, duty and handling frequently are not, which flatters margin on exactly the products where it matters most.
- What about the discounts you give? A list price margin and a realised margin after settlement discounts and rebates are different numbers, and sales usually quotes the first.
- And the ones you receive? Supplier rebates that arrive quarterly rarely find their way back to the product they came from.
The practical advice is to put margin on the dashboard last rather than first, and to label plainly which cost it uses. A screen that shows revenue and volume reliably is more useful than one that shows margin unreliably, because the second gets disputed in a meeting and takes the rest of the screen down with it.
Start with the Monday meeting
The most useful specification you can write is already being produced every week by whoever builds the spreadsheet.
- Take the actual Sunday spreadsheet and mark which figures anyone looks at. It is usually a fraction of what is in it.
- Ask what decision each one drives. A number nobody acts on does not need automating, it needs deleting.
- Watch where the builder makes a judgement call, because those are the definitions you will have to settle, and they are currently living in one person's head.
- Build only those figures first. Three or four, not thirty, and let the rest wait until people ask.
- Keep the spreadsheet running alongside for a month and compare. Where they disagree, one of them is wrong and finding out which is the point.
- Then stop building it, deliberately, and make sure the person who did gets their Sunday back rather than a new job maintaining a dashboard.
What it costs
| Scope | Cost | Note |
|---|---|---|
| Reports from one system you already own | Nothing, plus an afternoon | Check this first. Modern systems report better than people assume. |
| Two systems joined, a handful of figures | $15,000 to $40,000 | The common case, and where most of the value is. |
| Three systems, with definitions settled and role-based views | $40,000 to $100,000 | Realistic for a wholesale business with real complexity. |
| The above plus forecasting | $80,000 and up | A different project. Get the reporting trustworthy first. |
Illustrative ranges from the kind of work we quote, not a price list. Roughly four fifths of the effort is connecting and agreeing; the dashboard itself is the visible fifth. What does a business dashboard cost explains that split.
Forecasting is worth a caution. It is the thing everyone wants once the reporting works, and it is a real capability, but a forecast built on numbers people do not yet trust inherits the distrust and amplifies it. Get the current view believed first. That is what forecasting work depends on.
Not for you if
One risk worth naming, and it is the one that usually decides this in the end. The Sunday spreadsheet is a single point of failure wearing a cardigan. It works precisely because one person understands it, and the day they leave or fall ill, nobody can rebuild the assumptions inside it. That is frequently a better reason to do this than the reporting itself, and it is the one least likely to be in the business case.