Why Three Systems Give You Three Different Numbers

Somebody asks how many deals closed last month. Sales says eleven. Finance says eight. The spreadsheet the operations team keeps says nine.

Everybody is right. Sales counted verbal agreements. Finance counted invoices raised. Operations counted jobs that started. Three defensible answers to one question, and the meeting is now about whose number is correct rather than what to do about it.

This is the most common reporting problem in businesses between five and fifty million, and it is almost never solved by buying software.

The dashboard that stopped being used

Most companies have already tried once. Somebody built a dashboard. It looked impressive for about six weeks. Then people drifted back to their own spreadsheets and it now gets opened when a consultant asks about it.

Two reasons, and neither is the tool.

It was built from whatever data happened to be available rather than from the decisions anyone actually makes. Twenty numbers, four of which mattered, none defined tightly enough for two departments to agree.

And it went stale. If refreshing it requires somebody to export and paste, it is out of date within a fortnight, and a number that might be old is a number nobody will act on.

Write the definitions down

This is the unglamorous part, and skipping it is why the expensive version fails.

A small number of terms have to mean one thing across the business. What counts as a lead: anybody who filled in a form, or somebody who met your criteria? When is a deal closed: verbal agreement, signature, or payment? Does revenue land on the date invoiced or the date received? Is a returning client a new sale? Does a cancelled job come back out of last month's numbers?

None of these have universally correct answers. They only need consistent ones. An afternoon spent agreeing them removes most of the disagreements that make reporting useless, and it costs nothing but the afternoon.

Then write them somewhere findable. A definition living in one person's head is the same problem in a different shape.

Start from the decision

The right way to build reporting is backwards. Not what can we measure, but what do we decide, how often, and what would we need to know to decide it well.

For most owner run businesses the list is short. Where does next month's marketing budget go. Do we need to hire. Which clients are quietly unprofitable. Is the pipeline enough to cover the quarter. Which service line is actually carrying the business.

Five questions, perhaps three numbers behind each. That is your requirement, and it is far smaller than the dashboard you were about to commission. Everything else is interesting rather than useful, and interesting is what makes people stop opening the page.

Where the numbers actually live

Before connecting anything it helps to know what you are connecting. The map is usually shorter than expected and always contains one surprise.

Pipeline sits in the CRM, assuming people fill it in. Money sits in the accounting system and rarely agrees with the CRM, because one records invoices and the other records optimism. Marketing spend sits in the ad platforms, each with its own definition of a conversion. Delivery sits in a project tool, a spreadsheet, or somebody's calendar. And a surprising amount sits in email, which is where exceptions get agreed and never recorded anywhere else.

Doing that inventory usually reveals that one number everybody quotes has no system behind it at all.

The weekly summary beats the dashboard

The highest value output is not a dashboard. It is a short written summary that arrives before the week starts.

A dashboard needs somebody to remember to open it, and busy people do not. A summary in an inbox on Monday morning, carrying five numbers, what changed, what caused it and what needs attention, gets read because reading it is easier than not reading it.

It also changes the meeting. When everyone has seen the same five numbers before sitting down, the hour goes on what to do rather than on establishing what happened.

Traceability is the trust test

One rule protects the whole thing. Every figure has to be traceable back to the record it came from.

The first time a number looks wrong and nobody can explain where it came from, the system loses credibility permanently. People go back to their spreadsheets and you are worse off than before, because now there is a fourth version of the truth.

If a figure cannot be followed back to a specific record in a specific system, do not put it on the page.

Who owns it afterwards

Reporting decays faster than almost anything else, because the business changes underneath it. A new service line appears that does not fit the categories. Somebody renames a pipeline stage. A definition drifts.

Somebody has to own the definitions, not the dashboard. Their job is to notice when the business has changed shape and update what the numbers mean before the reporting quietly becomes wrong. In a small company that is fifteen minutes a month, and it is the difference between a system people trust in year two and one they abandoned in month seven.

Attribution, honestly

Connecting revenue back to what produced it is the thing most businesses want and the thing most likely to be oversold to them.

Some of it is straightforward plumbing. If enquiries carry their source, and that source survives into the CRM and onto the closed deal, you can answer which channel produced the revenue. That alone is more than most companies have.

What is not straightforward is splitting credit across several touches. A client who saw an ad in March, read something in May and was referred in July is not cleanly attributable to any one of them, and any model claiming otherwise is making a choice you should know about. Use attribution to decide where to put more money, not to prove which channel deserves it. The first is a decision. The second is an argument.

What AI does once the definitions exist

With the terms agreed, the rest is the part that was never practical by hand, and it is where AI earns its place.

It reads the material your systems cannot report on. Invoices as documents, free text notes in a CRM, supplier statements, email threads where the exceptions were agreed and never recorded anywhere else. That unstructured material is usually the reason a number had no system behind it, and turning it into structured fields is genuinely hard to do any other way.

It also explains rather than displays. A chart tells you cost per client rose eighteen percent. AI tells you it rose because one channel scaled while converting worse and the other three were flat, every week, which is the sentence you actually needed.

And it watches continuously, so a problem starting on the second of the month is raised on the second rather than found on the thirtieth.

AI Optimize starts from the decisions you make each week and works back to the numbers that answer them, connected, live, traceable and summarised before Monday. That work sits under Reporting & Data.