A Stock Count Corrects the Records. The Work That Waits Needs the Cause Fixed.

The plant closed for two days in March for the yearly stock count. Forty people counted every bin. The system was corrected line by line.

For about six weeks the figures were a pleasure. Nobody walked to the stores to check.

It is now the second week of May. This morning the line stopped for a washer the system said there were two thousand of. The count was done properly. The planner has started a private list again, and the finance head wonders what the firm bought with those two days.

What We Found

For this article we went through the public record on what waits while a manufacturer’s figures cannot be trusted: a survey of manufacturers on what holds back their use of AI, a report of an industry summit in India, one survey on why firms keep old systems, and the European Commission’s pages on a new rule that reaches exporters through their buyers.

Data comes first among the barriers to AI. In one survey of manufacturers, run by an industry body, respondents picked their top three barriers to using AI. Problems with data came first, named by about six in ten. The page gives no sample size, and the survey is about AI, not about stock.

Without the groundwork, a pilot stays a pilot. A report of an industry summit held in Bengaluru in 2026 says that “smaller and mid-sized Indian producers encounter a deeper barrier: fragmented data, aging machinery, and a lack of standardized operating frameworks”, and that without that groundwork “a successful pilot might remain merely a pilot”. It names no firm and gives no figure.

Replacement waits on the data too. In one survey of about 500 IT professionals in the US, run by a firm that sells software services, about a third named concerns about moving data among their reasons for not replacing old systems. It covers every sector, not manufacturing alone.

A new EU rule asks importers to declare the emissions in their imports. The European Commission’s page on the rule (updated 26 June 2026) says that “the CBAM definitive regime applies from 1 January 2026”. Under it, “EU importers declare the emissions embedded in their imports and surrender the corresponding number of certificates each year”. It covers iron and steel and aluminium among other goods, for importers of more than 50 tonnes of covered goods. In its questions and answers of 17 December 2025, the Commission also says: “If embedded emissions of goods, as calculated by a third country producer, are not verified by an independent verifier, the importers have to use default values.” For steel and aluminium those default values carry a mark-up of “10% in 2026, 20% in 2027, and 30% in 2028”, and “CBAM certificates will be sold only from 1 February 2027”. The pages place these steps on EU importers, and do not mention India.

The Work That Waits on Figures Nobody Can Vouch For

A count is judged by how right the figures are the week after. Our reading is that the larger cost is the work that cannot start.

• The AI pilot that stays a pilot. A pilot can look good on a sample and fail to spread, because the plants it would spread to hold the same doubtful records.
• The replacement that carries the faults across. A new system filled from old records starts with the old mistakes.
• The buyer who asks for figures per shipment. Our reading, and only that: an Indian steel or aluminium exporter whose production and dispatch records cannot tie a batch to a shipment may find it hard to give its EU buyer figures for each shipment. Where a producer’s figures are not verified, the Commission says, “the importers have to use default values”. Whether the rule touches a particular exporter is for that firm and its advisers. The part we would look at is the records: the supplier data, and the trail from batch to shipment through the firm’s own programs. We do not measure or verify emissions, or advise on the rule.

This is our reading of the evidence. It is not a measured market figure.

Four Ways to Answer Figures That Keep Going Wrong, Weighed

There are four real options. Each one is right for somebody.

Option	When it fits	What it risks
Count again, and count more often	The differences are small and nothing else waits on the figures	It is paid for every year, and the cause is still running
Bring in a data-cleaning specialist before the next project	Years of records must be put right at once, and the causes are already fixed	If the rule that spoiled the records is still running, they drift again
Trace a wrong figure through the code to the rule or step that makes it	The same parts are short at every count, and nobody can say why	A trace corrects nothing on its own. It names the cause; somebody still has to change it
Rebuild the points where figures are entered, with checks, one module at a time	The cause is known to sit in the firm’s own entry screens or programs	Without a trace first, the wrong screen may be rebuilt. A screen that refuses entries annoys people at first, and a check that is too strict is worked around

Where You Are Now, and Where You Should Be

Before the next count is approved, we would set a benchmark for the decision itself.

Measure	Typical starting point	Where it should be	How it is checked
The cost of living with wrong figures	The count’s budget alone	The counts, the checking by hand and the adjustments, in one place	The cost comparison
The cost of fixing the causes	Never priced	A price for each stage	The staged plan and quote
The cause of each kind of difference	”The stores” or “the data”	Named: the rule, conversion or manual step that produces it	The specification and the risk register
What an entry screen accepts	Almost anything	For each rebuilt part, the checks written down, and an entry that cannot be right refused with a reason	The specification and the traceability matrix
The trail from batch to shipment	Not written down	Each step a batch passes through, from production record to dispatch, named in the specification	The specification

The starting points above are what we expect to find. In an engagement they are replaced by your own figures, taken in the first review. The targets are what we would work towards with you. They are not results we have already achieved for a manufacturer.

How We Work With You

1. We listen: a free Reliability Review. A first conversation about the counts and the differences that keep coming back. You leave with a short written note of what we found. You can start with a conversation about the system; we ask for read-only code access only when the assessment begins.
2. We research. We research your sector, the regulator’s position and what the public record says about firms in your situation. This article is an example of that work.
3. We read the system and propose: the Legacy System Specification Assessment. Here the system is the stock, production and dispatch programs you built or had built beside your ERP, including the screens where figures are entered. It produces a map of the system as it stands, a written specification of what it does, a risk register, a cost comparison between keeping and replacing the system, and a staged plan with a fixed quote. AI does a first pass, our senior engineers review it and trace it to the code, and your own experts confirm it. Code access is read-only and logged, never used for training, and deleted within 30 days of the evaluation ending unless we engage. You get the options and our recommendation.
4. If you approve, we build. We replace a system one module at a time while the old one keeps running. Here that means the part that makes a figure wrong, rebuilt with checks at the point of entry, while your own staff correct the existing records. We never do a single cutover. Each module has a traceability matrix. The work is billed by milestone, each milestone tied to a module you accept, and you can pause after any accepted module.
5. We review against the benchmark. Before work starts we set the benchmark. After each module we show you where each measure stands, and the filled-in benchmark is yours at the end.

The Review is free. Every paid step is fixed price, agreed at scoping. Stop after any step. The findings are yours. The assessment fee is credited in full against the first build stage if it starts within 90 days of the report.

What we do not do: we do not clean, correct, merge, reconcile or migrate records, and we do not count stock. Our terms on access cover your code, not your stock, production or customer records. We cannot see inside or change a packaged ERP product, and we do not connect to machines or read counts from them. Errors that begin on the floor are untouched, and nothing here blames the people who work there. We do not measure or verify emissions, and we offer no compliance or legal service: what the EU rule asks of an importer, and what an importer asks of you, is for you and your advisers. We promise no level of stock accuracy and no result from AI. We have no published case of this work for a manufacturer.

What changes for the business:

Role	What they gain
CEO	The yearly count stops being the only time the figures can be believed.
CTO or CIO	The rule that spoiled the figure is replaced, and an impossible entry is refused where it is made.
CFO	Counting and adjusting stop being a cost that returns every year with no end in sight.
COO	A count, once done, stays true for longer, and planners can plan from the screen again.
Head of digital	Figures that are checked as they are entered are figures a pilot can be built on.

The cases in this article, and what IOanyT does about that kind of problem:

The case	What IOanyT does
The European Commission says the definitive regime of the EU’s carbon border rule applies from 1 January 2026 and, in its questions and answers of 17 December 2025, that where a producer’s figures are not verified “the importers have to use default values”, with a mark-up of 10% in 2026 for steel and aluminium. No organisation is named.	Our assessment shows where a firm’s own programs record batches, suppliers and dispatches, and whether a batch can be followed to a shipment. Measuring emissions, verification and what the rule asks of anyone are outside our work.
A report from an Indian summit says that, without groundwork on data, “a successful pilot might remain merely a pilot”. No organisation is named.	We promise no result from AI. We rebuild the part of a firm’s own programs that makes figures wrong; your own staff correct the records.

IOanyT is a consultancy with its own engineering team: we study the problem, recommend, and prove it by building only what you approve. A count corrects what was made wrong; we would rather fix what keeps making it.

Book a free Reliability Review →

Get a fixed quote →

See the Legacy System Specification Assessment →