
Quality deadlines stay buried in spreadsheets and surface only during audits. Why manual management fails and what to get in order before the visit.
Delays in quality certifications in SMEs appear when renewals, laboratory tests and compliance documents remain scattered across spreadsheets, inboxes and shared folders. There is no single overview of what is expiring, who needs to sign and which evidence is incomplete. The certification body is not the problem: the problem is the lack of daily oversight. The audit arrives and exposes gaps that have been open for months. What is needed is assigned deadlines, up-to-date evidence and procedures aligned with the actual text of activities, not a calendar forgotten in a spreadsheet.
In small and medium-sized enterprises, the quality manager often shares the role with production, purchasing or management. Renewals, calibrations, licences and test reports live in silos. There is no continuous querying of data: there is only the memory of whoever handled the last visit. When that person is absent or changes role, the deadline stays in the file and nobody asks about it. The practical result is a scramble to recover evidence the week before, with different versions across the management system, folders and paper registers.

Manual management fails because deadlines live in tools that nobody queries every day and because those responsible change without leaving a single shared register.
A shared spreadsheet seems sufficient while the entries are few. Then product variants arrive, new suppliers, instruments to calibrate and courses to renew. Every row depends on a person who must remember to open it. If the reminder sits in a personal inbox, the deadline dies with the holidays. If it sits in a generically named folder, nobody knows which version is valid.
The auditor does not evaluate intentions: they compare procedures, records and physical items. When the procedure text refers to a monthly test and the last quarter is missing from the folder, the delay is already a non-conformity, even if in production "it has always been done that way." The gap originates earlier, in the way tasks are assigned and evidence is filed.
In many SMEs the quality role is a secondary assignment. Whoever produces, buys and signs quotes also accumulates renewals and forms. Without a single list of obligations, every operational urgency wins over the silent deadline. Control only becomes visible again when the certification body's letter sets the date of the visit.
Those who have little time for the daily document routine tend to close the order and postpone the filing. The evidence exists, perhaps on the workbench or in the inbox, but it never enters the correct dossier. At the audit, the link between batch, instrument and signature is missing. This is not laziness: it is a workflow that demands human memory instead of a mandatory step in the management system.
A postponed audit or one closed with findings costs key people's time, blocks orders that require a valid certificate and reopens work that was already closed.
The visible cost is the day spent accompanying the auditor. The invisible cost is the downtime of those who must reconstruct evidence, rewrite procedures and call suppliers back. Meanwhile, a customer who only buys from certified suppliers puts the order on hold. Preparation concentrated into a few days also produces version errors: the old form gets uploaded, an outdated list gets signed, a calibration expired weeks ago gets presented.
Procedures and records work together: if the document in the folder does not match what has been declared, the visit stalls regardless of the tool used for filing.
The same mechanism appears when internal systems do not exchange facts. A certificate renewed in one folder and a supplier master record frozen in the management system tell two different stories. On this point, the same logic applies as described in the signals that business software systems are not talking to each other: the data exists, but it is not the same everywhere.
Automation prevents delays when it turns every deadline into an assigned task, with evidence to upload and a visible block if the document is missing.
A reminder the day before is not enough. What is needed is a rule that, on a defined date, requests a file, a signature and a responsible person. If the evidence does not arrive, the status remains open and appears on the control dashboard, not in a private conversation. The procedure text and the management system fields must say the same things: the same item code, the same date, the same form version.
Artificial intelligence enters after the facts, not in place of the facts. It can read a report, propose a deadline entry, flag an empty field or a discrepancy between a procedure and a record. It cannot invent a calibration or a signature. Those who confuse text generation with compliance evidence arrive at the audit with polished but incomplete documents. The same risk appears in visual quality control that slows down production: without structured data, only the judgement of the moment remains.
Consider a workshop or laboratory with a few dozen references under certification. No department is needed. What is needed is for every code to have its evidence linked and for a supplier's expired document to block use. When references grow, memory does not scale. The spreadsheet that worked with ten rows becomes the blind spot of the audit.
When quality documents are produced by hand, the delay comes from the time spent drafting them, not only from forgetting the date.
At About medically s.r.l., a parapharmaceutical company, the intervention on document generation delivered an 85% increase in document generation speed within 90 days. The principle is the same one needed for an audit: fewer hours spent producing documents by hand leave more time to verify that every dossier is complete before the visit.
At About medically s.r.l., document generation became 85% faster in ninety days.
Another case involves ERP Costruzioni, in the construction sector: drafting a quote went from 8 hours to 5 clicks in 30 days. A quote is not a certificate, but the mechanism is similar. A lengthy document, with data scattered across multiple archives, becomes a guided click when master records, line items and rules all live in the same system. Audit preparation requires the same kind of order, not another hero in the office.

Start with an honest inventory of obligations and evidence, assign a responsible person to each deadline and close each entry only with a verified document.
Before buying a new tool, open the last visit report and copy out the points that have already generated observations. Alongside it, list the real deadlines: management system certificate, calibrations, analyses, licences, qualifications. For each one, write down where the file is today. If the answer is "in someone's inbox," that entry is already at risk.
Then decide on a minimum rule: no non-conformity closed without a date and verification; no active supplier with an expired qualification; no instrument in use with an out-of-period calibration. Those who want to accelerate with artificial intelligence must first know which fields are missing. Otherwise, the disorder gets automated. On budgets and unrealistic expectations, the five budget mistakes that block automation in SMEs remain useful: what gets funded is the theatre of control, not the workflow that produces the evidence.
The procedure gets updated while the old form stays on the shop floor. The PDF gets filed with a name that only makes sense to whoever saved it. A deadline is considered closed because it "was requested," not because the file was accepted. Extra evidence gets accumulated the day before, without linking it to the batch or the instrument. The auditor reads inconsistency, not commitment.
Leomat simplifies SME processes with artificial intelligence automations and the development of tailor-made ERP systems, so that deadlines and documents stop living in isolated folders.
Leomat was created to bring frontier technology to those who run businesses. For a mid-sized Italian company without a structured IT office, the point is to translate quality obligations into measurable workflows, without over-engineering. Among the services is the development of tailor-made ERP systems, designed to bring master records, documents and operational rules together in the same system, not across three parallel archives.
On the document side, the reference remains About medically s.r.l. and, on the shift from hours of drafting to a few clicks, ERP Costruzioni. To understand whether a management system built around real processes can close the gap between procedure and evidence, you can speak with Leomat about business process automations.
Delays are avoided by listing every obligation with a date, a responsible person and an expected document, then closing the entry only when the file is verified. Reminders in a personal inbox are not enough. An open status visible to whoever coordinates production and purchasing is needed. Artificial intelligence can flag empty fields or misaligned texts; it does not replace the evidence. The delay originates when the deadline is known to only one person and the evidence sits in a folder that nobody queries.
Calibrations linked to the wrong instrument, expired supplier qualifications, unsigned training records and non-conformities without a closure verification are the gaps that emerge. Shop-floor forms also go missing when the procedure has only been updated in the text. The list varies by sector, but the mechanism is identical: the fact exists in one place and is absent from the dossier the auditor compares. An inventory based on the last visit report is the most honest starting point.
A spreadsheet handles few entries and a single curator. When references, shifts and suppliers grow, the spreadsheet does not request the file and does not block the use of an expired instrument. A specific product name is not necessarily required: what is needed is for the management system and the archive to show the same dates. A tailor-made ERP makes sense if quality facts currently live outside the master records. A new tool without closure rules simply replicates the delay inside a tidier interface.
No. Automation removes the work of chasing dates and copying forms; it does not remove the judgement on a non-conformity or on the completeness of evidence. The responsible person remains the one who accepts or rejects the evidence. What changes is visibility: an open status no longer depends on the memory of whoever is on holiday. Without a responsible person to validate, the result is only a faster archive of still-incomplete documents.
This post was created with AI
Content (text, processing, quotations and images) generated or artificially manipulated by artificial-intelligence systems. Notice provided under the transparency obligations of Article 50 of Regulation (EU) 2024/1689 (AI Act), applicable from 2 August 2026.