Part of the LIMS software guide
Method · Control plan

Quality control plan: method and examples in the food industry.

Parameter, method, frequency, threshold, owner, record: the method to build and manage a control plan for your raw materials, products and suppliers, with an example from one product line.

A control plan is the document that specifies, for each raw material, product and supplier, what is checked, with which method, how often, against which threshold, by whom, and where the result is kept. This page describes a method for building and managing it, from defining requirements to following up deviations, with an example from one product line. It is aimed at food manufacturers who must monitor their raw materials, products and suppliers, not at the testing laboratory that performs the tests.

Key points
  • A quality control plan is judged on its consistency (requirement → parameter → threshold → action), not on its number of analyses.
  • Six headings are essential: parameter, method, frequency, threshold, responsibility, record.
  • A plan is built from the requirements, then prioritised by risk, supplier by supplier and origin by origin.
  • The planned control plan and the executed control plan must remain comparable: the gap between the two is the first management indicator.
  • A plan not revised after a change of supplier, origin or process is an obsolete plan, even if it is being applied.

What is a control plan, and where does its scope end?

A control plan is the document that sets out, for each material, product or step, what to check, how, how often, against which threshold, by whom and with which record; it stops where carrying out the checks and deciding on the batch begin.

Operational definition

The control plan is a table of decisions. Each line links an object to monitor (a raw material, a finished product, a packaging material, sometimes a contact surface or an environmental parameter) to a parameter, a method, a frequency, an acceptance threshold, an owner and a record.

A raw material control plan and a finished product control plan can sit in the same document or in two separate documents: what matters is that the scope is explicit, stable over time and known to the quality, purchasing and laboratory teams. A useful plan can be read without instructions: if someone who did not write it cannot say who samples, who analyses and who decides, the responsibility heading is incomplete.

What the control plan is not

  • An analysis plan: this is the analytical part of the system (parameters, matrices, methods, laboratories). It feeds the control plan, it does not replace it.
  • A sampling plan: it describes how to sample (sample size, number of units, sampling method). A frequency without a sampling rule is not reproducible. The method is detailed on our sampling plan page.
  • A monitoring plan: it pursues a different objective (see below) and is not managed with the same indicators.
  • A complete documentation system: the control plan refers to procedures, methods and forms; it does not contain them.

A regulatory framework that imposes obligations, not a template

No European regulation imposes the name “control plan”, nor a standard format. The obligations it rests on are found elsewhere:

  • Regulation (EC) No 178/2002, Article 17: food business operators are responsible for compliance with food law at all stages of production and distribution under their control. The control plan is one of the documents that evidence this responsibility.
  • Regulation (EC) No 852/2004, Article 5: operators put in place, implement and maintain permanent procedures based on HACCP principles, with documents and records appropriate to the nature and size of the business. The regulation also provides for these procedures to be reviewed when a product, process or step changes.
  • Regulation (EU) 2017/625, Article 9(1): competent authorities perform official controls on operators regularly, on a risk basis and with appropriate frequency, taking into account in particular the reliability and results of their own checks. A well-kept, documented control plan therefore affects the intensity of official controls, without exempting from them.

Thresholds come from specific regulations: maximum residue levels of pesticides, Regulation (EC) No 396/2005; maximum levels for contaminants, Regulation (EU) 2023/915, applicable since 25 May 2023 and repealing Regulation (EC) No 1881/2006; microbiological criteria, Regulation (EC) No 2073/2005; plastic materials in contact with food, Regulation (EU) No 10/2011; food information to consumers and allergens, Regulation (EU) No 1169/2011.

References consulted on 12 November 2026: Regulation (EC) No 178/2002, Regulation (EC) No 852/2004, Regulation (EU) 2017/625.

Which headings should a quality control plan contain?

A control plan line reads like an instruction: it must allow the check to be carried out and then recorded without interpretation. Six headings are the minimum; the others make the difference at audit and when handling deviations.

HeadingWhat it must make explicitQuestion to ask
Object and scopeMaterial, product, packaging or surface concerned; reference, supplier, originWould I know, without ambiguity, what this line applies to?
ParameterThe hazard or characteristic measured (contaminant, allergen, micro-organism, aw, net weight, integrity)Does this parameter really cover the identified risk?
MethodStandard or in-house method, laboratory chosen, scope of accreditation, limit of quantificationIs the limit of quantification below the threshold to compare against?
FrequencyNumber of checks per batch, per delivery, per campaign or per year, with the justificationIs this frequency proportionate to the risk and the volume received?
ThresholdRegulatory limit, customer specification and internal threshold, explicitly distinguished, with a warning level and a blocking levelWhich threshold triggers a decision, and which decision?
ResponsibilityWho samples, who orders the analysis, who interprets, who decides what happens nextWhat happens if the named person is absent?
RecordMedium, format, retention period, filing locationWould I find the result in under five minutes, two years later?
Action in the event of a deviationBlock, isolate, inform the supplier, re-analysis, root cause analysis, deadlineIs the action decided in advance, or improvised in the heat of the moment?
RevisionTriggers (result, complaint, change, new standard) and date of last reviewWhen was this line last reviewed?

Two points of method avoid most internal disputes. First, clearly separate the three sources of threshold: a regulatory limit is not negotiable, a customer specification is read in the specification document, an internal threshold is a management choice, often stricter than the other two. Giving each threshold two levels, warning and blocking, avoids treating a result approaching the limit in the same way as one exceeding it. Second, write down the comparison rule: a result close to the threshold is read in the light of measurement uncertainty and the limit of quantification, otherwise two people will reach different conclusions on the same report.

From a requirement to a written decision

  1. RequirementRegulation, customer specification or internal rule
  2. Parameter and methodLimit of quantification below the threshold
  3. Two-level thresholdWarning and blocking, with its source and date
  4. Planned actionBlock, re-analysis, inform the supplier
Threshold with no sourceUnjustified frequencyAction improvised in the heat of the moment

Each line of the plan links these four elements; a break in any one of them makes the line unworkable or indefensible at audit.

How to build a control plan in seven steps

1. List the requirements. For each product, list the applicable regulatory requirements, contractual customer specifications, audit standard requirements and internal rules. A requirement with no corresponding plan line is an uncovered requirement: exactly what an auditor looks for.

2. Map object × hazard pairs. Cross raw materials, finished products, packaging and surfaces with the relevant hazards: biological, chemical, physical, allergens, authenticity, labelling compliance. This matrix, derived from the hazard analysis, is the source of the plan.

3. Prioritise by risk. Not all pairs call for the same intensity of checking. Cross the severity of the hazard, its likelihood given the origin and process, the supplier's history (previous results, non-conformities, audits, complaints) and the volume received. A new supplier or a new origin automatically moves up the ranking.

4. Choose parameter and method. For each line, choose the available reference method and check its limit of quantification against the threshold. A method whose limit is above the threshold does not allow a conclusion: it must be discarded or supplemented.

5. Set frequency and sampling. Frequency is justified, not copied from one site to another. Write down the reasoning (“one analysis per delivery and per origin for the first six months, then review”) and attach the corresponding sampling rule. The sampling plan is a separate but inseparable document.

6. Name responsibilities and records. A line without a job title cannot be applied. Specify where the result is kept and for how long, including contract laboratory reports and supplier certificates of analysis.

7. Approve, distribute, revise. The plan is approved by quality management, distributed to the teams concerned and revised according to written triggers: new supplier, new origin, new product, process change, regulatory change, repeated non-conformity, audit result.

Free templateQuality control plan template in Excel format

Tabs, columns, formulas and drop-down lists ready to adapt to your materials, products and suppliers.

A concrete example: a dry biscuit line

Take an SME that makes dry biscuits and buys wheat flour, butter, chocolate chips and a laminated packaging film. The table below illustrates three blocks of the same control plan. The frequencies shown are examples to adjust: they depend on the hazard analysis, customer specifications and the volume received.

BlockParameterMethodFrequencyThresholdOwner / record
Raw material - wheat flourPesticide residuesMulti-residue method in an external laboratoryExample: one analysis per origin and per campaign, increased for a high-risk originRegulatory limit: annex to Regulation (EC) No 396/2005 for the food concernedIncoming quality / test report filed in the supplier file
Raw material - wheat flour and chocolate chipsMycotoxinsStandard method in an external laboratoryExample: one analysis per batch for a new origin, then one monthly analysis per supplierMaximum levels: Annex I to Regulation (EU) 2023/915Incoming quality / report and release decision
Raw material - butterComposition (fat content), no spoilageSupplier technical data sheet + sensory check on receiptAt every deliverySupplier specification (example: fat ≥ 82%)Incoming quality / goods-in inspection sheet
Raw material - allAllergens and material labellingTechnical data sheet and supplier control planAt every change of supplier or formulationDeclaration compliant with Regulation (EU) No 1169/2011Purchasing + quality / supplier file
Finished product - biscuitsAverage net weightWeight check on a representative sample of the batch, in-house methodExample: one check per production batch and per hour of productionInternal specification and labelling toleranceProduction + quality / finished product inspection sheet
Finished product - biscuitsMicrobiological contaminationStandard method in an external laboratoryExample: one analysis per week of productionRegulation (EC) No 2073/2005 criteria where they apply, then internal and customer criteria, written with their sourceQuality / release report
Packaging - laminated filmSuitability for food contactSupplier declaration of compliance + migration evidenceAt approval and at every material changeRegulations (EC) No 1935/2004 and (EU) No 10/2011 depending on the materialPurchasing + quality / packaging file
Packaging - laminated filmLegibility and accuracy of printed informationVisual check on a sample, internal templateAt start-up and end of run, at every product changeApproved artwork and Regulation (EU) No 1169/2011 particularsLine quality / start-up sheet

How to link the control plan, the monitoring plan and customer standards?

The two terms are often used interchangeably. In an industrial setup, they do not answer the same question.

Monitoring planControl plan
ObjectiveEstimate an average level or track a trend across a population of batches, suppliers or originsCheck the compliance of specific objects and detect risk situations
SamplingRepresentative, often randomTargeted by risk, history, suspicion or change
DecisionAdjust the strategy, revise frequencies, alert a supplierRelease, block, isolate, open a non-conformity
IndicatorsTrend, drift, comparison between originsCompliance rate, completion rate, lead times

In practice, the same sample can feed both: a pesticide analysis on a batch of flour produced at regular intervals serves the monitoring plan (trend by origin) and, if it is linked to an identified delivery, the control plan. What changes is the intention and how the result is used; write it down, otherwise the data will be counted twice or not at all.

Standards then shape the content. IFS Food version 8 (published in April 2023, mandatory since 1 January 2024) requires documented, risk-based analysis and monitoring plans covering at least raw materials, relevant semi-finished products, finished products, packaging materials, contact surfaces and relevant environmental monitoring parameters (requirement 5.6.1). ISO 22000:2018 deals with hazard control in clause 8.5.4 and verification in clause 8.8, with associated records. Customer standards add parameters, minimum frequencies and mandatory methods.

The most robust method is a correspondence matrix: a “requirement” column (regulation, clause of a standard, article of the customer specification), a “control plan line” column, an “evidence available” column. Any requirement without a line, or any line without a requirement, shows up immediately.

Why distinguish the planned control plan from the executed control plan?

The planned control plan is the programme: the lines and frequencies set for the period, supplier by supplier, origin by origin. The executed plan is what was actually done: sampling dates, laboratories, results, decisions taken, deviations and follow-up.

Comparing the two is the most useful management information in the system. A 100% completion rate with a stable compliance rate does not tell the same story as a 70% completion rate covered by reassuring talk. The gap reads both ways: a planned line not carried out must carry a written justification (supply disruption, batch not received, laboratory unavailable), and an analysis carried out outside the plan must be integrated, otherwise the history becomes fragmented.

Three rules make the reconciliation possible:

  • A unique reference per plan line, used identically in records and reports.
  • A fixed periodic planned/executed reconciliation (monthly or quarterly), run by quality with purchasing.
  • A consistent retention period across the plan, reports and decisions, so that reconstruction remains possible several years later.
With Tracklab

Tracklab tracks the coverage of your analyses: frequency and date of the last analysis per product × supplier pair. For each report received, it calculates the compliance status of each parameter against your regulatory and internal standards, and keeps a trace of the threshold applied. The control plan and its implementation remain in your procedures. See Tracklab's LIMS software page.

Which common mistakes weaken a control plan?

Copying a generic plan. A plan imported from another site, another product line or the previous year describes risks that are not yours. The first review must start again from the hazard analysis.

Confusing the three sources of threshold. Mixing regulatory limit, customer specification and internal threshold leads to inconsistent decisions from one operator to another, and to undetected deviations.

Working on an outdated regulatory threshold. The annexes to the pesticide and contaminant regulations are amended several times a year. A threshold copied without its version or date produces a wrong status.

Setting frequencies without justification. A “historical” or copied frequency does not stand up to a simple question at audit: why this number, for this supplier, at this time?

Forgetting the action in the event of a deviation. A plan that says what to measure but not what to do when a limit is exceeded produces results with no effect. Blocking, re-analysis and informing the supplier must be decided in advance. The non-conformity report is the medium for this.

Separating the plan from the records. Results in inboxes, reports in local folders, a spreadsheet kept by one person: the planned and executed plans become impossible to compare, and reconstruction impossible.

Not revising after a change. New supplier, new origin, new line, new formulation, new customer standard: each change is a revision trigger, not a forgotten memo.

Treating packaging and environment as appendices. Contact materials and environmental parameters are part of the scope expected by the standards, and are frequent causes of non-conformity.

Leaving the plan to the laboratory alone. Quality, purchasing and laboratory teams must share the same reading of thresholds and supplier consequences, otherwise decisions contradict one another.

Which indicators manage a control plan?

A control plan is managed with few indicators, but stable ones that are comparable from one period to the next.

IndicatorHow it is calculatedWhat it revealsFrequency
Completion rateChecks carried out / checks planned over the periodThe gap between intention and executionMonthly
Compliance rateLines with a compliant result / lines with a resultThe level of control, to be analysed by supplier and by originMonthly
Result lead timeResult date − sampling dateThe ability to decide before releaseQuarterly
Deviation handling timeClosure date − detection dateThe actual responsiveness of the systemMonthly
Concentration of non-conformitiesNumber of non-conformities by supplier, origin, parameterWhere the risk liesQuarterly
Plan obsolescenceLines reviewed in the last twelve months / total linesThe documentation riskAnnual
Cost of controlCost of analyses / tonne produced or per itemThe sustainability of the system, to be related to the risk coveredAnnual

Calculating the compliance status of a test report - comparison with the threshold, taking the limit of quantification into account, distinguishing “not detected” from “not analysed” - deserves a written, shared rule, otherwise two readers of the same result will reach two different statuses. This is the scope of Tracklab's test report compliance calculation: each parameter is compared with your thresholds and the four built-in European regulations, with the regulation, version and date applied. Tracklab calculates and traces; the decision stays with your quality teams.

With Tracklab

Tracklab gathers reports from your suppliers and laboratories, calculates the compliance status of each parameter and keeps a trace of the threshold applied. Discover Tracklab's LIMS software, built for the manufacturer who receives analyses.

Key takeaways

A control plan is not a catalogue of analyses: it is a set of written decisions, linked to identified requirements, prioritised by risk and applicable by named people. Three concrete actions let you move forward straight away: build the requirements × plan lines matrix to spot the gaps, check that each line carries a threshold, an owner and a record, and set up the periodic reconciliation between planned and executed plans.

On your own reports

To see the compliance calculation applied to your own test reports, book a meeting with the Tracklab team.

Before / after

A copied plan or a constructed plan.

The difference lies not in the number of analyses, but in the consistency between requirement, threshold, frequency and action.

Copied plan

A spreadsheet inherited from another site or from last year

Constructed plan

A set of written, reviewed decisions

  1. Copied plan: “Historical” frequencies nobody can justify

    Constructed plan: Each line linked to an identified requirement

  2. Copied plan: Regulatory, customer and internal thresholds mixed together

    Constructed plan: Sourced, dated, two-level thresholds

  3. Copied plan: No action planned when a limit is exceeded

    Constructed plan: A periodic reconciliation between planned and executed

Frequently asked questions

Control plan: your questions.

Is a control plan mandatory in the food industry?

No European regulation requires a document called a “control plan”. However, Regulation (EC) No 852/2004 (Article 5) requires permanent procedures based on HACCP principles, with associated documents and records, and Regulation (EC) No 178/2002 (Article 17) places responsibility for compliance on the operator. Audit standards (IFS Food, BRCGS Food Safety, ISO 22000) and customer specifications, for their part, require documented plans. In practice, a control plan is therefore expected everywhere, under varying names and formats.

What frequency should be used for a raw material control plan?

There is no standard frequency: it is justified by risk (severity of the hazard, origin, process), the supplier's history, the volume received and customer requirements. Two principles help decide: temporarily increase the frequency for any new supplier or new origin, then review it based on results; and link each frequency to a written sampling rule, otherwise the check is not reproducible.

Who writes and approves a control plan?

Writing is led by quality, with purchasing for the supplier part and the laboratory for methods and analytical thresholds. Approval rests with quality management, which commits to the system. Implementation is shared: goods-in and production carry out checks, purchasing handles supplier consequences, the laboratory interprets results. A plan written by a single function is generally incomplete on responsibilities.

Is a spreadsheet control plan enough?

A spreadsheet is enough to set up and run a low-volume control plan, provided three points are locked down: a single identified version, stable line references and an accessible backup. The limits appear with the number of suppliers, items and reports: re-keying results, collecting reports and comparing with thresholds then become costly and fragile. That is the point at which software takes over from the spreadsheet.

How do you revise a control plan?

Set written triggers rather than an anniversary date alone: new supplier, new origin, new product, process or packaging change, regulatory change, repeated non-conformity, audit finding, customer complaint. At each revision, keep the previous version, the date, the reason and the effect on frequencies. An annual review remains useful to check overall consistency and obsolete lines.

Laboratory analyses module · demo on your data

Your test reports, compared with your thresholds?

Tracklab gathers reports from your suppliers and laboratories, calculates the compliance status of each parameter and keeps a trace of the threshold applied. The decision stays with your quality teams.

Our clients

Reliable supplier data, managed with Tracklab.

Their quality and purchasing teams collect, control and leverage their supplier data with Tracklab.

150+

companies supported

40,000+

suppliers referenced

90,000+

products and ingredients

Customer testimonial · Bouvard
“Tracklab has eliminated the blind spots that existed in our document monitoring. Every document is analysed, every change is detected and every decision is traceable. This has enabled us to take a real step forward in terms of professionalism and credibility, particularly with auditors and customers.”
EVEllen V.Quality Director, Bouvard Pro
CyberVadis
CyberVadis Excellence rating

Independent assessment of our cybersecurity framework and of the protection of the data entrusted to Tracklab.

They use Tracklab

  • Somapro
  • Jean Niel
  • La Fournée Dorée
  • Nautilus
  • Cérélia
  • Will&Co
  • Neta
  • Diffusions Aromatiques