Most mid-market variance commentary never reaches a genuine root cause — phrases like “timing differences,” “one-off,” and “phasing” dominate monthly reports without diagnosing the underlying issue. Root cause analysis adapted for financial variances requires decomposing total-level deviations into granular components before diagnosis, then classifying causes into four categories: operational, commercial, external, and planning — each demanding a different corrective response. Building a root cause library over successive periods reveals systemic patterns that surface-level analysis consistently misses. Critically, sometimes the root cause is the plan itself — a flawed budget produces meaningless variances, and the only corrective action is plan revision. Finance teams that move beyond proximate causes to genuine root causes convert variance reports from passive commentary into decision-triggering analysis.
Open any monthly variance report in a mid-market company and count the explanations. “Timing differences.” “One-off.” “Phasing.” “Market conditions.” These four phrases account for the vast majority of variance commentary in most organisations. They appear month after month, quarter after quarter, applied indiscriminately to variances large and small, favourable and unfavourable.
This is not analysis. When every problem receives the same diagnosis, no diagnosis is happening.
Root cause analysis — the systematic process of identifying the fundamental reason behind a deviation — is well established in manufacturing (Six Sigma, Lean) and IT (incident management). It has never been systematically adapted for financial variance analysis in mid-market companies. This article fills that gap.
What Root Cause Analysis Means in a Financial Context
Root cause analysis (RCA) is the process of moving beyond the surface description of a variance to identify the fundamental reason it occurred. The distinction has three layers:
| Layer | Description | Example |
|---|---|---|
| Symptom | The variance itself | Revenue is £150,000 below plan |
| Proximate cause | What happened | A key customer delayed their Q1 order |
| Root cause | Why it happened | Single-source dependency on a supplier caused delivery delays, leading the customer to postpone |
Most financial variance commentary stops at the proximate cause — and frequently does not even reach that far. “Revenue was below plan due to lower sales” is the variance restated, not explained. “Revenue was below plan because Customer X delayed their order” is a proximate cause. “Customer X delayed because our delivery lead time extended from 3 weeks to 7 weeks due to single-supplier dependency in our supply chain” is a root cause. Only the root cause points to something the organisation can change.
Four Categories of Root Cause
Not all root causes are the same, and each category requires a different response:
Operational root causes relate to internal processes, efficiency, and execution. Examples: production yield decline, delivery delays, quality failures, staffing shortages. These are typically within management control and addressable through operational improvement.
Commercial root causes relate to pricing, customer behaviour, and market positioning. Examples: price erosion from competitive pressure, customer churn, mix shift toward lower-margin products, channel mix changes. These require commercial decisions — pricing reviews, customer retention initiatives, portfolio adjustments.
External root causes relate to factors outside the organisation’s control. Examples: regulatory changes, foreign exchange movements, commodity price shifts, macroeconomic conditions. These are not directly addressable but must be documented, quantified, and — where possible — hedged or incorporated into revised plans.
Planning root causes relate to flawed assumptions, methodology errors, or unrealistic targets in the original budget. Examples: revenue growth assumption unsupported by pipeline, cost inflation not reflected in plan, seasonal patterns not properly modelled. When the root cause is the plan itself, the corrective action is not operational change but plan revision.
This fourth category — planning root causes — is almost never formally recognised. Yet every finance professional knows the feeling: “The budget was wrong to begin with — so what exactly are we comparing to?” Acknowledging planning methodology as a legitimate root cause category is a significant shift. It means that some variances are not performance failures but planning failures, and the appropriate response is different.
Why This Matters
Recurring variances persist when root causes go unidentified
The same £40,000 material cost overrun appears in January, February, March, and April. Each month it receives the same explanation: “supplier price increase.” By May, the cumulative impact is £200,000 — but no one has investigated whether the supplier contract should be renegotiated, whether alternative suppliers exist, or whether the budget assumption on material costs was simply wrong. The variance was reported four times. It was diagnosed zero times.
IMA (Institute of Management Accountants) research confirms that fewer than 25% of mid-sized companies decompose variances beyond the total-level budget comparison. Without decomposition, root cause analysis is structurally impossible — you cannot diagnose what you have not broken apart.
Decision quality degrades with superficial explanations
When leadership receives variance reports with shallow commentary, two things happen. First, they stop reading the commentary because it adds no information beyond the numbers. Second, they lose the ability to make informed corrective decisions because the analysis never identifies what specifically went wrong and what specifically could be changed.
Ventana Research finds that companies with structured variance analysis are 2.4x more likely to meet or exceed their financial targets. RCA is what makes variance analysis “structured” — it provides the diagnostic rigour that converts numbers into understanding.
Budget credibility erodes when “one-offs” recur
If the same item is classified as a “one-off” in three consecutive quarters, it is not a one-off. It is either a recurring cost that the budget does not capture or an operational issue that nobody is addressing. Either way, the label has become a way of avoiding analysis rather than performing it.
Aberdeen Group research shows that organisations with driver-based planning — which requires understanding the performance drivers behind results — report 24% improvement in forecast accuracy. Driver-based thinking is the foundation of effective RCA because it provides the causal framework for diagnosis.
A Five-Step RCA Process for Financial Variances
Step 1 — Decompose before diagnosing
Break the total variance into meaningful components before attempting to explain any of them. A total-level variance — “costs are £300,000 over budget” — is too aggregated for diagnosis. Decompose by:
- Type: price, volume, mix, efficiency
- Category: cost of goods, operating expenses, overhead
- Responsibility: department, division, project, product line
- Time: monthly trend, cumulative, year-over-year
Each component variance is a more specific question. Specific questions produce specific answers. Aggregated questions produce “timing differences.”
Step 2 — Classify the variance type
For each material decomposed variance, classify it as operational, commercial, external, or planning. This classification determines who owns the investigation, what kind of evidence to gather, and what range of responses is available.
| Classification | Owner | Evidence Required | Response Range |
|---|---|---|---|
| Operational | Operations / Production | Process data, efficiency metrics, incident reports | Process improvement, resource reallocation, training |
| Commercial | Sales / Commercial | Customer data, pricing analysis, market intelligence | Pricing review, customer engagement, portfolio adjustment |
| External | Finance / Risk | Market data, regulatory filings, macro indicators | Hedging, plan revision, scenario updating |
| Planning | Finance / FP&A | Assumption documentation, methodology review | Budget revision, assumption updating, process improvement |
Step 3 — Apply the five whys
The “five whys” technique, adapted from manufacturing quality management, works by asking “why” iteratively until reaching an actionable root cause. In financial variance analysis, this typically requires three to five iterations.
Example:
- Revenue is £150,000 below plan. Why?
- Volume is down by 300 units. Why?
- Key customer delayed their Q1 orders to Q2. Why?
- Our delivery lead time extended from 3 weeks to 7 weeks. Why?
- Primary raw material supplier had a production disruption, and we had no qualified alternative. Root cause: single-source supplier dependency in procurement.
The first “why” is financial (revenue is short). The second is commercial (volume is down). The third is customer-specific (one customer delayed). The fourth is operational (delivery lead times extended). The fifth is structural (procurement risk concentration). Only the fifth points to something that can be systematically addressed.
A second example — where the root cause is the plan:
- Labour costs are £80,000 over budget. Why?
- Headcount is 5 FTEs above plan. Why?
- The business hired to meet demand that materialised in Q4 of the prior year. Why?
- The budget was built in September, before Q4 demand was known. Why?
- The budget cycle closes before Q4 actuals are available, and no mechanism exists to update headcount assumptions. Root cause: budget timing methodology — the plan does not incorporate the most recent information.
In this case, the corrective action is not to reduce headcount. It is to revise the budgeting process so that late-year information feeds into headcount assumptions — perhaps through a January reforecast that updates the annual plan.
Step 4 — Distinguish controllable from uncontrollable
Not every root cause requires corrective action, but every root cause requires a conscious decision:
- Act: The root cause is within management control and the cost of correction is justified. Assign an owner, define the action, set a deadline.
- Accept: The root cause is real but the cost of correction exceeds the benefit, or the variance is within acceptable risk tolerance. Document the decision to accept.
- Adjust the plan: The root cause is a planning failure. Revise the budget or forecast assumptions to reflect reality.
The critical discipline is making this decision explicitly rather than allowing variances to persist in a grey zone where nobody has decided whether to act, accept, or adjust.
Step 5 — Build the root cause library
Over time, categorise and catalogue root causes. This serves three purposes:
Acceleration: When a familiar variance pattern appears, the root cause library provides a starting hypothesis. Instead of investigating from scratch, the analyst checks whether a known root cause applies.
Pattern recognition: Recurring root causes across periods reveal systemic issues. If “single-source supplier dependency” appears as a root cause for three different variance lines, procurement risk concentration is a systemic issue — not three separate problems.
Institutional memory: When the analyst who investigated last quarter’s variance leaves the organisation, the root cause library preserves the diagnosis. Without it, the next analyst starts from zero.
A root cause library does not require specialised technology. A structured spreadsheet with columns for date, variance line, decomposition, root cause category, root cause description, decision (act/accept/adjust), and follow-up status is sufficient.
The Common Excuse Decoder
Surface-level variance explanations can be translated into diagnostic starting points:
| Common Explanation | Diagnostic Question |
|---|---|
| “Timing differences” | Is this a genuine timing shift (the revenue or cost will appear in a future period) or a systematic recognition issue? Has this “timing” explanation appeared before for the same line? |
| “One-off” | Has this “one-off” appeared in prior periods? If so, how often? At what point does a recurring item stop being one-off? |
| “Market conditions” | Which specific market condition? Can it be quantified? Is the impact temporary or structural? Did competitors experience the same effect? |
| “Phasing” | Was the original phasing assumption documented? If the phasing was wrong, should the annual plan be re-phased? |
| “FX impact” | What is the underlying operational performance excluding FX? Is the FX exposure hedged? Should it be? |
| “Lower than expected sales” | This is the variance restated, not explained. Which customers, products, regions, or channels underperformed? What driver changed? |
None of these translations require sophisticated analysis. They require the discipline to ask one more question instead of accepting the first answer.
Common Pitfalls
Accepting the first explanation as the root cause. “Revenue is down because sales were low” is a tautology. The first explanation is almost never the root cause — it is the symptom rephrased. Apply at least three iterations of “why” before concluding.
Attempting RCA on undecomposed variances. A total-level variance is too aggregated for meaningful diagnosis. Attempting to explain why “total costs are over budget” without first breaking costs into components, categories, and responsibility centres produces vague explanations that satisfy nobody.
Assigning all variances to external factors. External causes are real — commodity prices move, exchange rates shift, regulations change. But defaulting to “external factors” for every unfavourable variance is a way of avoiding accountability, not a way of understanding performance. Validate external attributions with evidence: market data, competitor benchmarks, regulatory records.
Treating RCA as a one-time exercise. The value of root cause analysis compounds over time. A single investigation is useful. A root cause library built over twelve months reveals patterns that no single-month analysis can show. Treat RCA as a repeatable process, not an occasional deep-dive.
Confusing root cause with blame. RCA is a learning process, not a performance review. If people associate root cause analysis with blame, they will resist providing honest explanations. The purpose is to understand what happened and decide what to do — not to assign fault. The planning root cause category is particularly important here: acknowledging that the plan was wrong is not blaming the planner. It is recognising that the analysis baseline needs updating.
Skipping the planning root cause category entirely. Every finance professional has experienced the situation where the budget was unrealistic from the start. Yet “flawed budget assumption” almost never appears in formal variance commentary. Including planning methodology as a root cause type means that some variances are reclassified from “performance failure” to “planning failure” — and the corrective action shifts from operational change to process improvement in how plans are built.
Industry Considerations
Manufacturing: RCA is well established for production variances — quality failures, yield losses, efficiency deviations — through Six Sigma and Lean methodologies. The gap in most manufacturing companies is not production-floor RCA but the connection between production root causes and financial variance explanations. The quality team knows why yield dropped. The finance team reports a material cost variance. The two analyses rarely meet.
Professional services: Utilisation and rate variances frequently have root causes in project scoping (unrealistic estimates), resource allocation (wrong skill mix), or commercial decisions (discounting to win work). These root causes cross functional boundaries and require collaboration between project management, sales, and finance.
Retail and distribution: Markdown and shrinkage variances benefit significantly from RCA. Markdown root causes often trace back to demand planning accuracy — overbuying leads to markdowns, and the root cause is the buying decision, not the markdown itself. Shrinkage root causes may be operational (warehouse processes), systemic (inventory management), or external (theft patterns).
Frequently Asked Questions
How is root cause analysis different from variance analysis? Variance analysis is the broader discipline of comparing plan to actual, decomposing the difference, and deciding what to do. Root cause analysis is the diagnostic step within variance analysis — specifically, the process of identifying why a variance occurred rather than simply reporting that it occurred. RCA operationalises Step 3 (Diagnose) of a structured variance analysis process.
How deep should we go with the five whys? Deep enough to reach an actionable cause. If the answer to “why” at a given level points to something the organisation can change, you have likely found the root cause. If it points to another question, continue. In practice, three to five iterations are typical for financial variances.
Do we need to perform RCA on every variance? No. Apply materiality thresholds. Perform RCA on variances that exceed defined absolute or relative thresholds, that recur across periods, or that affect strategic priorities. A £2,000 one-month variance in office supplies does not warrant five-whys analysis. A £200,000 cumulative material cost overrun does.
What if the root cause is genuinely external? Document it with evidence, quantify the impact, and decide whether to hedge, absorb, or adjust the plan. External root causes are legitimate — the discipline is validating them rather than assuming them. If “market conditions” is the root cause, specify which market condition, measure its impact, and assess whether competitors experienced the same effect.
Can RCA be done in a spreadsheet? Yes. The five-whys analysis is a thinking process, not a technology process. The root cause library is a structured list. Both can be maintained in standard spreadsheets. The value of RCA is in the structured thinking, not in the medium.
Related Reading
- Variance Analysis — A Practical Guide — the structural framework within which RCA operates
- Plan vs Actual — Beyond the Table — moving from BvA reporting to BvA analysis
- Driver-Based Performance Analysis — identifying the operational variables that explain financial variances
- Margin Erosion — Causes and Prevention — applying RCA specifically to margin decline
- Management Reporting Framework — structuring reports so that root cause findings reach decision-makers
Sources
- IMA (Institute of Management Accountants), “Statements on Management Accounting: Variance Analysis and Decomposition,” imanet.org
- Ventana Research, “Structured Variance Analysis and Financial Performance,” ventanaresearch.com
- Aberdeen Group, “Driver-Based Planning and Root Cause Discipline in Mid-Market Finance,” aberdeen.com
Martin Duben is the founder of OneTribe Advisory, where he works with mid-market finance leaders on performance analysis, management reporting, and financial governance. He writes about the diagnostic practices that separate variance reporting from variance understanding.