Performance management: linking strategy, decisions and results

Performance management is the set of processes an organization uses to turn its strategy into decisions, track whether those decisions are working and change course when they are not. Management accountants sit at the center of it. They build the measures, run the models behind major decisions, prepare the budgets and forecasts, and explain the gap between what was planned and what happened. In the United States, much of this material appears in the CMA examination and in the Business Analysis and Reporting (BAR) discipline section of the CPA examination; in Canada it forms the CPA Performance Management elective. This course takes an integrated view: it treats measurement, decision analysis, pricing, capital investment and risk as parts of one system rather than separate techniques.

The course builds on earlier work rather than repeating it. ACCT-222 covered cost systems, activity-based costing, standard costing and variances. ACCT-321 covered management control systems, responsibility centers, return on investment, residual income and economic value added, transfer pricing, the balanced scorecard and incentive design. FIN-201 covered the mechanics of discounted cash flow, and BUS-450 covered strategic analysis. Here the focus is on using those tools to make and evaluate decisions: which products to make, what to charge, which investments to fund, how to plan under uncertainty, and how to tell whether the whole system is moving the organization toward its goals.

The running example

Every lesson follows one invented company. Northgate Equipment Inc. is a privately held manufacturer based in Minneapolis, Minnesota, with about 640 employees and revenue of 148 million dollars in fiscal 2026. It has three divisions. Farm Systems makes grain augers, conveyors and bin sweeps sold through about 400 independent dealers across the northern Plains and the Canadian Prairies. Industrial Drives makes gearboxes and drive units, partly for Farm Systems and partly for outside customers. Parts and Service sells replacement parts and service contracts. Around 30 percent of sales are invoiced in Canadian dollars.

The chief financial officer, Anita Sandhu, has been asked by the board to raise return on capital employed (ROCE) from just over 12 percent to 15 percent within two years, without taking on new debt. The financial planning and analysis (FP&A) team is led by Marcus Lavoie, and its newest member is Grace Odum, a recent accounting graduate. Over the course, the team will work through the decisions that could close the gap: a bottleneck in the paint line, a make-or-buy question on gearbox housings, a new connected-sensor product called GrainSense, a robotic welding cell, a constrained capital budget and a volatile steel market. The final lesson brings the pieces together.

What performance management involves

Performance management can be described as a cycle with five linked activities. First, the organization sets direction: its mission, its strategy and the few objectives that matter most. Second, it translates that direction into plans, targets and resource allocations, through budgets, capital plans and operating plans. Third, managers make decisions within those plans, such as what to produce, what to charge and what to buy. Fourth, the organization measures results, both financial and non-financial, and compares them with the plan and with what it learns about its environment. Fifth, it learns and adapts, revising decisions, plans or even strategy. Then the cycle repeats.

Two features of this cycle deserve emphasis. The first is that measurement is only one part of it. Organizations that equate performance management with a dashboard of key performance indicators tend to measure a great deal and change very little. The second is the distinction between single-loop and double-loop learning, terms introduced by Chris Argyris. Single-loop learning corrects deviations from the plan: if costs are over budget, find out why and bring them back. Double-loop learning asks whether the plan, the targets or the strategy were right in the first place. A company whose margins are falling because the market is shifting needs double-loop learning; cutting costs to hit an outdated budget will not save it.

From strategy to measures and decisions

Strategy is a set of choices about where to compete and how to win. BUS-450 discussed the main frameworks: industry analysis, resources and capabilities, and generic strategies such as cost leadership and differentiation. For performance management, the key step is to identify the critical success factors that the strategy depends on, and then the measures and decisions that drive those factors.

Northgate's strategy, approved by its board in 2025, is to move from selling equipment to selling uptime. Farmers lose money when an auger or conveyor breaks during harvest. Northgate believes it can earn higher and steadier margins by combining robust equipment with fast parts supply, service contracts and, eventually, connected sensors that warn of problems before they happen. The critical success factors that follow include product reliability, on-time delivery to dealers during the spring and harvest peaks, a strong dealer network, growth in service revenue and cost competitiveness against a low-cost importer that entered the market in 2024.

Each critical success factor points to measures, which lesson 2 develops, and to decisions, which the rest of the course analyses. Reliability points to warranty cost, field failure rates and the design choices made through target costing. On-time delivery points to the paint line bottleneck and the product mix. Cost competitiveness points to make-or-buy decisions and automation. Service growth points to pricing and to the GrainSense investment. The value of making these links explicit is that it lets the accountant ask of any proposed decision or measure: which success factor does this serve, and how would we know if it worked?

Value driver trees

A value driver tree breaks a financial objective into the operational drivers that determine it. At the top sits the objective, here ROCE. Beneath it are its two components, operating margin and capital turnover, because ROCE equals operating profit divided by revenue, multiplied by revenue divided by capital employed. Each of those breaks down further: margin into price, volume, material costs, conversion costs and overheads; turnover into fixed assets, inventory, receivables and payables. At the bottom are drivers that managers can actually influence, such as steel cost per metric ton, inventory days or the dealer payment terms.

Grace builds Northgate's value driver tree as a small program, so that the team can test how much each driver matters. She uses the fiscal 2026 figures and asks a simple question: if each driver improved by a plausible amount, how much would ROCE rise?

# Northgate value driver tree: return on capital employed (ROCE) built up from its drivers.
# All figures are invented, in millions of dollars, for fiscal 2026.
base = dict(revenue=148.0, material_pct=0.41, conversion_pct=0.30, conv_variable=0.60,
            sga=31.12, fixed_assets=52.0, inv_days=95, rec_days=58, pay_days=45)

def roce(d):
    material = d['revenue'] * d['material_pct']
    conversion = d['revenue'] * d['conversion_pct']
    profit = d['revenue'] - material - conversion - d['sga']
    cogs = material + conversion
    inventory = cogs * d['inv_days'] / 365
    receivables = d['revenue'] * d['rec_days'] / 365
    payables = material * d['pay_days'] / 365
    capital = d['fixed_assets'] + inventory + receivables - payables
    return profit, capital, profit / capital

p, c, r = roce(base)
print(f"Operating profit {p:6.2f}  margin {p / base['revenue']:.1%}")
print(f"Capital employed {c:6.2f}  turnover {base['revenue'] / c:.2f}")
print(f"ROCE {r:.2%}  (target 15.00%)")
print()

def scenario(name, **chg):
    d = dict(base)
    for k, v in chg.items():
        d[k] = v(d) if callable(v) else v
    _, _, r2 = roce(d)
    return name, (r2 - r) * 100

def volume_up(d):
    # 1% more units: revenue and variable costs rise, fixed conversion and SG&A do not
    fixed_conv = base['revenue'] * base['conversion_pct'] * (1 - base['conv_variable'])
    var_conv = base['revenue'] * base['conversion_pct'] * base['conv_variable'] * 1.01
    d['revenue'] = base['revenue'] * 1.01
    return (fixed_conv + var_conv) / d['revenue']

tests = [
    scenario('Price +1%', revenue=base['revenue'] * 1.01,
             material_pct=base['material_pct'] / 1.01, conversion_pct=base['conversion_pct'] / 1.01),
    scenario('Volume +1%', conversion_pct=volume_up),
    scenario('Material cost -1%', material_pct=base['material_pct'] * 0.99),
    scenario('Conversion cost -1%', conversion_pct=base['conversion_pct'] * 0.99),
    scenario('SG&A -1%', sga=base['sga'] * 0.99),
    scenario('Inventory -10 days', inv_days=base['inv_days'] - 10),
    scenario('Receivables -5 days', rec_days=base['rec_days'] - 5),
]
print('Change in ROCE (percentage points) for each improvement:')
for name, d in sorted(tests, key=lambda t: -t[1]):
    print(f"  {name:22s} {d:+.2f}")

Operating profit  11.80  margin 8.0%
Capital employed  95.39  turnover 1.55
ROCE 12.37%  (target 15.00%)

Change in ROCE (percentage points) for each improvement:
  Price +1%              +1.52
  Material cost -1%      +0.65
  Volume +1%             +0.58
  Conversion cost -1%    +0.48
  Inventory -10 days     +0.38
  SG&A -1%               +0.33
  Receivables -5 days    +0.27

The model reproduces the reported figures: operating profit of 11.8 million on capital employed of 95.4 million gives a ROCE of 12.37 percent. The sensitivity results are revealing. A 1 percent price increase, holding volumes, adds 1.52 percentage points to ROCE, more than twice as much as any other single change of similar size. That is because a price increase falls straight to operating profit, while a volume increase brings extra variable costs with it. Material cost matters next, because materials are 41 percent of revenue. The working capital levers, cutting inventory by ten days or receivables by five days, add less but require no change to customers or suppliers' prices and release cash as well.

The model also shows why the board's target is hard. To close a 2.6-point gap, Northgate would need something like a 1 percent price increase plus a 1 percent material saving plus a meaningful cut in inventory, all at once, and sustained. No single initiative will do it. That is why the course treats performance management as a portfolio of decisions whose combined effect has to be estimated and tracked, and why the capstone returns to this tree to bridge from 12.4 to 15 percent.

Levels of performance management

Performance management operates at several levels, and each needs different information. At the strategic level, the board and senior executives need a small number of measures that show whether the strategy is working, plus the analysis behind major commitments such as acquisitions or large capital projects. At the business unit or divisional level, managers need measures that reflect what they control, which ACCT-321 discussed under responsibility accounting. At the operational level, supervisors need timely, specific information: the paint line's output per shift, the scrap rate on a welding station, the backlog of dealer orders.

A common failure is to push strategic measures down to operational levels where nobody can act on them, or to roll operational measures up to the board where they create noise. A useful test is whether the person who receives a measure can name at least one action they would take if it moved. Grace finds that Northgate's monthly management pack, inherited from a previous controller, contains 64 measures, sent to the same 30 people. Few of the recipients can explain what they would do with most of them. Lesson 2 returns to how she redesigns it.

Financial and non-financial measures

Financial measures such as operating profit, ROCE and cash flow have obvious strengths. They are comparable across units, they connect directly to what owners care about and they are audited or at least reconciled to the books. Their weaknesses are equally familiar. They report outcomes after the fact, they can be improved in the short term at the expense of the long term, for example by cutting maintenance or training, and they say little about why results occurred.

Non-financial measures such as on-time delivery, defect rates and customer satisfaction fill some of those gaps. They are often leading indicators: a fall in dealer satisfaction this quarter predicts lower orders next season. They are usually closer to what operational managers control. Their weaknesses are that they are harder to compare, easier to define loosely and sometimes only loosely connected to value. A balanced set, the idea behind the balanced scorecard covered in ACCT-321, combines both kinds and makes the hypothesized links between them explicit.

Where the management accountant adds value

The management accountant's role in performance management has shifted over the past two decades. Routine reporting is now largely automated, and analytics tools, discussed in ACCT-341, make it possible for operational managers to explore data themselves. What remains distinctive is judgment: framing decisions correctly, identifying relevant costs and benefits, modeling uncertainty honestly, challenging optimistic assumptions and explaining results in terms that lead to action.

The Institute of Management Accountants' Statement of Ethical Professional Practice sets out the duties of competence, confidentiality, integrity and credibility; the AICPA Code of Professional Conduct applies similar principles to CPAs in business. In this context, integrity and credibility mean resisting pressure to shade an analysis toward a decision someone has already made, and disclosing the limits of a model rather than presenting a single precise number as if it were certain.

Throughout the course, every major analysis will end with the same three questions. What does the number say? How confident are we in it? What would change the decision? Those questions are the core of the management accountant's contribution.