Fastenal’s First FP&A Team Starts Lean With AI

Fastenal did not have a dedicated financial planning and analysis team when Max Tunnicliff became chief financial officer last year. He is now building the company’s first one, and he told The Wall Street Journal that artificial intelligence should let him start with a smaller team than the same effort might have required in the past.

The timing makes the operating challenge concrete. Fastenal reported that third-quarter net sales rose 15.5% to $2.46 billion and net income increased 19.7% to $401.6 million. Growth with larger customers also pushed net accounts receivable up 19.9%, while trade working capital rose 8.1%. A company growing at that pace needs more forward-looking finance capacity, not merely faster production of historical reports.

Starting an FP&A function in the AI era creates a real design advantage. The team can build around automated data preparation, scenario generation and narrative drafting from the beginning. But “fewer people” is not the control objective. The objective is a planning system that gives management better decisions with explicit ownership, traceable assumptions and reliable review.

The judgment

• Implication: A new FP&A team can use AI to avoid building a large manual reporting factory. The benefit appears only if the function is designed around decisions, drivers and accountable review rather than automated spreadsheet assembly.
• What would change the conclusion: Persistent data reconciliation problems, rising forecast error, untraceable model outputs, confidentiality failures or heavy rework would weaken the case for a lean design. Rapid expansion into new markets or business models could also require more domain specialists even if transaction work is automated.
• Management action: Before finalizing the hiring plan, write a one-page FP&A charter that names the decisions supported, forecast cadence, system of record, assumption owners, human approval points, override documentation and measures of cycle time, forecast quality and capacity created.

Growth makes the missing function visible

The Journal reported that Tunnicliff wants the new team to give business leaders forward-looking information rather than concentrate on prior-period reporting. That is an organizational distinction, not a change in spreadsheet format.

Fastenal’s latest results show why it matters. The company said third-quarter operating margin improved to 21.5% from 20.7% as SG&A leverage more than offset pressure from customer mix. At the same time, larger customers tend to carry longer payment terms, contributing to a $267.9 million year-over-year increase in net receivables. Capital spending is also expected to rise to between $310 million and $330 million in 2026 as Fastenal invests in facilities, industrial vending equipment, IT and trucking.

Those facts create linked planning questions. How much of sales growth is volume, price or mix? What does a shift toward larger customers do to margin and cash conversion? Which capital projects remove constraints, and when? How much working capital should the company finance under different growth cases?

An FP&A function earns its place by connecting those drivers before management commits cash. AI may reduce the time required to assemble the evidence, but it does not decide which trade-offs Fastenal should accept.

Starting lean is not the same as cutting a mature team

A company creating a function from scratch has a cleaner opportunity than one automating a large inherited process. It does not have to reproduce every legacy report, spreadsheet handoff or monthly ritual. It can ask first which decisions require finance support and then build the minimum recurring process around them.

That sequencing matters. If a team automates a weak reporting process, it receives weak information faster. If it begins with decision rights and a common driver model, automation can prevent manual work from becoming institutionalized.

The lean design should therefore be measured by output and control quality, not simply analyst count. A small team that spends its time reconciling inconsistent source data is not efficient. It is merely understaffed. A small team that can explain forecast changes, test scenarios and challenge operating assumptions has created leverage.

Five layers should exist before the tool stack expands

1. Decision map. List the decisions the team must support: inventory, pricing, customer economics, working capital, capital spending, hiring and capacity. Each recurring output should have a named decision user and decision date.
2. Driver and definition library. Define volume, price, mix, customer tier, payment terms, backlog, productivity and capital-project stages once. AI cannot reconcile competing definitions that management has never resolved.
3. Scenario ownership. Assign every material assumption to an operating owner and a finance reviewer. The system should retain the source, date, range and rationale for each override rather than replacing judgment with an unexplained point estimate.
4. Review gates. Separate machine-assisted preparation from management approval. Sensitive data access, forecast publication, board narratives and material assumption changes should require named human review.
5. Performance measures. Track forecast bias and range calibration, cycle time, manual adjustments, reconciliation exceptions, repeat data defects and decisions influenced. Hours saved are useful, but they are not sufficient evidence of better planning.

This structure is consistent with the principle in Stop Pretending the Forecast Is One Number: the purpose of a forecast is to make uncertainty usable for decisions. AI can produce more scenarios cheaply. Finance still has to determine whether the scenarios are coherent and what management should do under each one.

AI can accelerate preparation without owning the forecast

Early use cases are likely to be practical rather than spectacular. AI can help classify variance explanations, summarize operating comments, compare actuals with prior assumptions, find inconsistencies across files, generate first drafts of management narratives and suggest scenario combinations for review.

Those activities are valuable because they consume analyst time without necessarily requiring analyst judgment at every step. They also create measurable baselines: minutes per variance, time from close to forecast, number of manual reconciliations and percentage of narratives rewritten after review.

AI should not become the unnamed owner of revenue, margin or cash assumptions. It lacks accountability for sales commitments, supplier behavior, operational constraints and risk appetite. It can surface patterns and draft explanations. Business leaders and finance must own the claim that those patterns should drive the plan.

AI Is Already in Finance makes the same distinction between adoption and governance. The new Fastenal team has an advantage because it can embed approved tools, confidential-data rules, review requirements and output accountability before informal practices harden.

Headcount should follow the bottleneck

A lean starting point does not establish the right long-term team size. That should emerge from the constraints the function cannot automate or redesign.

If the bottleneck is data extraction, fix integration. If it is inconsistent commercial assumptions, clarify ownership. If it is weak challenge of operating plans, hire analysts with business depth. If it is demand from a growing set of decision makers, add finance partners where the marginal decision value exceeds the cost.

This is also how management should evaluate the AI investment. The free AI & Automation ROI Calculator can compare implementation and operating cost with capacity, speed and error improvements. For an FP&A build, add measures for forecast cycle time, decision coverage and avoided rework rather than treating every saved hour as cash savings.

The strongest outcome is not an FP&A team with fewer people. It is a finance function that scales its judgment faster than its reporting workload. Fastenal’s blank sheet makes that possible, but only if the control design is built as deliberately as the technology.

Sources

• The Wall Street Journal, “Fastenal’s CFO Is Building a Financial Planning Team. AI Is Helping,” October 20, 2026.
• Fastenal, “Fastenal Company Reports 2026 Third Quarter Earnings,” October 14, 2026.
• U.S. Securities and Exchange Commission, Fastenal Form 8-K filing, October 14, 2026.

Fastenal’s reported financial results and Tunnicliff’s comments about the team are attributed to the company and the Journal. The function-design, control and staffing recommendations are Numbers & Judgment analysis.