Calculating true customer acquisition cost for capital equipment

Two numbers, same machine. The marketing dashboard says you acquired that customer for $26,000. The channel P&L says the machine left the factory at 22 points off list, and the distributor who found the plant manager, ran the demo and wrote the quote kept most of that discount. Both numbers are arithmetically defensible. Only one of them survives a conversation with the CFO about whether to open a direct territory.

This lesson builds the second number: a CACCACCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → you can defend for a machine that sells a few hundred times a year, with dealer margin, show spend, field application engineering and a multi-year attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition → window loaded in.

What belongs inside a capital-equipment CAC

Textbook CAC arithmetic assumes fast, cheap, single-channel deals, and the funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition →-metrics lesson has already shown why that logic does not transfer here. The practical consequence is a costing problem: four large pools of acquisition cost sit outside the line item marked "marketing".

• Dealer and distributor margin. When a machine ships through a channel partner at 20 to 25 points off list, part of that discount buys demand creationdemand creationCreating and stimulating demand for your offer, often upstream of the buying process to generate interest and awareness before prospects are ready to buy.View full definition →: local prospecting, the demo machine on the dealer floor, the quotation, the tooling package proposal. Haas Automation sells its machining centres exclusively through Haas Factory Outlets, so almost all of that field cost sits with the channel rather than in a corporate budget. Atlas Copco covers compressors and industrial tools through a mix of direct sales and distribution depending on product and territory, which means the same unit can carry two different cost structures in two countries. Comparing those territories on marketing spend alone tells you nothing.
• Field application engineering. Hours spent on a proof-of-process trial (test cuts, fixture design, cycle-time studies, sample parts returned to the prospect's quality lab) are acquisition cost, because they happen before the purchase order exists. On a spec-heavy machine, one trial absorbs a week of engineer time plus machine hours and consumed tooling.
• Show cost spread over the right period. IMTS runs every two years at McCormick Place in Chicago; Hannover Messe runs annually. A calendar-year CAC that charges the full biennial stand to one year alternates high and low for reasons that have nothing to do with efficiency. Amortise a biennial show across the 24 months of pipelinepipelineAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.View full definition → it feeds.
• Time. Spend incurred in Q1 2025 closes revenue in Q3 2026. Dividing this quarter's spend by this quarter's wins compares two unrelated populations.

The corrected formula: cohort-based blended CAC

Allocate spend to the cohort of customers who closed in a given period, using the spend incurred across their actual buying window.

Blended CAC (cohort) = (Marketing spend + selling spend attributable to the cohort) ÷ Number of new customers in that cohort

Attributable spend includes paid mediapaid mediaVisitors arriving via paid ads or sponsored placements, where you pay a platform to display your message rather than earning visits organically.View full definition → and content tied to campaigns that touched the cohort, show costs prorated to converting leads, application engineering hours logged against those opportunities, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → (Customer Relationship Management) and automation platform costs allocated proportionally, and the demand-creation share of channel margin.

The denominator counts first-time accounts only. A second machine sold into an existing plant is expansion and belongs to the installed-base metrics, not here. A single purchase order covering four identical machines is one customer, so track CAC per customer and cost per unit shipped as separate lines.

Worked example

A machine tool builder shipping roughly 300 units a year closes 8 new accounts in Q3 2026, sourced over the prior 9 months.

Marketing-controllable CAC = $210,000 ÷ 8 = $26,250 per customer

Now load the channel. Six of the eight closed through distributors. On a $450,000 list price at 22 points off, roughly $99,000 per unit sits with the dealer. Split it: allocate 40% to demand creation and 60% to fulfilment, installation, financing and local service readiness. That adds 6 × $39,600 = $237,600, and fully loaded CAC becomes $447,600 ÷ 8 = $55,950, more than double the dashboard figure.

Keep both numbers and say which is which. The controllable figure manages campaigns. The loaded figure decides whether a direct territory, a second demo machine or a richer dealer program is worth funding. The 40/60 split is an assumption, so write it down, agree it with finance once, and stop relitigating it every quarter.

Avoiding double-counting: the attribution rule

The most common error: charging the full show budget against every lead from that show, then charging the full ABM budget against the same accounts because they were retargeted.

Fix: use one attribution model consistently, and document it.

Two defensible approaches for industrial buying committees (typically 3 to 7 stakeholders per B2B purchase, per Gartner's B2B buying research):

• First-touch weighted: 40% of credit to the channel that created initial awareness, the remainder split across nurture and sales touches.
• Multi-touch linear: equal credit across every logged touchpoint on the account.

Set the lookback window from the 90th percentile of your closed-won cycle length, not the average. If the average cycle is 9 months but a fifth of deals take 20, a 12-month window silently deletes the show that started your largest orders and charges those customers to campaigns they never saw. Switching models or windows between quarters makes the trend line meaningless.

CAC benchmarks in manufacturing (estimates)

Sector CAC benchmarks are scarce and vary by equipment category, so treat these as directional as of 2025-2026:

• CAC-to-first-order-value: commonly cited as 5% to 15% for capital equipment above $100,000, marketing-controllable basis. Loaded for channel margin, 20% to 30% is normal and not a sign of failure.
• Sales cycle: 6 to 12 months for mid-complexity equipment, 18 months and beyond for custom process lines.
• Show cost per qualified lead: often estimated at $300 to $900 for major US industrial shows. A top-tier stand with machines running and rigged is a six-figure commitment before travel, which is why the prorating rule matters more than the per-lead number.

Reading CAC against lifetime value, and against noise

CAC on its own tells you little; judge it against the multi-stream lifetime value the LTV lesson builds, where the machine is the smallest of several revenue lines. A 3:1 lifetime-value-to-CAC ratio is the widely cited floor across B2B, and capital equipment usually clears it comfortably once aftermarket revenue is counted.

The bigger risk at this volume is statistical, not strategic. With 8 closed customers in a cohort, one anomalous deal moves CAC 20% or more. A single prospect who demanded three proof-of-process trials across two plants can add $40,000 of engineering time to one quarter. Report rolling four-quarter cohorts, show the median alongside the mean, and never let a board deck compare a quarter with an IMTS in it to a quarter without one.

The second-order effect is organisational. Once application engineering hours enter CAC, the number stops being marketing's private metric: those people usually report to engineering or operations, and someone has to agree an hourly transfer price. Skip that conversation and you will conclude that spec-heavy leads are cheap when they are the most expensive kind you generate.

Practical steps to fix your CAC tracking

1. Tag every lead source at first contact, including badge scans, gated downloads and distributor referrals. Untagged leads make cohort analysis impossible, and badge IDs need to stay alive across two show cycles.
2. Have field application engineers log hours against opportunities. This is usually the largest hidden component of industrial CAC.
3. Agree the demand-creation share of dealer margin with finance and freeze it for at least a year.
4. Set the lookback window from the 90th percentile cycle length and review it annually rather than quarterly.
5. Segment CAC by equipment category and deal size. A blended figure across $20,000 parts orders and $2 million lines hides more than it reveals.
6. Resolve anonymous web sessions to accounts before the CRM record exists. A customer data platform such as Twilio Segment (a vendor that sells exactly this identity-resolution tooling) is one way; a disciplined UTM and IP-to-account process is another.

🎬 [VIDEO: "B2B Marketing Attribution Explained" - youtube.com/results?search_query=b2b+marketing+attribution+explained - search this term for current explainer videos on multi-touch attribution models applicable to long B2B sales cycles]

Key takeaways

• Publish two CAC figures: marketing-controllable, and fully loaded with the demand-creation share of dealer margin. The loaded number is usually more than double the dashboard number.
• Match spend to the cohort that closed, not to the calendar month it was incurred, and amortise biennial shows like IMTS across the 24 months of pipeline they feed.
• Field application engineering hours before the PO are acquisition cost, and pricing them requires an agreement with whoever owns those engineers.
• Set the attribution lookback from the 90th percentile cycle length; an average-based window deletes the touchpoints that started your longest, largest deals.
• At a few hundred units a year, one unusual deal swings quarterly CAC by a fifth, so report rolling four-quarter cohorts with a median beside the mean.