Calculating true client acquisition cost when sales cycles run 18 months

A managing partner at a mid-sized strategy consultancy pursued a Fortune 500 supply chain transformation mandate for 22 months. Three partners, two engagement managers and a proposal team logged an estimated 900 hours across four RFP (request for proposal) rounds. The firm lost the first bid, was invited back, and won on the second attempt. Marketing's recorded spend on the account: about $8,000 in event sponsorship and one targeted LinkedIn campaign. Load in partner time and the acquisition costacquisition costCustomer 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 → lands closer to $280,000. A model that counts only the $8,000 is measuring a rounding error and calling it CAC.

Why standard CAC formulas fail in professional services

The textbook formula:

CAC = Total sales and marketing spend / New customers acquired

holds up for a 30-day cycle with a self-serve signup. Three things break it here:

1. The "salesperson" is a $400-to-$1,200-an-hour partner. Their time is the dominant cost and it never sits in the marketing budget.
2. Pursuits span fiscal years. An 18-month cycle attaches year-one cost to year-three revenue.
3. Most pursuits lose. Win rates on competitive RFPs in consulting and engineering commonly sit in the 20 to 40 percent band (estimate; varies by firm tier and deal size). The cost of the losses has to land somewhere.

The tooling reinforces the error. Segment, which sells the customer data plumbing many firms pipepipeAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.View full definition → into their CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →, is built to stitch anonymous web events onto a known contact: it will record the LinkedIn click, the gated report download and the webinar registration precisely. It has no field for 900 partner hours. Whatever the pipeline can see becomes the number in the board pack unless someone adds what it cannot.

Building the fully-loaded model

Step 1: define the pursuit cohort

Group every opportunity opened in a period, whatever the outcome. Picking out the winners after the fact is survivorship bias, and it understates cost by exactly the amount you most need to know.

Example cohort: 10 enterprise pursuits opened by a consulting firm's industrials practice in one year.

Step 2: capture every cost category

A reasonable internal cost rate for a partner is fully-loaded compensation and overhead divided by roughly 1,600 to 1,800 annual billable-equivalent hours (estimate; firms define this differently).

Step 3: allocate lost pursuits into the CAC pool

Most firms skip this step and it carries most of the signal. Pursue 10, win 3, and the cost of the 7 losses belongs to the 3 wins. Venture portfolio math runs the same way. Ben Horowitz's blog covers portfolio-level thinking accessibly, applied to a different asset class.

Step 4: compute fully-loaded CAC

Fully-loaded CAC = (Sum of all costs across the full pursuit cohort) / (Number of clients won from that cohort)

Worked example

Industrials practice, one year, 10 pursuits:

• Partner/senior time: 10 × 90 hours average × $500 internal rate = $450,000
• Proposal team costs: 10 × $12,000 average = $120,000
• Marketing-attributed spend: $60,000 (events, content, campaigns aimed at named accounts)
• BD travel: 10 × $8,000 average = $80,000

Total cohort cost = $710,000

Wins from cohort = 3

Fully-loaded CAC = $710,000 / 3 = $236,667 per client

The marketing-only version is $60,000 / 3 = $20,000, off by roughly 12x.

Dating the cohort against the revenue it produced

Accenture reports new bookings and revenue as separate quarterly figures because the two do not belong to the same period: work sold in one quarter converts to revenue over the quarters and years that follow. Cohort CAC needs the same discipline.

Anchor each pursuit to the date it opened, not the date it closed. A pursuit opened in FY23 stays in the FY23 cohort even when the engagement letter is signed in FY25. Then publish the cohort's maturity next to the number: "FY23 cohort, 70 percent resolved" warns the reader that the figure will move. A firm that closes a cohort at 12 months on an 18-month cycle flatters itself twice, excluding both the slow losses (understating cost) and the slow wins (which then belong to no cohort at all).

Open pursuits are the edge case that breaks the spreadsheet. At cohort close, some deals are neither won nor lost. Park their cost in a suspense line, restate quarterly, and treat a young cohort's CAC as an upper bound.

The lag also carries a financing cost nobody invoices. Cash spent on a pursuit sits dead for eighteen months before margin arrives; at an 8 percent cost of capitalcost of capitalThe blended rate a company pays to finance itself through debt and equity. It sets the minimum return an investment must clear to create value.View full definition → that adds roughly a tenth to the real cost of each win. Small next to the partner-hours correction, and the reason a pursuit sliding from 12 months to 30 hurts more than the extra hours suggest.

Matching CAC to lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition →

CAC means something only against the lifetime value figure the sibling lesson models for lumpy, irregular engagement patterns. Take that number as given and read the ratio: 3:1 or better is the directional band cited across B2B (business-to-business) services (borrowed from SaaS, so treat it as a check rather than a sector standard).

A $236,667 CAC client generating $1.5 million of lifetime margin across two expansion phases sits near 6:1. The same client who buys one $250,000 engagement and never returns sits under 1.2:1, and that pursuit model does not survive repetition however good the logo looks on the credentials page. Two things hide inside a healthy-looking ratio: expansion revenue that exists only in the forecast (you are underwriting today's pursuit decision with unsigned phases), and relationship concentration, where the departure of one partner turns paid-for acquisition cost into a sunk cost.

Payback period: the metric partners actually feel

CAC payback period = how long gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → from the client takes to repay the acquisition cost.

On 18-month cycles, payback commonly runs 24 to 36 months from first contact, because you are repaying costs incurred long before the first invoice. Enterprise SaaS benchmarks of 12 to 18 months (commonly cited estimate) are the wrong yardstick.

Infrastructure and design pursuits stretch it further. AECOM competes for multi-year public programmes where the first win is a place on a framework or panel rather than a funded scope. The bid cost is fully incurred; recovery waits on a call-off that may never be issued. If your practice bids that way, run two figures: cost per panel place and cost per funded instruction. Report only the first and you look efficient right up to the year the call-offs dry up.

Break CAC down by channel and practice

Blended CAC hides the signal. Cut it by:

• Practice area: an M&A advisory team chasing $2M+ mandates carries a structurally higher CAC than a compliance practice selling $150,000 engagements. Comparing the two is meaningless; comparing each against its own prior year is not.
• Origination channel: partner-network referral, competitive RFP, inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → thought leadership.
• Client type: new logo versus expansion inside an existing relationship. Expansion CAC runs far lower because access already exists.

Firms that measure this usually find inbound, content-originated pursuits carry a lower loaded CAC than competitive RFPs, largely because they skip rounds of unpaid proposal work. Widely observed across B2B services; the multiples are firm-specific.

Two mechanics decide whether the segmentation holds. Cross-practice pursuits need one owner: assign the cost to the practice that owns the client relationship and log the other practice's hours as a note, or the cost either doubles or disappears. And the moment fully-loaded CAC is used to judge partners rather than to price pursuits, pursuit hours stop being logged. A calendar audit surfaces materially more pursuit time than timesheets do, and that gap widens as soon as the number carries consequences. Keep the model with the people choosing which pursuits to run, and out of compensation.

A simple tracking snippet

For firms building this in a spreadsheet or lightweight database, the core logic per cohort:

total_cost = partner_hours * internal_rate 
           + proposal_team_cost 
           + marketing_spend 
           + bd_travel

fully_loaded_cac = total_cost / clients_won

ltv_cac_ratio = client_lifetime_margin / fully_loaded_cac

Run it per cohort, per practice, per year, and hold open-pursuit cost in a separate field so restatements are visible. Trends matter more than any single number.

Key takeaways

• Standard CAC formulas undercount by an order of magnitude here, because partner time and lost-pursuit cost sit outside the marketing budget and outside the attribution tooling.
• Build CAC at cohort level, anchored to the date each pursuit opened, and report how much of the cohort is resolved.
• Use internal cost rates (comp plus overhead over billable-equivalent hours), never billing rates, when valuing partner time.
• Where a win is a panel place rather than funded work, report cost per panel place and cost per instruction separately.
• Segment by practice and origination channel, and keep the number out of partner compensation, or the hour data degrades within a quarter.