Calculating true acquisition cost across asset management channels

A wholesaler flies to a regional broker-dealer conference, expenses a $4,000 booth, buys three dinners, and comes home with a verbal commitment from an advisor to allocate to the mid-cap fund. Two quarters later that commitment lands as $12 million of net new flow. Finance books the trip under travel and entertainment. Nobody books it against the $12 million. So when the CEO asks what a pound of new AUM cost to buy, the honest answer is that nobody in the building knows.

Most firms track gross sales on one side and marketing spend on the other and never divide one by the other, channel by channel. This lesson does that division twice: once per client won, once per pound of assets gathered. The two calculations rank your channels differently, and that gap is where distribution budgets are actually decided.

The two denominators

Fully loaded 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 → is every sales and marketing pound spent to win new business, divided by what the spend won. There are two denominators worth computing:

CAC per client     = Total acquisition spend / New clients acquired
Cost per £ of AUM  = Total acquisition spend / Net new assets gathered

Which "client" you count (an advisor, a home-office gatekeeper, a self-directed investor) follows from the channel mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → the distribution battlefield lesson sets out, so assume it here. The second denominator is the one this lesson insists on. Revenue arrives as basis points on an asset base, so a channel that wins many small clients cheaply can still be an expensive way to gather your next billion.

Loading the full cost stack

Four buckets, and the first is the one most often left out.

1. Distribution headcount

Wholesalers are the largest line. In the US an external wholesaler's fully loaded cost (base, commission, benefits, travel, entertainment) commonly runs into the mid six figures a year. Internal sales desks, sales management and the national accounts people who negotiate platform slots belong here too. If people are your acquisition engine, their cost is acquisition cost, not overhead.

2. Platform and shelf-space fees

To be sold through a US broker-dealer or wirehouse, managers frequently pay revenue sharing and sub-transfer-agency fees for shareholder servicing and record-keeping, plus negotiated access fees to sit on a preferred list.

One nuance that trips up most models: sub-TA fees scale with assets and keep running for as long as the money stays, which makes them partly a servicing cost. Load the one-off access payments into CAC, and treat the ongoing basis-point payments as a haircut to revenue. Do both and you count the same pound twice, once as cost of acquisition and once against lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition →.

In the UK and EU the money moves differently. The Retail Distribution Review (2012) and the MiFID II (Markets in Financial Instruments Directive II) inducement rules restrict the payments that dominate US shelf economics, which pushes the European cost stack toward salaries, brand and media. A European manager with a low shelf-fee bill is not more efficient than an American one, it is regulated differently, and its costs have simply moved into bucket 1.

3. Marketing and media

Conference sponsorships, advisor dinners, webinars, trade press, paid search and LinkedIn. LinkedIn sits at the expensive end of the major ad platforms, and financial audiences sit at the expensive end of LinkedIn: clicks priced in pounds, not pennies. Defensible when one click can start an institutional conversation, indefensible when the average account you are buying is a £3,000 ISA.

4. Data and content infrastructure

CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →, marketing automation, the salaries of the people writing commentary and fund one-pagers, and the data subscriptions used for targeting. Allocate a share of this to each channel rather than parking it in a central pot nobody owns.

Two channels, two sets of economics

Take the wholesaler-led advisor channel from the hook and set it against a direct, self-serve channel where buyers research and allocate with almost no human contact. Same fund, same fee, same firm. Figures below are illustrative teaching numbers, not benchmarks.

Channel A: wholesaler-led advisor channel (one quarter)

New advisor relationships that produced net inflows: 16. Average allocation: $8 million.

CAC (Channel A) = 320,000 / 16 = $20,000 per advisor relationship

Channel B: direct self-serve channel (same quarter)

New accounts won: 40. Average ticket: $1.5 million.

CAC (Channel B) = 80,000 / 40 = $2,000 per account

The blended number, and why it misleads

Blended CAC = 400,000 / 56 = ~$7,143 per new client

Useful for the CFO, dangerous for the marketer. It averages two businesses with a tenfold difference in unit cost and hides both. Compute per channel first, then blend, never the reverse.

Cost per pound of AUM

Now run the same spend against assets gathered.

Channel A: 320,000 / 128,000,000 = 0.0025 = 25 bps of assets gathered
Channel B:  80,000 /  60,000,000 = 0.0013 = 13 bps of assets gathered

Per client, Channel B looks ten times cheaper. Per pound of AUM it is roughly twice as cheap. That collapse in the gap is the single most useful output of this calculation, and it is why firms that manage distribution on cost per client keep starving the channel that actually moves assets.

Turn it into payback. At a 50 basis point management fee, a 25 bps acquisition cost is repaid by about six months of gross revenue on the assets it bought; Channel B needs about three. Neither figure allows for portfolio management, operations, custody or compliance, so the true break-even sits well beyond that.

Then change the fee. Several core Vanguard index funds charge under 10 basis points, and Vanguard's structure (owned by its funds, run close to cost) exists to keep them there. At 8 bps, a 25 bps acquisition cost takes more than three years of gross revenue to repay. Field-sales distribution is arithmetically impossible at that price point, which is why cheap passive products are sold through brand, default platform status and adviser inertia rather than wholesalers with expense accounts. Your fee level decides which acquisition channels you are even allowed to consider.

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 →: the hard part

The tables assume you know which channel produced which win. In practice an advisor reads three months of commentary, attends a conference, then allocates a fortnight after a wholesaler visit. Who gets the $320,000?

• First-touch: credit whoever engaged the buyer first. Simple, and it flatters awareness spend.
• Last-touch: credit the final interaction. Flatters the wholesaler, who sometimes only signs the paperwork.
• Multi-touch: split credit across interactions. More honest, considerably more work.

Pick one and hold it steady for at least four quarters. Switching models mid-year produces a "CAC improvement" that is pure accounting. For the underlying logic, the Google Analytics Help documentation on attribution models is a free, clear primer, and it transfers even if you never point GA at fund flows.

CAC alone ranks nothing

A 25 bps acquisition cost is cheap or ruinous depending on how long the assets stay and how the balance behaves, which is the value side the lifetime value lesson models. Pair the two and watch the ratio; the 3:1 rule of thumb borrowed from subscription marketing is a floor, not a target, and it was never calibrated for a revenue line that falls 20% when markets do.

Nutmeg is the cautionary arithmetic. At its 2021 acquisition by JP Morgan it managed roughly £3.5 billion for around 140,000 clients, so an average pot near £25,000. At its published fees (roughly 0.45% to 0.75% depending on portfolio type), that pot generates somewhere between £110 and £190 of gross annual revenue. A £300 blended acquisition cost, modest by consumer fintech standards, needs two to three years of gross fees before it clears, and Nutmeg reported losses in every year of its independent life. Direct-to-consumer asset gathering does not fail on CAC per client. It fails on how few pounds each cheap client brings.

Sanity-checking against the sector

You will rarely find published, apples-to-apples CAC benchmarks in this industry, because firms guard distribution economics and definitions vary. Treat any precise "industry average CAC" as a guess.

What you can benchmark:

• Your own trend, per channel, quarter over quarter.
• Channel against channel inside your own firm, on one consistent allocation rule.
• Acquisition cost as a percentage of first-year revenue won. Channel A's $20,000 against $40,000 of first-year fees is 50%.
• Basis points of gathered assets, which is the only figure that survives comparison across a $1.5 million ticket and a $50 million mandate.

The extreme counter-example is worth keeping in view. Zerodha has said publicly that it spends essentially nothing on customer acquisition, growing through referrals and its free Varsity education material to over ten million clients before extending into fund management. That is not a tactic most managers can copy, but it sets the floor: if a competitor's CAC is close to zero, your 25 bps has to buy something theirs does not.

For structural context on US fund fees, the SEC's investor education material on mutual fund fees is a free, authoritative reference.

Common mistakes that wreck the number

• Excluding wholesaler salaries. The most common omission, and the one that makes advisor distribution look free.
• Counting gross sales instead of net. A client who allocates and redeems within two quarters was never acquired, only rented.
• Counting internal switches as new business. Assets moving from your own fund into your new strategy cost real wholesaler time and add nothing. Net them out or your cost per pound of AUM is fiction.
• Ignoring the lag. Spend in Q1 often lands as flow in Q3. Match cost periods to your observed conversion window or CAC will swing 40% between quarters for no operational reason.
• Blending too early. The average of two different businesses describes neither.

Key Takeaways

• Run both calculations. Cost per client and cost per pound of AUM rank channels differently; in the worked example the tenfold per-client gap shrinks to roughly twofold per pound.
• Load the whole stack: distribution headcount, platform access fees, events, media and data infrastructure. Headcount is the biggest line and the most frequently omitted.
• Split one-off access fees from ongoing sub-TA payments. The first is acquisition cost, the second is a revenue haircut. Counting both as CAC double-charges the same pound.
• Your fee level constrains your channels. At single-digit basis points, wholesaler-led acquisition cannot pay back inside a plausible holding period.
• Fix one attribution model and keep it for a year. Changing it produces improvements that exist only on the spreadsheet.
• Distrust published industry averages. Benchmark your own trend, and acquisition cost as a share of first-year revenue won.