How to Align Ecommerce, Finance, and Marketing on Growth

A practical system for building one commercial plan that finance, marketing, and ecommerce all make decisions from, instead of three separate scorecards.

Updated on: 2026-09-08

The clearest sign of misalignment I see in a growth audit is three people describing the same month in three incompatible ways. Marketing shows a 4.2 blended ROAS and calls it a good month. Finance shows revenue up 18% but cash down and gross margin flat. Ecommerce shows two SKUs stocked out and a return rate creeping past 20%. Nobody is lying. They are all reading a different number and calling it the truth.

The fix is not another dashboard. It is one commercial plan that all three functions make decisions from. Ecommerce owns the customer and the trading system. Marketing creates and captures demand. Finance defines the economic constraints and the cash. When those three optimize separately, you get busy teams and flat contribution. When they run off one model, spend decisions start to make sense.

Here is the chain the plan should follow, in order:

Growth target → demand and conversion assumptions → order and customer economics → marketing investment → inventory, fulfillment, and cash requirements → owners and review cadence.

Everything below is how you build and run that.

Why one shared target beats three departmental KPIs

Most brands set revenue as the company goal because it is easy to see and politically comfortable. It is also the target most likely to hide a problem. Revenue can climb while cash tightens and margin erodes, and the people watching revenue will not notice until the bank balance forces the conversation.

A shared objective should combine at least six things:

• Net sales
• Contribution margin (in euros, not just percent)
• New customers
• Repeat-purchase contribution
• Cash position or cash conversion
• Inventory and fulfillment capacity

McKinsey's advice to establish a quantifiable "North Star" and then connect growth levers with cost levers lands here. A workable version reads like this:

Generate €X in net sales and €Y in contribution margin from Z new customers by end of Q4, while holding a minimum cash buffer and a defined service level.

That single sentence gives marketing, finance, and ecommerce the same decision target. "Increase ROAS" does not.

This is the pattern the approach behind Miguel Casteleiro's advisory is built around: one commercial target, clear responsibilities, and a weekly decision rhythm, rather than each function reporting local wins while the company misses plan.

Define contribution margin before you argue about ad spend

You cannot set an allowable acquisition cost until you agree what a customer is worth. Contribution margin is revenue minus variable costs, and the ratio is contribution divided by revenue times 100. Simple formula, endless arguments about what counts.

For ecommerce, the model should normally strip out net product revenue, discounts, refunds and returns, product and landed cost, packaging, pick-and-pack, outbound shipping subsidy, payment processing, marketplace commissions, and variable service or sales commissions.

The single most useful thing you can do is name two layers and stop mixing them:

There is no universal rule about whether marketing sits inside contribution margin. Some finance teams report contribution before marketing. Operators often add a second layer after variable acquisition spend. Either works. What kills you is subtracting CAC twice because two teams defined the layer differently and never reconciled it. Name the layer, document the policy, and make finance sign it off.

Cash is a constraint, not a year-end report

I have watched a brand grow revenue 30% in a quarter and nearly run out of cash doing it. Inventory was bought before it sold. Marketplace settlements lagged. Ad spend went up front. Returns came back weeks later. Every one of those is normal, and together they created a squeeze nobody modeled.

A short-term cash forecast belongs next to the growth plan, not in a separate finance file. A 13-week forecast split into money-in and money-out is enough for most brands at this stage. Build a cash gate into the plan:

• Increase spend only if the next 13 weeks stay above the minimum buffer.
• Delay acquisition if inventory, payables, or refund obligations create a squeeze.
• Model downside, base, and upside cases instead of trusting one forecast.

The build: eight steps that actually stick

1. Set the decision mandate

Appoint one accountable owner before you build anything. Founder, commercial director, head of growth, or a fractional head of growth. The owner does not do every task. The job is to stop marketing, finance, and ecommerce from running separate plans.

Write a one-page decision charter: shared objective, metric definitions, planning horizon (usually 90 days for execution against a 12-month financial view), decision rights, meeting cadence, escalation rules. If it does not fit on a page, it will not get used.

2. Reconcile the numbers before setting targets

Do not start from desired ROAS. Start by reconciling the last three to twelve months across every system that touches money.

The goal is not to force every system to match. It is to explain the gaps and decide which number is authoritative for each decision.

One measurement point matters here. Platform-reported ROAS is an attribution number, not proof the ad caused the sale. Google's guidance on incrementality testing defines incremental ROAS as incremental revenue divided by media spend, measured with a treatment-and-control design. Keep that distinction in mind before you reallocate budget on attributed numbers.

3. Build a shared unit-economics model

The model should run at SKU, order, customer, channel, cohort, and total level. Core calculations:

Order contribution before acquisition:
Net order revenue − product cost − fulfillment − shipping subsidy
− payment fees − expected return cost

Allowable CAC:
Expected customer contribution over payback period − required buffer

Observed channel CAC:
Channel acquisition spend ÷ new customers acquired

Observed channel CAC is not the same as incremental CAC. If a channel gets credit for customers who would have bought anyway, it looks cheaper than it is.

Segment by new versus returning, SKU, first-order product, channel, market, offer, and cohort month. McKinsey's recommendation for cost visibility at customer, item, channel, and transaction level (including returns, markdowns, and payment fees) is worth taking seriously here.

The cost-policy calls that must be explicit and finance-approved: whether customer-paid shipping is revenue or a cost offset, whether returns hit revenue or sit as variable cost, whether agency retainers are overhead or acquisition, whether brand media enters CAC, and whether CAC is judged on first-order or a defined payback horizon.

4. Convert the target into a demand plan

Now translate the financial target into operational drivers:

Target orders = target net sales ÷ expected AOV
Required sessions = target orders ÷ expected conversion rate
Required new customers = target orders − expected returning orders
Required acquisition budget = required new customers × allowable CAC

Then test feasibility against inventory, supplier lead times, warehouse and service capacity, cash, and seasonality. This is where ecommerce becomes the bridge. Finance can approve a budget marketing can spend, but ecommerce has to confirm the business can fulfill and support the demand it creates.

Run three cases, not one:

5. Give every channel a role and an economic rule

Stop treating every channel as a direct-response machine. Assign roles: demand creation (paid social, creators, PR), demand capture (search, shopping, marketplaces), conversion support (CRO, merchandising), retention (email, SMS, loyalty), and measurement (experiments, holdouts).

Marketing keeps its channel expertise. The plan just stops the ad platforms from grading their own homework.

6. Build a three-level measurement hierarchy

One dashboard cannot answer every question. Use three levels.

Operational (frequent): spend, orders, net sales, new customers, conversion, CAC, contribution, stock, refunds, cash.

Diagnostic (to explain movement): SKU profit, funnel conversion, new vs returning, cohort retention, discount effects, creative fatigue, fulfillment economics. Shopify's customer cohort report groups customers by first-order date, which is the cleanest way to check whether a high-CAC source pays back later rather than judging it on first-order ROAS alone.

Causal (for budget decisions): geo holdouts, conversion-lift tests, matched markets, and marketing mix modeling for larger multi-channel brands. Google's Conversion Lift documentation covers user and geo designs, and geo studies can run on first-party finance data without relying on cookies. Nielsen positions marketing mix modeling as a way to assess investment impact and simulate future allocation.

The rule: attribution for operational tuning, cohorts for customer economics, incrementality or modeling for budget moves. Do not reallocate the whole budget off a last-click number.

7. Install a cross-functional cadence

Weekly growth review with ecommerce, finance, marketing, merchandising, operations, and analytics in the room. Review actual versus plan across net sales, new vs returning, contribution, CAC, inventory, cash forecast, and live tests. Every meeting ends with a decision, an owner, a deadline, and the metric that will verify it. No decision, no point meeting.

Monthly business review covers forecast versus budget, contribution by product and channel, cohorts, working capital, promotions, and assumption changes.

Quarterly reset revisits whether the North Star, the allowable CAC, and the channel roles still hold.

8. Make decision rights explicit

Input can be shared. Accountability is singular.

Where these plans usually break

What I would do first

If you have one week, reconcile the numbers (step 2) and write the contribution-margin policy (step 3). Almost every alignment problem I have unpicked traces back to teams using different definitions of revenue, CAC, and profit without knowing it. You cannot set a shared target on numbers people quietly disagree about.

After that, run one weekly review with a real decision log. The cadence exposes the gaps faster than any audit.

How to know it is working

Alignment is not "everyone can recite the target." It is behavioral. Interview each function after four to six weeks:

• What is the shared growth target?
• Which contribution-margin definition are we using?
• What would make you increase or reduce spend?
• Who owns the next decision?
• Which dashboard number do you not trust, and why?

You have alignment when finance can explain how spend moves contribution and cash, marketing can explain how activity moves new-customer economics, and ecommerce can explain which products and offers create or destroy contribution. And when budget changes happen against agreed thresholds instead of the loudest opinion in the room.

FAQ

Should marketing cost be inside contribution margin or not?

Both are defensible, which is exactly why it causes fights. The trick is to run two layers: contribution before acquisition (to set allowable CAC) and contribution after acquisition (to judge whether the customer paid off). The failure is not picking the "wrong" layer. It is using both loosely and subtracting CAC twice.

Isn't a weekly meeting just more overhead?

It is overhead if it only reports numbers. The weekly review earns its place when it produces decisions with an owner, a deadline, and a verification metric. If your last four reviews ended without a single budget or priority change, the meeting is the problem, not the cadence.

Can a $2M to $5M brand run this without a data warehouse?

Yes. A reconciled, governed spreadsheet is enough early on if the definitions are agreed and finance has signed off the cost policy. A warehouse becomes worth it once you have multiple markets, currencies, or large order volumes. Buy the tooling when the spreadsheet breaks, not before.

How is this different from what our agency already reports?

Agencies optimize their channel against the metric they own, usually attributed ROAS. That is useful for tuning campaigns and misleading for allocating the whole budget. A shared growth plan sits above channel execution and forces every function, agency included, onto one contribution and cash target. This is the gap a growth advisory or fractional head of growth engagement is built to close: directing agencies and internal teams under one commercial plan rather than replacing them.