Ecommerce & Digital
How to Build an Ecommerce Budget That Survives Contact With Reality
1 March 2025
Most ecommerce budgets are built backwards. Start with the revenue target, work back to the required traffic, apply a conversion rate, add a cost structure that produces acceptable margin, and present the result as a plan. The problem is that every number in that chain is an assumption, and the assumptions are almost always optimistic.
The revenue target is what the business needs, not what the market will deliver. The traffic number is achievable in the model but not in practice without the advertising spend that has not been accounted for. The conversion rate is the industry average rather than the rate the business will actually achieve with its specific product, listing quality, and review count. The cost structure omits the costs that were not anticipated when the budget was built.
A budget built this way does not survive first contact with reality. It becomes a source of frustration rather than a management tool, and the exercise of producing it next year becomes an exercise everyone goes through without believing in the result.
Start with costs, not revenue
The budget that survives is built from the bottom up, starting with what you know rather than what you hope.
The cost structure is more knowable than the revenue, and understanding the ecommerce P&L structure is the foundation. Platform fees are fixed percentages. Fulfilment costs can be modelled from the carrier rates and your average order weight. Payment processing fees are a known margin. Staff costs are contractual. Software subscriptions are recurring. Starting with these produces a cost base that is close to reality regardless of what revenue does.
Then add the variable costs that scale with revenue: advertising, returns, customer service volume. For each of these, use actual rates from your own data or, if you do not have it, conservative estimates from comparable businesses. The advertising cost per order in your category is knowable. Use it.
The revenue model that is honest
Revenue should be modelled from traffic and conversion, not from a growth percentage applied to last year.
Traffic breaks into organic and paid. Organic traffic is knowable from your analytics and grows slowly unless you invest in SEO. Paid traffic is a function of your advertising budget and your cost per click in your category. If you have not run paid advertising before, your CPC assumptions are guesses. Flag them as such and build a sensitivity table.
Conversion rate should be based on your own historical data if you have it, and benchmarked against category data if you do not. A new brand with limited reviews should not be budgeting at category average conversion rates. It will not achieve them until review volume is there to support them.
The costs that kill ecommerce budgets
Customer acquisition costs that compound. Paid advertising costs tend to increase as you scale, not decrease. The cheapest customers come first. As you exhaust the easiest audiences, cost per acquisition rises. Budget for this rather than assuming flat CPA across the revenue range.
Returns at realistic rates. Model returns at your actual category rate, not at zero or at an optimistic assumption. Include the full cost: refund, return shipping where you bear it, processing cost, and stock disposition write-down where relevant.
The cost of building Amazon velocity. Product launch budgeting matters: in the first three to six months on Amazon, the cost of building review volume and search ranking through Vine enrolment, aggressive PPC, and promotional pricing should be treated as a launch investment, not a recurring cost. But it needs to be in the budget.
Seasonal working capital. If the business is seasonal, the working capital requirement at peak needs to be funded. A budget that does not model the monthly cash position through a seasonal cycle will miss the funding requirement until it arrives.
Scenario planning, not point estimates
Present the budget with three scenarios: conservative, base, and stretch. The conservative case should be survivable. The base case should be achievable with disciplined execution. The stretch case should require things to go better than expected.
The question is not which scenario will occur. It is whether the business can fund itself through the conservative case. Having the right finance infrastructure to monitor actuals against scenarios is what makes the budget a living tool. If the conservative case cannot be funded, the budget needs to be reworked before it is approved, not after.
Maebh Collins is a Chartered Accountant (FCA, ICAEW) with twenty years of operational experience building and managing ecommerce budgets across multiple channels and markets.
Maebh Collins is a Chartered Accountant (FCA, ICAEW), Big 4 trained, with twenty years of experience building and running international businesses. She specialises in finance transformation, ecommerce operations, and digital strategy.