How to make Amazon sales projections

When I’m in my Socratic mom mode, I answer my kids’ questions with another question:
“If you knew the answer, what would you do?”

They usually sigh, but then they start reasoning, thinking out loud, connecting what they know. And more often than not, land on an acceptably right answer.

Making projections for your business isn’t that different. You start with what you know: prior data, reasonable assumptions, and a few goalposts. The rest is more about thinking clearly and thoroughly, not necessarily predicting perfectly.

With 2026 planning right around the corner, here are a few principles for building Amazon projections that are actually useful. The kind that help you make decisions, not be a spreadsheet exercise.

What’s the purpose?

First and foremost: why are we doing forecasting? To help plan budgets and cash flow? Raise money? Or be accountable to current investors? There is no wrong answer, but it’s important to be honest for it to be useful.

A question to ask here is: what decisions are we trying to make?

I’ve given input on sales projections for brands that needed it to raise money or report to stakeholders. And to brands that needed forecasting to plan production for their multiple channels. Both were sales projections, but very different mindset and underlying reasons for doing them.

Purpose of creating sales projections is to help make decisions, not to be predictive about future. So sales projections should then be reverse-engineered from the decisions they are intended to support.

Inputs before outcomes

Amazon is a marketplace, and like any marketplace, it is influenced not only by the all powerful Amazon self, but by myriad of other participants. Everybody exists in that ecosystem, impacting and influencing each other all the time.

On top of that we have macro factors (tariffs, supply chain, etc.) that we can’t control.

To make sales projections useful, it’s best to focus on what you can control.

Modeling based on inputs, rather than guesstimating for outcomes. Which means using existing data to drive projections.

Main top line inputs: Sales per SKU, Average Order Value, Average Selling Price

Main traffic and conversion inputs: Sessions, Conversion Rate, Ad Spend, Sales from Ads, TACOS, ACOS, Buy Box Rate, Category Share

Main cost inputs: COGS, Amazon fees (note they change every year Q1), Fulfillment and Storage, Returns

Unit economics is a huge input in terms of impact to sales projections.

I recommend track ad spend as part of your contribution margin. This aligns your Amazon P&L with actual operational profit, not just top line.

Cushion for Changes

There are qualitative factors within your control that you also need to account for.

New hire, switching Amazon agency/team? Ramp up of 1-3 months cushion.

Switching 3PL? Allow for some out of stock.

Getting into a national retail chain? Build in out of stock or less marketing budget for Amazon, since getting into national retail is capital intensive.

In short, changes present risk, so adding a cushion of impact to performance on Amazon to your baseline scenario would be wise.

Speaking of baseline..

Scenario planning over single goal setting

Building 3 coherent futures using different assumption sets of inputs is very helpful.

Because again, the real benefit of sales projections is not getting it right, but forcing clarity and making decisions.

You can call it good/better/best, baseline/maintain/push, conservative/base/ambitious, but the principle is to think through various cases, and be prepared accordingly.

Baseline projections can be projected on the existing data. For example, taking last 12 months, and assuming same growth, same performance on main KPIs. THen layer with any qualitative factors changes, changes in Amazon fees, changes in COGS/landing cost, change in your SKUs portfolio, or advertising budget.

Value of scenario planning is in helping prepare for a range of outcomes.

Growth and Optimization Cycle

Every brand wants the same thing: margins that hold-or even improve - as sales grow.
But in practice, growth rarely follows a straight upward line. It moves in cycles: growth and optimization.

Growth phases usually come from one of three sources:

• Doing more of what already works (which delivers diminishing returns over time).

• Trying something new, such as a marketing strategy or channel expansion (which introduces short-term inefficiencies while the learning curve plays out).

• Launching new products or entering new markets (which often requires upfront investment before payback).

Each of these creates turbulence in your P&L—spikes in ad costs, temporary dips in efficiency, or increased working capital needs. Then comes the optimization phase, where you refine what was built, extract efficiency, and stabilize margins before the next push.

A realistic forecast builds in these ebbs and flows. It anticipates periods of push and recovery instead of assuming a smooth upward line. If you plan to drive growth on Amazon next year, also plan time and budget to optimize afterward. Growth comes in waves, and your financial plan should move with them.

A note on marketing budget planning

One of the most common questions I get from younger brands is:
“What kind of return can I expect on Amazon marketing? If I knew it was profitable, I’d spend whatever it takes.”

Yes, me too. If returns were guaranteed, you’d always find the money.
But business doesn’t move in linear, predictable lines.

Advertising costs (especially CPCs) on Amazon continue to rise. And while strong brand awareness or off-Amazon marketing can help delay the need for heavy PPC spend, at some point you have to advertise to grow. To avoid a plateau, you have to embrace it.

A more sustainable approach is to plan your marketing as a percentage of sales, then focus on improving efficiency within that budget. For most CPG brands, that means allocating roughly 10–15% of monthly sales to Amazon advertising. You can control that variable (i.e your spend) but not the return. Instead of chasing predictability, manage for consistency, and optimize from there.

Projecting sales is part science, part strategy, part innate business instincts and tacit business knowledge. The best approach to useful forecasting is when it’s there to guide decisions, rather than be a goal setting or prediction exercise.

Saludos,

Irina

Strategic level guidance for CPG manufacturers on Amazon. Reach out for help.