Amazon Ad Budgets — ACOS, TACOS, and Where Money Should Go

Part 1

The only three ways to grow sales

Before any budget mechanics, anchor on the three — and only three — variables that determine total sales on Amazon. Every budget decision ties back to this equation:

Traffic — the volume lever

Traffic (sessions in Seller Central, glance views in Vendor Central) is total visits to your detail pages. Double it, hold the rest constant, and sales double. Its drivers: category search volume (largely outside your control, but monitor it), your share of that search — organic and paid, ad spend efficiently deployed, click-through rate (top-of-search placements alone can multiply CTR several times over), and off-platform pushes from Facebook, Google, or TikTok that spike Amazon traffic.

Conversion rate — the efficiency lever

Conversion is how often a visit becomes an order, and small swings move sales disproportionately: page quality (images, bullets, A+ content), reviews — slipping from 4.5 to 4.0 stars can gut conversion, pricing and promotions — on higher-ticket products a modest price rise can trigger a disproportionately larger conversion drop, traffic mix — non-brand and upper-funnel traffic converts at a fraction of branded traffic, so a mix shift alone drags the average down, and broader price sensitivity in the market.

Average sales price — the revenue-per-unit lever

Price sets revenue per unit and the profit pool available for advertising. It moves through price changes, product-mix shifts, and promotions. And because price and conversion are tightly linked, the compounding can surprise you: a +10% price rise that triggers a −20% conversion drop leaves total sales down 12%.

Part 2

Breakeven and target ACOS

Before a single dollar goes into ads, answer one question: at what ACOS do I start losing money? Everything — targets, investment posture, tactical calls — flows from that number.

Pre-ad profit per sale (PAPS)

A $20 product with $6 in Amazon fees and $4 landed COGS has a $10 PAPS — the absolute maximum you can spend on ads per sale before losing money. FBM sellers pay lower fees (no FBA fulfillment fee) so their PAPS runs higher, but slower shipping usually means lower conversion, which caps how much traffic they can profitably drive. Vendors calculate from shipped COGS rather than retail price, while the ad console reports ACOS against retail — mind the denominator mismatch.

From breakeven to target

Target ACOS is a strategic choice about investment level. Below breakeven, every ad sale is profitable — margin protection mode. At breakeven, ad sales make no profit but lose none; the profit arrives as the organic sales your ad velocity stimulates. Above breakeven, you are deliberately investing per ad sale to buy rank and market share.

Part 3

ACOS vs TACOS — what actually matters

Plenty of voices claim ACOS is a vanity metric. That is only true when campaigns are set up badly. Saying ad-sales efficiency doesn't matter is really saying you don't understand what's driving attribution in your account.

When ACOS and TACOS disagree

ACOS can fall while TACOS — ad spend over total sales — rises. That divergence means ads are cannibalizing organic sales instead of creating incremental ones. The usual culprits: view-based Sponsored Display attribution quietly claiming organic sales as its own, brand campaigns spending against search terms you already own, and paying for clicks on keywords where you already hold the top organic spot — buying sales that were coming anyway.

Using one to steer the other

The relationship tracks your organic-to-ad ratio. If organic softens — a competitor's deal, lost rank — TACOS rises even with ACOS flat. That is your cue to pull target ACOS down until the balance recovers. If organic is surging, you have headroom to push ACOS up and capture more growth.

Part 4

Why lowering CPC can increase ACOS

The most misunderstood dynamic in Amazon PPC: you cut bids expecting ACOS to fall, and it spikes instead. Two mechanisms cause it.

Reason 1 — placement shifts

Placements convert very differently; top of search converts several times better than product pages. A $1.50 bid might win a mix of both — some expensive, high-converting top-of-search clicks and some cheap product-page ones. Cut the bid to $1 and the expensive placements vanish first. What's left is spend concentrated in the worst-converting real estate. CPC fell; ACOS rose.

Reason 2 — search term shifts

In auto and broad campaigns, the same thing happens at the query level. The strong term needed $1.50 to stay in the auction; at $1 you no longer win it. The terms you still win are the low-volume, low-intent leftovers. Cheaper clicks, worse buyers.

The fix is to never move bids in isolation: when cutting base bids, raise the top-of-search modifier so you keep the placement that converts; and harvest strong terms into exact match, where keyword-level bids don't ripple across everything else. In a well-managed account CPC and ACOS track each other closely — where they don't, the placement and segmentation hygiene is usually the gap.

Part 5

Budget management fundamentals

The purpose of budgets is not to control spend. Budgets are safety rails — bids are how you actually control spend, because bids set CPC and CPC × clicks is spend. A budget cap is a governor on the engine, not the accelerator.

The four states of every campaign

A cash-flow reality check: sometimes the money to extend a profitable campaign simply isn't there. That's a real constraint, not a failure. Fund the best-return campaigns first, trim the rest, and scale as revenue allows. The principle is about strategic intent — never let winners cap out because of lazy bid management.

Raising a budget changes nothing

If a campaign spends $100/day against a $500 cap, raising the cap to $1,000 changes nothing — it will still spend $100. Spend grows only through bids (winning more auctions) or keywords (more chances to spend). And no, a big budget does not signal anything to Amazon or earn ad rank — rank comes from bid × relevance, not the budget field.

Pacing and starting budgets

Three levels of control exist: campaign caps (set generously — they're backstops, and summing them is a terrible spend forecast), portfolio budgets (monthly caps per product line — useful for large catalogues), and the account-level daily cap in Seller Central (a master backstop that trades away granular control).

Pace with bids, not budgets: over-pacing with good ACOS → fund it or trim bids slightly to protect margin; over-pacing with bad ACOS → cut bids (efficiency improves and campaigns run all day); under-pacing with good ACOS → raise bids while performance supports it; under-pacing with bad ACOS → fix efficiency first, then scale. For new campaigns, start at or slightly below your account's average spend per active campaign — a lower cap is protection while you learn the campaign's temperament. For brand-new products with no history: target CPC × desired daily clicks.

Part 6

Budget by ad type — SP, SB, SD

Not all ad types earn budget equally. The natural allocation for most accounts:

Sponsored Products dominates for concrete reasons: it converts best, it looks native — shoppers often can't tell it's an ad — and the inventory is vastly larger: multiple SP slots at the top of search and dozens across the page, versus a single SB banner. More placements means more chances to win a click.

Sponsored Brands carries the most expensive placements in the console. It earns meaningful budget only when you have genuinely strong video and imagery — without good creative, you're paying premium CPCs to be ignored. Treat it like a flashier Sponsored Products ad with the same ACOS bar, and only go heavier (up to ~20%) for deliberate awareness plays judged on impressions and new-to-brand metrics.

Part 7

Budget by match type

When bids are managed properly, match-type allocation settles into a natural pattern on its own — roughly two-thirds exact, a small slice of broad/phrase, and a fifth in auto. It isn't a rule you enforce; it's what healthy optimization produces.

Exact earns the majority for three reasons: scalability (one bid change on a broad keyword ripples across hundreds of terms; exact gives per-term control), certainty (you know precisely which term you're buying), and concentration — in most categories a small fraction of queries drives the large majority of volume, and once you've harvested those into exact, that's naturally where the spend lives.

Two exceptions. New launches run exact-heavy from day one — 80–100% — because you know what the product is and which terms matter, while auto on a freshly indexed listing mostly buys irrelevant traffic until Amazon learns the product. And Sponsored Brands should run nearly all exact: it's your most expensive ad type, so premium CPCs should never leak into broad matches — feed it the best terms from your SP exact campaigns.

Part 8

Death by a thousand cuts

The most common and least discussed budget leak: thousands of low-visibility keywords, each wasting a little, adding up to a lot. No single keyword looks alarming — fifty cents here, a dollar there — but multiplied across an account it can quietly eat a tenth or more of total spend.

What causes it

Over-harvesting stuffs exact campaigns with terms that never had volume. Runaway escalation — automation or a manager raising bids on keywords because they “aren't getting impressions” — walks a $0.30 keyword up to $5 over weeks of five-percent nudges. And with no bid ceiling anywhere, nothing stops the spiral.

How to find it

Filter for keywords with fewer clicks than your ACTC — they haven't had a fair chance to convert — and bids above the target CPC. That intersection is your thousand cuts: terms too young to judge, priced as if they were proven. Overpaying for unproven clicks is the definition of the leak.

The fix — dynamic bid ceilings

Not an account-wide cap — too blunt, since a $10 product and a $100 product afford wildly different CPCs. Set the ceiling per ad group, from that ad group's own economics, at the level its worst placement can afford. Below the ceiling, low-visibility keywords are safe to raise 5–10% to gather data. Above it, bring them down to the ceiling — that's the whole discipline.

Part 9

How to lower ACOS — the only two levers

There are exactly two levers, and every tactic in this guide maps to one of them:

Diagnose first

Compare CPC and RPC period over period. CPC up and RPC flat → spend-side problem. RPC down and CPC flat → conversion problem. Both moved → the bigger mover first, because the two compound: a 10% CPC rise plus a 10% RPC drop is roughly a 20% ACOS jump.

Reducing CPC

Raising RPC

The Pareto workflow

You do not need to optimize 170 campaigns. In a typical account the top 8–10 campaigns drive half or more of all spend and sales. Sort by spend, take the ~20% that moves ~80% of the money, work their keywords, placements, and search terms, and only then descend to the next tier. Same effort, radically more impact.

The whole guide reduces to one sentence: spend where you make money, stop spending where you don't, and use math instead of guesswork. That's how we run every account — and if you want yours reviewed against these numbers, the first diagnosis is free.