PPC campaign structure: the foundation that makes optimization possible
- You know the three ad formats and what to start with (and what not yet).
- You understand match types and the division of labor: auto collects, broad researches, exact delivers.
- You build the proven three-campaign structure with a clean naming convention.
- You set starting bids and budgets that produce data instead of just spending.
After launch (Getting-started L12) your auto and one manual campaign are running. Now they become a system. The key insight: optimization requires structure. Throw every keyword into one campaign and you can neither see what works nor steer anything. This lesson builds the foundation — the optimization routine follows in lesson 2.
1The three ad formats
| Format | What it is | When it matters for you |
|---|---|---|
| Sponsored Products (SP) | Your products in search results and on product pages — looks like an organic result | From day one. The bulk of every FBA ad budget; the rest of this lesson is about SP |
| Sponsored Brands (SB) | Banner with logo + several products above search results, video ads | From Brand Registry and multiple products |
| Sponsored Display (SD) | Ads on product pages (including competitors'!) and off Amazon, retargeting | From Brand Registry; sensible once SP is maxed out |
2Match types: how closely must the query match?
- Exact: your ad shows only for exactly this term (plus close variants like plurals). Maximum control, maximum efficiency — for proven winner keywords.
- Phrase: the query must contain your phrase in order (“electric spice grinder” also matches “small electric spice grinder”). Medium control.
- Broad: Amazon serves on anything related. Finds terms you'd never think of — but burns money without supervision.
- Auto campaign: no keywords needed; Amazon matches based on your listing across four target groups (close match, loose match, substitutes, complements) — your data collector.
The same keyword “spice grinder set” in three match types, booked for the AURELO spice grinder set: as exact, your ad shows only for “spice grinder set” (plus close variants like “spice grinder sets”). As phrase, also for “spice grinder set wood” or “small spice grinder set” — the phrase stays intact. As broad, additionally for “grinder for spices” or “salt and pepper set” — good for discovery, but also for “electric spice grinder”, although your set is manual. The wider the match type, the bigger the reach — and the waste.
3The proven structure: three campaigns per product
| Campaign | Role | Content |
|---|---|---|
| 1. Auto (“collector”) | Discover new search terms and product placements | One ad group, moderate bid; runs permanently |
| 2. Manual broad/phrase (“research”) | Test your keyword set broadly | 10–20 terms from the keyword map in broad or phrase |
| 3. Manual exact (“performance”) | Scale proven winners efficiently | Only terms with proven orders; higher bids, biggest budget |
The cycle between them: auto and broad find converting search terms → you promote them into the exact campaign → and negative them in auto/broad so two campaigns never bid on the same term. That harvesting cycle is the core of lesson 2.
Think of the three campaigns as a football club: the auto campaign is your scouting department, searching everywhere for talent. Broad/phrase is the youth squad where candidates prove themselves. Exact is the first team — only proven performers play here, and most of the budget flows here. When a talent moves up to the first team, it stops playing in the youth squad (that's the negativing) — otherwise two of your teams stand on the pitch at once, taking the ball off each other.
Name campaigns by a scheme, e.g. “GM-SET | SP | Auto” / “GM-SET | SP | Exact”. In six months you'll have dozens of campaigns — without a system you search yourself to death; with one, the ad report reads like a dashboard.
4Starting bids and budgets
- Starting bid: use Amazon's suggested bid as orientation and start in the lower-to-middle range — typically €0.50–1.00 for a €25 product. After the first hundred clicks the formula from lesson 2 takes over anyway.
- Budget split: while exact is still empty, most flows into auto + broad. As the winner list grows, budget migrates toward exact — where every euro works hardest.
- Enough budget for data: a campaign exhausting its €5 budget by noon delivers skewed data (morning buyers only). Fewer campaigns with viable budgets beat many starved ones.
- Control value: your break-even ACOS from the unit-economics lesson. Without that number, every bid decision is guessing.
Starting lineup for the AURELO spice grinder set (€24.99, roughly €7 unit profit, so a break-even ACOS of ~28 %): Amazon's bid suggestion sits at €0.45–1.10 — you start at €0.70. Daily budgets: €8 auto, €8 research, €4 exact (still empty). At €0.70 per click that's a good 20 clicks a day at first, around 150 a week — enough data for the weekly routine from lesson 2 to make real decisions. Six weeks later, with ten proven winners in exact, the picture flips: €5 auto, €5 research, €10 exact.
5Two terms you now keep apart
- ACOS: ad spend ÷ advertised revenue. Measures campaign efficiency.
- TACOS: ad spend ÷ total revenue (organic + advertised). Measures how dependent your business is on ads. A falling TACOS at stable revenue means your organic rankings are taking over — the goal of the whole game (lesson 3).
One AURELO month in numbers: €300 ad spend, €1,200 advertised revenue, €3,000 total revenue. ACOS = 300 ÷ 1,200 = 25 % — the campaigns run efficiently. TACOS = 300 ÷ 3,000 = 10 % — the business depends only moderately on ads. Three months later: still €300 ad spend, but €4,000 total revenue → TACOS 7.5 %. The ACOS has barely moved, and yet the falling TACOS shows: the organic rankings are taking over.
The everything-in-one campaign: 50 keywords, all match types, one budget. Result: the two strongest terms eat the budget, the rest never gets enough data, and you can neither see nor steer what's happening. Structure isn't bureaucracy — it's the precondition for optimization to exist at all.
- Break-even ACOS calculated and noted.
- Three campaigns per product: auto, research (broad/phrase), exact.
- Naming convention defined and applied everywhere.
- Starting bids in the lower/middle suggestion range, budgets viable.
- Keyword map loaded into the research campaign.
- Weekly optimization slot in the calendar (lesson 2).
6Expert insight: placement modifiers as a structural lever
Why campaign layout truly decides money comes down to an often overlooked mechanic: daily budget and placement modifiers live at campaign level. An ad group has neither its own budget nor its own modifiers — it is just a container for keywords and bids. Everything you want to steer per placement, you in fact steer through how you cut your campaigns.
The math: a modifier raises the base bid for a placement by 0 to 900 % — it can only raise, never lower. From that follows the most important bidding rule for advanced sellers: calibrate the base bid to the weakest placement and lift the strong ones via modifier — never the other way round. The profitable maximum CPC per placement comes from the formula conversion rate × price × break-even ACOS. For the AURELO spice grinder set (€24.99, break-even ACOS 28 %, conversion per placement as an assumption):
| Placement | Conversion (assumption) | Profitable max CPC | Setting |
|---|---|---|---|
| Top of search | 12 % | 0.12 × 24.99 × 0.28 ≈ €0.84 | Modifier +200 % |
| Rest of search | 6 % | ≈ €0.42 | Modifier +50 % |
| Product pages | 4 % | ≈ €0.28 | Base bid €0.28 |
Set €0.84 as the base bid instead and you pay that price even where a click is only worth €0.28 — with no lowering modifier to fix it. Second layer of the same mechanic: with “dynamic bids — up and down”, Amazon may additionally raise the result, by up to 100 % top of search and up to 50 % elsewhere. A €0.28 base with a +200 % modifier can thus peak at €1.68 — always calculate your maximum through to the end before combining both levers.
- One modifier applies to every keyword in the campaign. Keywords with different placement behavior in one campaign force a compromise value that fits none of them.
- Isolate hero keywords. If one term drives a large share of your advertised revenue (rule of thumb: from about a third), it deserves its own exact campaign — guaranteed budget, its own modifier, a crystal-clear report.
- Don't fragment. Thirty single-keyword campaigns splinter data and maintenance; one to three isolated heroes plus a bundled long tail is the proven balance.
This fine-tuning needs data: start with the bids from this lesson and neutral modifiers. Replace the conversion assumptions with real values from the placement report only once each placement has roughly 100+ clicks — before that you are calibrating on noise.
Check yourself
6 quick questions — one at a time, instant feedback. With a free account your progress is saved.
Frequently asked
What total ad budget should I expect after launch?
In steady state, healthy accounts often settle at 5–15 % of revenue (TACOS). More concretely, calculate backwards: target ACOS × planned advertised revenue — and when in doubt, fund few campaigns properly rather than starving many.
Should I bid on my own brand keywords?
As soon as your brand gets searched: yes, with a small budget. Brand clicks are dirt cheap, convert excellently — and defend the position competitors otherwise bid on (which is allowed and common).
What about click fraud — do I pay for fake clicks?
Amazon filters invalid clicks automatically and credits detected invalid activity. Individual expensive clicks without a purchase are usually real prospects who changed their mind — the answer is conversion optimization, not conspiracy hunting.
The free Listing Check scores an ASIN from 0 to 100 in one minute: title, keywords, bullet points, images, A+ content and compliance — biggest weak spots first.
Everything in this academy comes from day-to-day selling practice — the same playbook behind Listimo, the tool that turns product photos into complete Amazon listings.