Module 6

Defining campaigns

16 lessons, 53 stepsOne instruction at a time, each with the result to expect. Tick steps as you go; Helix remembers.

Outcome: you can plan any ad campaign from scratch. You set the numbers it must hit, choose the goal and setup, pick the audience and budget, name it properly, brief the ads, test them fairly, read the results and decide: spend more, wait, make new ads, or stop.

This module connects your plan to Ads Manager (the place where you build Meta ads). The campaign brief template at the end is what Helix fills in with you when you press "New campaign" in Your ads.

Related SOPs: 05 Meta structure, 06 Meta daily optimisation, 07 Creative testing, 09 Google, 15 Promo calendar

In plain words

Stage: Attract (get seen by people who want what you sell). Every lesson below is tagged with its stage; see Attract, Convert, Grow.

What it is: How to plan one ad campaign from start to finish.

Why it matters: A plan with clear numbers tells you when to spend more, wait or stop, so you never guess.

Do this:

  1. Set your target cost per sale from your break-even numbers.
  2. Choose the goal (sales), the audience and the daily budget.
  3. Launch, wait for enough data, then decide: spend more, wait, new ads or stop.

Words to know:

  • Campaign: The top level in an ad account. It holds the goal, like getting sales.
  • Ad set: The middle level in Meta. It holds who sees the ads, where, and often the budget.
  • Target CPA: The cost per sale you aim for so each sale still leaves you profit.
  • Break-even CPA: The most you can pay in ads for one sale before that sale stops making money.

All words are explained in the glossary.


Lesson 6.1: Numbers firstAttract

Why this matters

Never open Ads Manager until you know the five numbers that say whether an ad makes or loses money. They come from Module 1 and Helix works them out from your daily numbers. With them, every later decision is a simple comparison.

0 of 3 steps done

Step 1 of 3

Open the Dashboard on Month and copy your target MER, break-even ROAS and break-even CPA from the metric cards.

Expected result

You have three numbers written on your campaign brief.

Open your key metrics

Good to know

Why cold campaigns get double: ad platforms miss many sales that cold ads cause, because people see an ad and buy days later another way. So the cost per sale they report looks worse than it really is.

Number What it means How to work it out Example (AOV $90, VCR 40%, FCR 15%, profit goal 15%)
Target MER The share of all sales you can spend on ads 100% - VCR - FCR - profit goal 30%
Break-even ROAS Sales per $1 of ads where a sale makes $0 1 ÷ (1 - VCR) 1.67
Break-even CPA The most you can pay for one sale before it loses money AOV x (1 - VCR) $54
Target blended CPA What you aim to pay per sale, across all customers AOV x target MER $27
Allowance for cold campaigns What a campaign aimed at new people may show per sale Target CPA x 2 $54

Lesson 6.2: Pick the job of the campaignAttract

Why this matters

Each campaign has one job, and the job decides the objective (the goal you pick in Ads Manager) and what the platform optimises for (the action it hunts for). Pick the wrong one and the platform finds clickers instead of buyers.

0 of 2 steps done

Step 1 of 2

Pick the job of your campaign from the table below, for example "Find new buyers".

Expected result

You have one job written on the brief.

Good to know

Job Objective to choose Aim for
Find new buyers Sales Purchase
Bring back visitors and subscribers Sales Purchase (if sales are few, sometimes add to cart)
Grow your email list Leads, or Sales sending people to a sign-up page Lead, or completed sign-up
Build excitement before a launch or sale Leads or Engagement Sign-ups
Test ads cheaply on a new platform Sales, aimed at an earlier step (view content, add to cart) when purchases are too few The earlier step, then switch to purchase
Show people products they looked at (catalog retargeting) Sales with your product catalog Purchase

Lesson 6.3: Testing campaigns vs scaling campaignsAttract

Why this matters

Keeping testing and winning separate stops new tests from upsetting what already works. Three kinds of campaign do the job: scaling, testing and build.

0 of 4 steps done

Step 1 of 4

Create a build campaign that is always switched off, and build each new ad there once.

Expected result

You have a paused campaign holding every ad you have made.

Lesson 6.4: Account structure by spendAttract

Why this matters

Cold means people who do not know you, warm means people who visited or engaged, and mixed means both. The right account structure depends on how much you spend a day. Too many campaigns for the budget starves each one of data.

0 of 3 steps done

Step 1 of 3

Find your daily spend band in the table below and set up the campaigns listed for it.

Expected result

Your account has the number and type of campaigns that suits your budget.

Good to know

Daily ad spend Suggested setup
Under $100 1 cold Sales campaign with broad targeting and your best 4 to 8 ads, plus 1 warm campaign. Put 70 to 80% of the budget into cold.
$100 to $500 Cold, mixed and warm campaigns, plus a testing campaign and a build campaign.
$500 to $1,000 Add a second cold campaign with different exclusions (people you leave out) or a different automated setup.
$1,000 to $3,000 Campaigns per country or product line. Separate sale campaigns during big events.
$3,000 or more A set test budget, creator and partnership ad campaigns, catalog video ads, more countries.

Lesson 6.5: AudiencesAttract

Why this matters

The audience is who sees your ads. Today broad audiences with strong ads usually win, while custom audiences and lookalikes help with warm and test campaigns. Set them up once and reuse them.

0 of 5 steps done

Step 1 of 5

For cold campaigns, start broad (Advantage+ audience, or country and age only) and only test interests after you have winning ads.

Expected result

Your cold ad sets have no stacked interests.

Lesson 6.6: BudgetsAttract

Why this matters

Budget decides how fast the platform learns and how much you can lose while it does. A few rules stop the classic mistakes of starving campaigns or resetting them.

0 of 4 steps done

Step 1 of 4

Set each new campaign's daily budget to 2 to 3 times your target CPA (for a $30 target CPA: $60 to $90 a day).

Expected result

The campaign can get enough purchases to learn.

Lesson 6.7: Naming conventionsAttract

Why this matters

A naming convention is a fixed pattern for names. It lets you, your team and Helix read the account at a glance and compare results later. It takes five minutes to set up.

0 of 3 steps done

Step 1 of 3

Name every campaign with this pattern: [Number]-[Auto/Manual]-[Cold/Mixed/Warm/Hot]-[Broad/Interest/LAL]-[None/Light/Harsh]-[BAU or Sale], for example 03-Manual-Cold-Broad-Harsh-BAU (BAU means business as usual; LAL means lookalike).

Expected result

Every campaign name follows the pattern.

Lesson 6.8: Creative concepts, angles and hooksAttract

Why this matters

Four words describe every ad: concept (the big idea), angle (why it matters to a specific person), hook (the first 3 seconds) and format. Testing one of them per batch is what makes a test fair.

0 of 3 steps done

Step 1 of 3

Write the concept for your next batch, for example "answering a one-star review".

Expected result

The batch has one big idea.

Lesson 6.9: Writing the creative briefAttract

Why this matters

A creative brief is a one-page set of instructions for each ad. It lets anyone, including a creator or an AI tool, produce an on-target ad without guessing.

0 of 3 steps done

Step 1 of 3

Open the template in SOP 08 and fill in the goal and who it is for (cold, warm or hot), the pillar, and the customer's problem in their own words.

Expected result

The top half of the brief is done.

Lesson 6.10: Testing methodologyAttract

Why this matters

Most ad tests never get enough visitors to be statistically certain, so you need fair default rules instead. These keep tests consistent and stop you calling winners too early.

0 of 2 steps done

Step 1 of 2

Set your test defaults from the table below: one thing changed per batch, the number of ads per batch for your budget, and the spend each ad gets before you judge it.

Expected result

Your test rules are written at the top of your testing campaign notes.

Good to know

Rule Default
Things you change per batch 1
Ads per batch 4 to 8 under $100 a day per campaign. 8 to 12 at $100 to $500. 12 to 15 above that.
Spend before you judge an ad 1 to 1.5 times target CPA (more for cold ads)
Expensive products Judge on early signs: outbound CTR, cost per click, add to cart rate, and how fast frequency rises
Budget for tests 20 to 30% of the total
Learning time Do not edit for 3 to 7 days after launch
Be careful with small numbers Under about 10 purchases per ad, treat results as a hint, not proof

Lesson 6.11: Reading metricsAttract

Why this matters

Each metric tells you about one part of the journey, from stopping the scroll to buying. Reading them in order shows exactly where an ad breaks.

0 of 3 steps done

Step 1 of 3

In Ads Manager click Columns > Customise columns and add hook rate, hold rate, outbound CTR, CPM, cost per outbound click, add to cart, cost per purchase, ROAS, frequency and reach, then save as a preset.

Expected result

You see all the metrics in the table below in one view.

Good to know

Metric What it tells you What good looks like
Hook rate (3-second views ÷ impressions) Does the first moment stop people? About 25 to 35% or more on video
Hold rate (views to 50%, or ThruPlays, ÷ 3-second views) Does the story keep them watching? Going up as you make new versions
Outbound CTR Does the ad make people want to click to your site? 1 to 1.5% or more. Below 0.5% is a problem.
CPM Cost to show the ad 1,000 times Compare with your own past and season, not other stores
Cost per outbound click CPM and CTR combined Must be low compared with RPV (what a visit earns)
Add to cart rate Does the page turn interest into carts? Compare with your site average
CPA (cost per purchase) How efficient the ad is Compare with the target for its audience
ROAS (as the platform reports it) Sales the platform credits to the ad, per $1 Compare with break-even, and with your MER trend
Frequency How often the same people see it Within the limit for its audience
Reach growth versus spend growth Are you reaching new people as you spend more? Reach should grow about as fast as spend

Lesson 6.12: Decision rules: scale, hold, refresh, killGrow

Why this matters

Four decisions cover every ad: scale, hold, refresh or kill. In Helix these show as Spend more, Wait, New ads needed and Stop. Using fixed rules removes emotion from the call.

0 of 3 steps done

Step 1 of 3

Open Your ads in Helix and read the verdict next to each campaign.

Expected result

Every campaign shows one of the four verdicts.

See what Helix found

Good to know

Decision When What to do
Scale (Spend more) Cost per sale is at or under target, frequency is fine, and MER for the last 3 days and the month so far is on target. Add 20% to the budget a day, or copy the ad into another campaign.
Hold (Wait) It is still learning, was edited recently, or results are mixed but getting better. Make no changes for 3 days.
Refresh creative (New ads needed) No strong ads are left, frequency is high, or fewer than 30% of the batch still perform. Launch a new batch from your list of ad ideas.
Kill (Stop) It has spent the "bad" amount (2 times the good CPA for its audience) with poor results or no purchases. Turn it off today.

Lesson 6.13: Google campaign definitionsAttract

Why this matters

Google campaigns each catch shoppers at a different moment, from searching your brand name to browsing YouTube. Knowing what each type is for stops you paying for the wrong clicks.

0 of 3 steps done

Step 1 of 3

Read the campaign types in the table below and list which ones you run today.

Expected result

You know your current Google setup by type.

Good to know

Campaign What it is for Key settings
Brand search Protect your own name cheaply when people search for it Small budget. Bid for impression share or set bids yourself. Your brand words only.
Brand shopping (optional) Show your products when people search your name Low priority, brand searches
Feed-only Performance Max or Shopping Find new customers from your product list Leave out your brand name. Group products by best sellers or margin level. Set a target ROAS above break-even.
Non-brand search Control over high-intent search words At most 5 keywords per ad group. A shared list of blocked words (negative keywords). Up to 15 headlines.
Competitor search Reach people comparing you with competitors Honest comparison ads. Watch costs closely.
Demand Gen and YouTube Visual ads to help people discover you Judge with view-through results (people who saw the ad and bought later)
Local Visits to a physical shop Only if you have a shop

Lesson 6.14: Promotional and seasonal campaignsAttract

Why this matters

Sale campaigns behave differently from your everyday ones. Building them separately, with daily budgets planned and judged as a whole event, protects your normal campaigns and your margin.

0 of 4 steps done

Step 1 of 4

Build sale campaigns separately from your BAU campaigns, copying your best cold campaign as the base for the sale's cold campaign.

Expected result

Your sale has its own campaigns, built and paused before launch.

Good to know

Black Friday plan (timeline):

When What to do
10 to 12 weeks before Set a sales target from last year and recent growth. Choose the offer. Order stock and free gifts. Start growing your email list.
8 weeks Brief the sale ads: the announcement, how the offer works, best sellers, gift guides, last chance. Plan landing pages.
6 weeks Plan the sale emails and texts. Book creators. Check site speed and checkout.
4 weeks Build the sale campaigns (switched off) and the excitement campaign. Load ads into the build campaign.
2 weeks Stop changing the website code. Test discount codes, free gift minimums and sale prices in your catalog.
5 to 3 days before Excitement phase: sign-up ads and teaser emails. Move normal budget across slowly.
Launch day Early access for your best customers, then everyone. Send the launch email and text. Turn sale campaigns on.
Middle of the sale Release something new or a limited bundle for 24 hours. Refresh the ads.
Last 48 hours Last-chance emails and texts, and ads that stress the deadline.
After Turn normal campaigns back up slowly. Welcome new buyers. Compare results with the target.

Lesson 6.15: Common campaign mistakesAttract

Why this matters

Most wasted ad money comes from a handful of repeat mistakes. Checking for them takes two minutes.

0 of 4 steps done

Step 1 of 4

Check you have break-even numbers written down before starting any campaign.

Expected result

Every live campaign has a break-even CPA on its brief.

Lesson 6.16: The campaign brief templateAttract

Why this matters

A campaign brief puts every decision on one page before you spend a dollar. Helix fills in the numbers from your scorecard when you press New campaign in Your ads.

0 of 4 steps done

Step 1 of 4

In Helix open Your ads > New campaign, or copy the template below.

Expected result

You have a blank brief with your numbers filled in.

Good to know

CAMPAIGN BRIEF

1. Name (use the naming pattern):
2. Channel: Meta / Google / TikTok / Pinterest / other
3. Job of the campaign: new buyers / warm / list growth / excitement / catalog / sale
4. Objective, and what it aims for:
5. Products and offer:
6. Numbers
   - AOV:            - VCR:           - Target MER:
   - Break-even CPA: - Target CPA:    - CPA allowance for this audience:
   - Break-even ROAS:
7. Audience: broad / interests / lookalike / custom. Who to leave out:
8. Countries and placements:
9. Budget: daily $____  test period ____ days  most you accept losing $____
10. Ad plan: concepts, angles, hooks, formats, number of ads, the one thing tested
11. Landing page: URL. Does it match the ad's message? (Y/N)
12. Tracking check: pixel and server events working, purchase is the only main conversion (Y/N)
13. Success looks like: CPA <= ____ after spending ____; outbound CTR >= ____
14. Decision date:
15. If it wins: +20% a day up to $____, then copy winners to ____
16. Stop rule: spend ____ with CPA above ____, or no purchases
17. Approvals: who approves the launch and budget changes

Self-check

  1. AOV is $120 and VCR is 35%. What is break-even CPA?
  2. Why keep testing and scaling campaigns apart?
  3. How many ads go in a batch at $250 a day?
  4. A cold ad has 0.4% outbound CTR after enough spend. What do you decide?
  5. Which Google campaign protects your brand name?
  6. What happens 5 to 3 days before a big sale?
Answers
  1. 120 x 0.65 = $78.
  2. So new ads do not reset proven campaigns, and tests get a fair, fixed budget.
  3. 8 to 12.
  4. Stop it, or rework the message. A click problem comes from the ad or the offer.
  5. Brand search.
  6. The excitement phase: sign-up ads and teaser emails, and a slow move of budget across.

Words to know in this module

Campaign
The top level in an ad account. It holds the goal, like getting sales.
Creative
The ad itself: the picture or video plus the words.
CPA
Cost per acquisition: how much ad money it took to get one sale.
Ad set
The middle level in Meta. It holds who sees the ads, where, and often the budget.
Target CPA
The cost per sale you aim for so each sale still leaves you profit.
Profit
What is left from sales after product cost, shipping, payment fees, discounts, refunds and ads. This is the number Helix cares about most.
Break-even CPA
The most you can pay in ads for one sale before that sale stops making money.
MER
All ad spend divided by all sales, as a percentage. MER 25% means $25 of ads for every $100 of sales.
Break-even ROAS
The ROAS where an order makes $0 profit after its costs. Below it, ads lose money.
ROAS
Return on ad spend: how many dollars of sales you get for each $1 of ads. ROAS 3 means $3 of sales for $1 of ads.
LTV
How much a customer spends with you over time, not just on the first order.
Margin
The share of each sale you keep after product cost. A $100 sale with $40 product cost is a 60% margin.
AOV
Average order value: total sales divided by number of orders.
Landing page
The page someone lands on after clicking an ad.
All words to know