Getting customers
The sales funnel: from stranger to customer
The three stages, which campaign you build in each one, which audience you choose and what content goes in, with the ads written out.
8 min read8 partsFree
01What a funnel is, in plain words
A funnel is a person's path from not knowing you to buying from you. It's called a funnel because fewer people are left at each step: out of a thousand who see your ad, a hundred come in to look and three buy. That's not a problem, it's how it works.
The mistake that wastes the most money on ads is treating those thousand people the same. You don't talk to someone who has never seen you the same way you talk to someone who abandoned their cart yesterday, you don't show them the same thing and you don't ask them for the same thing. One ad for everyone means most people get the wrong message.
The three stages are named after the person's temperature: cold (doesn't know you), warm (saw you but didn't buy) and hot (interacted a lot or already bought from you).
02The three stages, at a glance
Cold: doesn't know you
- Goal
- Get discovered, and sell if the product allows it
- Campaign
- Sales · Traffic · Engagement
- Message
- Create the need. Don't start with the product
Warm: saw you, didn't buy
- Goal
- Convince them and clear the doubt in the way
- Campaign
- Sales, optimized for purchase
- Message
- The product, the offer, the guarantee
Hot: almost bought, or already did
- Goal
- Close, and make them come back
- Campaign
- Sales, 100%
- Message
- The final push or the next product
03How to build it in Meta, campaign by campaign
This is where theory turns into clicks. The minimum structure that works is three separate campaigns, not three ad sets inside one, because each stage has its own budget and you don't want Meta taking money from retargeting to give it to cold.
| Campaign | Objective in Meta | Audience | Budget |
|---|---|---|---|
| 1: Cold | Sales, Purchase event | Advantage or Broad. If the product is niche, direct + indirect interests | 60-70% of the total |
| 2: Warm | Sales, Purchase event | Custom audience: website visitors 120 days + Instagram and Facebook engagement 120 days | 20-25% |
| 3: Hot | Sales, Purchase event | Added to cart 30 days + buyers 180 days | 10-15% |
Do
- Exclude buyers from the cold and warm campaigns: paying to show your ad to someone who already bought is throwing money away.
- Exclude cart abandoners from the warm campaign, so they don't get both messages at once.
- ALWAYS optimize for the event furthest down the funnel that you have enough data for. If you have few purchases, optimize for «Initiate checkout» until you build volume.
- Start each ad set at the minimum budget and raise it little by little. The minimum is what forces the platform to look for cheap results.
Avoid
- Building a single campaign with all three audiences inside: Meta gives all the money to whoever converts fastest, which is always the hot audience, and no new people ever enter the funnel.
- Giving retargeting the same budget as cold. The hot audience is small and gets saturated in days.
- Optimizing the cold campaign for «Traffic» and expecting sales. You get cheap clicks from people who never buy.
- Duplicating an ad set you already scaled without bringing the budget back to the minimum.
04The content for each stage, written out
This is what almost never gets explained with concrete examples. Let's take a real business: a brand that sells steel thermoses at $45,000.
05Cold: the ad that creates the need
Someone who doesn't know you doesn't care about your thermos. They care about something that happens to them. The cold ad starts there, and the product only shows up as the answer.
BAD: a photo of the thermos on a white background, «1L steel thermos, $45,000, 3 interest-free installments». It's a shop window for someone who isn't shopping for thermoses.
GOOD, script for a 20-second video: it opens with a hand pouring mate (the Argentine herbal drink) and lukewarm water comes out. Cut to the person's face, resigned. On-screen text: «11 in the morning and it's already cold». Only then does the thermos show up; they pour, steam comes out. Close: «12 hours. Really.» and the price below.
The difference isn't the production: it's where it starts. The first one informs; the second makes someone think «that happens to me».
Do
- Start with the situation, not the product. The first 3 seconds decide whether the rest exists.
- Show the problem happening; don't describe it. Seeing the lukewarm water works better than saying «it cools down fast».
- Put text on screen: most people watch without sound.
Avoid
- Starting with the logo. Nobody is waiting for you.
- Asking for the purchase in second one. In cold, that cuts the video short.
- Talking about yourself («we're a family brand with 20 years of history»). They don't care about you yet.
06Warm: the ad that clears the doubt in the way
This person already saw the thermos. You don't need to convince them they need one: you need to remove the reason they haven't bought it yet. And you know that reason, because they write it to you in messages.
If the doubt is «does it really last 12 hours?», the warm ad is the proof: a timelapse with the thermometer at 8am and at 8pm. If the doubt is «what if I don't like it?», the ad is the guarantee written big. If the doubt is the price, it's the comparison: «$45,000 once, or $1,200 a week at the corner café».
One doubt per ad. Putting all three in the same piece answers none of them.
Do
- Write down the five objections you hear most and make one ad for each.
- Show the whole product here: in warm, the good photo, the price and the installments do belong.
- Use real customer material if you have it. A screenshot of a real message convinces more than a polished production.
Avoid
- Repeating the cold video. They already saw it, and seeing it again confirms nothing new is happening.
- Lowering the price as the first move. If the doubt wasn't the price, you gave away margin without moving the needle.
07Hot: the push, and what comes next
There are two different audiences here that often get mixed up. The cart abandoner is one step away and needs a reason to do it today. The buyer doesn't need you to sell them the same thing again: they need what comes next.
For the cart: «You left the black thermos in your cart. We'll hold it for 24 hours», with a small coupon and a real date. The urgency has to be true: if the «last day» repeats every week, you lose credibility forever.
For the buyer: the bombilla (the mate straw), the yerba holder, the matching mate cup. And there you can talk to them as a customer: «Since you already have the thermos...». That same text in cold ruins the ad, because someone who never bought from you thinks «I didn't buy anything».
Do
- Separate cart abandoners from buyers: they're two different messages and two different windows.
- Use 30 days for the cart and 180 for buyers. Someone who already bought from you doesn't forget you: they have the product.
- Offer the buyer something complementary. It's the cheapest sale you'll ever make.
Avoid
- Showing the customer message to someone who never bought.
- Inventing urgency. Once people catch a fake «last day», none of them work anymore.
- Spending as much on hot as on cold: that audience is small and burns out in days.
08When selling to cold audiences does NOT make sense
You also sell to cold audiences: using cold only to «build brand» burns budget with no return. But there's one real exception worth having clear before you spend.
An impulse, low-ticket product sells to cold audiences without problems: a thermos, some mate cups, a T-shirt. An expensive product, or one that needs trust (a professional service, a course that costs thousands of dollars, a construction job), doesn't. Nobody hires someone they saw for the first time thirty seconds ago.
If your product is in the second group, cold doesn't sell: it brings people in. And the work of closing belongs to warm and hot, with more pieces and more time in between.
| Type of product | Does it sell to cold? | What to do |
|---|---|---|
| Impulse, low ticket | Yes, directly | A sales campaign from cold |
| Medium ticket, quick decision | Sometimes | Try selling directly; if it doesn't close, add warm |
| High ticket or trust-based | No | Cold brings contacts; a person or retargeting closes |
Case
Thermos brand, $45,000 per product
A single ad with a photo of the thermos and the price, shown to everyone. It spends $300,000 a month and sells 4 thermoses.
- It builds the cold campaign with 65% ($195,000): the lukewarm-water video, Advantage audience, optimized for Purchase, excluding buyers.
- It builds the warm one with 22% ($66,000): three ads, one for each objection people write to it about (how long it lasts, the guarantee and the price). Audience: 120-day visitors and Instagram engagement.
- It builds the hot one with 13% ($39,000): 30-day cart with a coupon and a real date; 180-day buyers with the bombilla and the yerba holder.
- It lets them run for two weeks without touching anything, and only then looks at the numbers by campaign, not in total.
The same budget, split into four messages that fit each person depending on where they are. The hot campaign closes sales the single ad never closed, because nobody was talking to the person who had left the cart.