Digim.pt

Como reduzimos o CPA em 42% sem aumentar o orçamento
Há uma pergunta que ouço muitas vezes:
«Será que precisamos de investir mais para conseguir melhores resultados?»
A resposta é... depende.
Neste caso, a resposta foi um simples não.
Quando comecei a trabalhar com este cliente, o problema não era o orçamento.
Era tudo o resto.
As campanhas estavam ativas há meses.
Havia vendas.
Havia leads.
Mas o CPA continuava a subir.
A primeira reação costuma ser:
«Vamos aumentar o orçamento.»
Percebo a lógica.
Mas aumentar o orçamento de uma campanha mal estruturada é como pôr mais água num balde furado.
O problema continua lá.
Antes de mexermos no investimento, fizemos o básico (que, curiosamente, muitas vezes é o que f**a por fazer):
✔️ Reestruturámos as campanhas.
✔️ Eliminámos públicos que estavam a competir entre si.
✔️ Revimos os criativos e a mensagem.
✔️ Garantimos que o utilizador encontrava exatamente a mesma promessa na landing page.
Nada de truques.
Nada de «hack do algoritmo».
Apenas estratégia e consistência.
O resultado?
👉 CPA reduzido em 42%.
👉 Sem aumentar o orçamento.
Esta é uma das coisas de que mais gosto no marketing digital.
Nem sempre a solução passa por gastar mais.
Às vezes passa simplesmente por deixar de desperdiçar o que já estamos a investir.
E, honestamente...
O botão «Aumentar orçamento» costuma ser muito mais fácil de encontrar do que o botão «Parar e analisar o que está mal.»
Mas é esse segundo que costuma fazer a diferença.

☕ Saturday Coffee & Marketing | 1 August 2026
I spent this week reading AI product announcements, so you don’t have to. 😄
Almost every major AI company released something new. Here’s what actually matters if you run a business or work in marketing.

🎬 Google AI (Gemini + Veo)
AI video editing is getting seriously good.
One product video can now become multiple versions for different audiences, formats and campaigns.
💬 The biggest reaction online?
“Everyone will have the same AI. Creativity becomes the competitive advantage.”
Couldn’t agree more.

🤖 OpenAI
Stop asking ChatGPT to write captions.
Start asking it to:
✔️ Review your landing page
✔️ Find conversion problems
✔️ Suggest A/B tests
Same AI.
Much smarter prompts.

🧠 Claude
Still my go-to for:
✔️ Marketing strategies
✔️ Client proposals
✔️ Competitor analysis
People describe it perfectly:
“The colleague who thinks before answering.”

🔍 Perplexity
Probably my favourite research tool right now.
Perfect for:
✔️ Competitor research
✔️ Statistics
✔️ Finding original sources
Huge bonus?
Every answer links back to its source.

☕ My takeaway
Stop asking:
❌ Which AI is the best?
Start asking:
✅ Which AI is best for THIS task?

My AI team looks like this:
🤖 ChatGPT → ideas & content
🧠 Claude → strategy
🔍 Perplexity → research
🎬 Google AI → visuals & video

Each tool has its own job.
Just like people in a marketing team.
And no... AI still doesn’t replace a good team.
It simply helps them work faster.

🎯 Challenge for next week
Choose ONE repetitive task and let AI handle it.
Then use that saved hour to talk to a real customer.
You’ll learn far more than from testing another AI tool.

👉 Read the full version on my LinkedIn (link in bio).
☕ See you next Saturday!



Why Your Ecommerce Store Doesn’t Need More Traffic (It Needs Better Conversion)
One of the biggest myths in ecommerce? “We need more traffic.” Maybe.

But before spending another €500 on Meta Ads, ask yourself one question:
What happens to the people already visiting your website?

Imagine 1,000 people land on your store every month.
If only 1% buy, that’s 10 orders.
Now imagine improving your website so 2% buy.

Without spending another euro on ads, you’ve just doubled your sales.
That’s the power of conversion optimization.

I often see ecommerce stores focusing on traffic while ignoring problems like:
❌ Slow websites
❌ Confusing navigation
❌ Weak product descriptions
❌ No customer reviews
❌ Complicated checkout
❌ Hidden delivery costs

More traffic won’t fix any of these.
It’s like pouring more water into a leaking bucket.

Before increasing your ad budget, ask yourself:
✔️ Is my website easy to buy from?
✔️ Can visitors find what they’re looking for in under 30 seconds?
✔️ Would I buy from my own store?

Sometimes the biggest growth doesn’t come from finding more customers.
It comes from helping the customers you already have click “Buy Now.”
That’s usually cheaper than buying another thousand website visitors.

💬 If you could improve just one thing on your website today, what would it be?

☕ Saturday Coffee & Marketing | Week 30
This week I filtered 40+ AI & Digital Marketing updates so you don’t have to.
Here’s what I’d actually recommend trying next week:
🤖 Meta Muse Image
Don’t let AI build your campaign. Let it create 5 creative variations of the strategy you’ve already planned.
✍️ Claude vs ChatGPT
I use both.

• ChatGPT → ideas, captions, Canva prompts.
• Claude → strategies, proposals, competitor research.
Different tools. Different jobs.

🎨 Canva
It’s becoming much more than a design tool. If you’re still switching between five different apps, maybe it’s time to simplify your workflow.

📈 SEO + AI
Stop chasing keywords.
Start answering real customer questions.
One good article can now bring traffic from both Google and AI search.

🛒 Shopify Magic
The quality of AI depends on your prompt.
More context = better results.

🧠 Tool of the week
Google Fact Check Explorer.
A free tool every marketer should have in their bookmarks.

🎯 Challenge
Let AI take over one repetitive task next week.
Use the time you save to talk to a real customer instead.
That’s still the best marketing research you’ll ever do.

👉 I share this roundup every Saturday.
The full version with practical examples, official sources and links is available on my LinkedIn (link in bio).

💬 Which AI tool saved you the most time this week?

“Is this real… or?” That’s probably the question I get asked most these days.

The honest answer? Sometimes I don’t know. And that’s OK.

AI has become incredibly good. Some videos and images are almost impossible to identify just by looking at them. But before you hit Share, here’s my 1-minute fact-check checklist.

🔍 1. Check the source first. Who posted it?
- An official news outlet?
- The original creator?
- Or a page that usually posts viral content?
If the source is unknown, be sceptical.

🖼️ 2. Reverse search the image
Use:
- Google Lens
- Bing Visual Search
- TinEye
- Google Fact Check Explorer
You might discover the same image was published years ago... in a completely different context.

🎥 3. Watch the video twice
The first time, watch the story.
The second time, watch the details.

Look at:
- Hands and fingers
- Teeth
- Glass reflections
- Jewellery
- Text on signs
- Background faces
- Shadows
- Objects appearing or disappearing

AI is improving fast, but small inconsistencies still happen. Also some verif**ation plugins like InVid can be helpful (link is below or just google it ;-)

📰 4. Can anyone else confirm it?
If it’s a major event, there should be more than one reliable source reporting it.
If only one random account has the “exclusive”...
🔴 That’s a red flag.

🤖 5. Ask AI… but don’t trust only one AI.

Upload the image or video to ChatGPT, Gemini or Claude and ask: “What details suggest this might be AI-generated?”

Treat the answer as an opinion, not a verdict.

⚠️ 6. Slow down
AI spreads because we share before we think. The easiest way to avoid misinformation? Wait five minutes. Check twice. Share once.

Technology is changing. Critical thinking isn’t. And honestly… That’s probably becoming the most valuable digital skill of all - THINK YOURSELF! USE YOUR OWN BRAIN!

Stop boosting posts.
Yes, I know. It takes two taps, Meta says “Reach more people” and suddenly it feels like you’re doing marketing.
But here’s the uncomfortable truth:
Likes don’t pay your bills.
Almost every week I hear:
“We’ve tried Facebook ads. They don’t work.”
Then I look at the account.
Most of the time, there isn’t a Meta Ads strategy.
There’s just a boosted post.
And those are two very different things.
A boosted post is designed to find people who are likely to like, comment or share.
A Meta Ads campaign is built to find people who are likely to take action:
✔️ Send an enquiry
✔️ Buy
✔️ Book a call
✔️ Become a customer
That’s a big difference.
And here’s another thing.
Sometimes the ad isn’t the problem at all.
People click.
Then they land on a slow website.
Or a landing page with six different messages.
Or a contact form asking for their life story.
Then they leave.
Poor ads get the blame.
Marketing isn’t about one magical button.
It’s about making everything work together:
→ the message
→ the ad
→ the landing page
→ the offer
Only then do your ads have a real chance to bring results.
So before you hit Boost Post again, ask yourself:
Do I want more likes... or more customers?
(They’re usually not the same thing.)
👇 Have you ever boosted a post? Did it actually bring you customers?

🗞️Weekly Meta Ads & AI Update
Another week, another reminder that marketing is changing faster than most businesses can keep up.

Here’s what actually matters.
🤖 Meta is pushing AI even further into ad creatives.

Meta introduced Muse Image and previewed Muse Video, its new AI generation models. The goal is simple: make it easier (and faster) to create ad creatives directly inside Meta’s ecosystem.

My take?
Great for testing ideas quickly.
Not so great if everyone starts using the same AI-generated visuals.
AI can help you create content. It still can’t create your brand.

🎯 AI is getting better. Strategy still wins.
I’m seeing more business owners asking: “Can AI run my marketing?”
My answer hasn’t changed.
AI can write.
AI can design.
AI can analyse data.
But AI still doesn’t know your customers better than you do!!
The businesses that will win aren’t the ones using the most AI.
They’re the ones asking better questions.

⚠️ Don’t switch your brain off
Some advertisers have reported strange AI-generated ad variations inside Meta - from distorted products to messaging that doesn’t match the brand.

It’s a good reminder that AI is an assistant, not an autopilot!

Always review your creatives before publishing.

Otherwise, you might discover your product has magically grown an extra handle or changed colour overnight 😕 Not exactly the type of A/B test we’re looking for.

My advice this week:
✔️Don’t chase every new AI feature.
✔️Choose one.
✔️Learn it well.
✔️Save one hour every week.
➡️Then move on to the next one.

Marketing has never been about using every new tool.
It’s about getting better results with the right ones.

Save this weekly digest and welcome back next Week!

Feel free to share with your colleagues and discuss in the comments your latest AI tools, both: good and annoying ones.

Last night I went to see Sting in Figueira da Foz.
Of course, the concert was incredible.
But somewhere between Englishman in New York and Every Breath You Take, my marketing brain refused to switch off.
I looked around and thought: “This isn’t just a concert.”
It’s a marketing campaign.
A sold-out event.
Thousands of people in one place.
Visitors from dozens of countries.
And, without even thinking about it, thousands of people doing the same thing:
📍 Tagging Figueira da Foz.
📸 Posting photos.
🎥 Sharing videos.
🌅 Showing the sunset.
Every Instagram Story.
Every Facebook post.
Every location tag.
Every hashtag.
They all became tiny ads for the city.
The best part?
Nobody was paid to create them.
That’s the power of experiences.
People don’t share cities.
They share moments.
And when those moments are good enough, the marketing continues long after the stage is gone.
That’s why I believe city marketing isn’t just about attracting tourists.
It’s about creating experiences worth talking about.
Because attention is expensive. Recommendations are priceless.
Congratulations to everyone who made this event happen.
Now comes the interesting part.
How do we turn one unforgettable night into thousands of reasons for people to come back?
That’s where marketing really begins ;-)