Sign up | Read online

Hey, it’s Jochem. Welcome back to the Ecom Handbook #12.

Quick question: is there anything I can do better with these newsletters?
Just hit reply and tell me. I read every message and it directly shapes what I write next, my goal is to make each edition more useful than the last to help you grow your online store with actionable tips.

-Jochem (@iamjochems)

There’s a quiet pattern I keep seeing.

Smart operators, strong brands, with solid budgets. Still frustrated with performance.

Not because automation doesn’t work.
But because they’re still trying to outmaneuver it.

Algorithmic ads aren’t tools you control, they’re systems you collaborate with. And collaboration requires a different posture.

Let’s break down how to actually work with them.

Previous mindset:

New mindset:

Advantage+, PMAX, auto-bidding stacks, they are prediction engines. They answer one question repeatedly:

Your job is not to micromanage the answer. It’s your job is to design the environment the answer emerges from.

That’s a very different skill.

Algorithms operate within boundaries.

Budget caps.
Conversion goals.
Event priorities.
Geographic scope.
Creative pool.

If results are unstable, it’s often because constraints are unclear.

Example:

The machine isn’t wrong. It’s maximizing what you defined. Clarity of constraints determines clarity of output.

Algorithmic systems need density.

Too many campaigns = diluted data.
Too many goals = diluted learning.
Too many exclusions = restricted exploration.

Consolidation isn’t laziness, it’s fuel.

If you want better performance:

Every major edit resets learning.

Budget spikes.
Creative swaps.
Event changes.
Structure rebuilds.

If you constantly “improve,” you constantly destabilize.

Working with algorithmic ads means:

The machine needs space to converge.

Platform ROAS is an output.
Not a strategy.

Algorithmic campaigns can look volatile inside the ad manager, zoom out.

Look at:

When automation works, it improves the system.

Not always the screenshot.

Run an Algorithm Alignment Audit, but not inside your ad account. Run it in your leadership team.

Ask these 5 questions in one meeting:

Then, and only then, adjust the ad account. Because most performance instability isn’t algorithm failure, it’s strategic ambiguity.

Automation amplifies clarity, it also amplifies confusion.

Before restructuring campaigns, use AI to pressure-test your logic.

Prompt:

Paste:

If the answer exposes trade-offs you hadn’t articulated internally,
that’s leverage.

AI is strong at modeling consequences, use it before you move money.

1. Feedonomics - Product feed optimization for Google and retail media
https://feedonomics.com/

PMAX performance is heavily influenced by feed structure.
Clean titles, attributes, and segmentation materially impact allocation.

2. Haus - Geo-based incrementality testing
https://www.haus.io/

When automation scales, incrementality questions grow.
Geo testing helps validate true lift beyond platform reporting.

AI‑driven search urgency for Shopify brands

A digital marketing agency warns that ecommerce merchants (especially on Shopify) need to adapt to AI‑powered search and recommendation systems to maintain visibility.

Read it here

Shein banned on university devices

The University of Texas at Austin has banned access to Shein and certain other Chinese‑owned e‑commerce apps/sites on institutional devices and Wi‑Fi as part of compliance with state directives.

Read it here

Mailchimp upgrades boost e-commerce marketing

Intuit rolled out major enhancements to Mailchimp designed to expand e-commerce marketing automation and analytics capabilities for sellers worldwide.

Read it here

After reading this edition of the newsletter, write one page.

Not in your ad account.
On paper.

Title it:

“How Our Algorithm Wins.”

Answer, in plain language:

• What economic outcome are we optimizing for?

• What trade-offs are we explicitly willing to accept?

• What signals matter most?

• What does success look like in 6 months?

• What will we ignore, even if it creates short-term noise?

If this isn’t documented,
your automation is operating on assumptions.

And assumptions create volatility.

Once it’s written:

• Share it with your leadership team.

• Align on it.

• Then check your campaigns against it.

If the constraints match the strategy, keep them. If they don’t, fix the inputs, not the algorithm.

Because the edge in algorithmic advertising isn’t better button pushing. It’s clearer thinking, and clearer thinking compounds.