Signal Stories.

Margin is invisible to most algorithms. That creates expensive growth.

Algorithms optimize for what they can see. If margin is not part of the signal system, the algorithm cannot optimize toward profitable business outcomes.

Products are no longer just inventory. They are instructions to the algorithm.

In AI-driven commerce, products are signals. We help companies structure products around profitability, demand, and strategic value so algorithms scale the right things.

Untangle Collective & Berghs School of Communication: Performance Marketing Strategy

An intensive 8-week remote course for professionals who want to take ownership of data, and profitable digital marketing in the AI era. Next start: October 19, 2026. The course is in Swedish.

Context engineering: the groundwork that makes AI useful

Prompt engineering without context engineering is like navigating without a map. AI doesn't just need instructions, it needs understanding, and that comes from well-structured data.

Invest as if you were already 30% bigger

With 50% churn, more than half of your 2027 revenue will come from customers you don't have yet. The growth budget you need already exists, inside your current performance spend.

Are you giving AI the right conditions to make business decisions?

AI already drives large parts of your marketing, but do you actually know what it is optimizing for? A Feed Prompt Strategy gives algorithms the business context they're missing.

The hidden waste in performance marketing

At SEK 1M/month in spend, even 15 percent of misaligned signals quietly costs SEK 1.8M a year. It rarely shows in ROAS. It shows in margin.

You are probably scaling the wrong signals

Most ecommerce brands optimize against metrics that platforms reward, not signals that reflect the business.

What a Cut Waste review actually reveals

Where money leaks in an ecommerce P&L, why teams miss it, and what changes when you see it for the first time.

AI amplifies whatever you give it

AI does not fix weak signals. It scales them faster. A short note for CFOs and CMOs sponsoring AI investments.

Scale what actually compounds

Once waste is out, reallocation is the next decision. A short framework for scaling the parts of the business that compound.

POAS: the metric that actually runs the business

ROAS tells you the ad worked. POAS tells you the business worked. Here is how to move from one to the other without breaking your stack.

Marketing Mix Modeling without the Coca-Cola budget

MMM is no longer a six-figure consulting project. Here is the lean version that works for a mid-sized ecommerce brand, and the three traps to avoid.

Return on ad spend is a vanity metric

ROAS made sense when margins were comfortable and returns were rare. Neither is true in 2026. Here is what to put in its place.

Marketing efficiency is a CFO conversation, not a CMO one

When marketing efficiency lives only in the marketing team, it gets defended. When it lives in the finance review, it gets fixed.

Incrementality testing: the cheapest truth in marketing

Most channels are taking credit for sales they did not cause. A simple incrementality test settles the argument in two weeks for less than a campaign budget.

Media efficiency is a leadership team question

Media efficiency stops being a media problem when leadership starts asking it. Three questions that put it back on the agenda where it belongs.

Profit on ad spend: the five-minute self-test

A short, honest self-test for any ecommerce leader who suspects their ad budget is earning less than it looks like it is. No tools, no consultants.