The system

How a shop has to be built if it is going to sell.

The Semantic Performance System describes which information, signals and structures a shop has to provide so customers can make a confident purchase decision. Not by taste, but against a rulebook that is openly documented and verifiable.

The sequence

The eight steps of a purchase decision

  1. 1Recognise the productWhat is this, who is it for, what problem does it solve?
  2. 2Create desireWhy should anyone want it?
  3. 3Fit, trust and contextDoes it suit me, can I trust the brand?
  4. 4Choose variant and sizeIs the choice a help or a hurdle?
  5. 5The act of buyingIs the buying moment clear, stable and trustworthy?
  6. 6Detail and value justificationDoes the decision hold up rationally too?
  7. 7UpsellDoes it complete the decision or interrupt it?
  8. 8Reinforce brand and trustDoes the brand come across as credible and relevant?

1Recognise the product

The first step is brutally simple: customers have to understand immediately where they are and what is on offer. What is the product? What is it for? Who is it relevant to? What problem does it solve?

Many shops lose clarity right here, because they lean too hard on mood, branding or visual staging. That can look premium, but it does not help if visitors cannot place what they are looking at, and why it might matter to them, within a few seconds.

2Create desire

Once the product has been recognised, it has to become desirable. Users need to understand why they should want it. This is not only about features, but about benefit, outcome, identification and emotional relevance.

Good product pages do not just show what a product can do. They show which better feeling, which status, which simplification or which result the purchase makes possible. Without that desire the shop stays informative but unconvincing. That holds across every industry.

3Fit, trust and context

Desire alone is not enough. Users have to be able to check whether the product really suits them. That takes clear information on audience, use, size, material, effect, compatibility, usage situation and possible limitations.

At the same time, trust signals have to become visible: reviews, guarantees, shipping, returns, payment methods, certifications, provenance, genuine customer voices or product promises you can verify. What matters is not whether this information exists somewhere, but whether it appears exactly where it has the greatest impact.

4Choose variant and size

Choosing variants, sizes, colours, bundles or quantities is not a technical detail. It is a central moment of the purchase decision.

Many shops treat variants like simple select fields. That forces users to jump back and forth between options to understand price differences, availability, discounts, properties or sizing. Good variant logic makes differences visible, reduces the work of comparing and helps people pick the right option faster.

5The act of buying

The buying moment has to be clear, stable and trustworthy. Primary CTA, price, availability, delivery information, payment methods and any remaining uncertainty all have to work together cleanly in this area.

This is exactly where no unnecessary friction can appear. Broken hover states, unclear disabled states, missing feedback after a click or contradictory information about price and availability damage trust in the very moment users want to buy.

6Detail and value justification

After the first intent to buy, many users look for confirmation. They want to understand whether the price is justified, which details make the product special and whether the decision holds up rationally.

This is where materials, construction, technical detail, ingredients, provenance, quality, sustainability, care, use or comparative advantages matter. That information should not read as an arbitrary block of text; it should justify the value of the product and reduce the doubts still open.

7Upsell

Upsells only work well when they support the original purchase decision. They must not feel like an intrusive add-on sale; they have to create value you can follow.

Good upsells complete the use case, increase the benefit or help someone buy a more complete solution. Bad upsells interrupt the decision, feel arbitrary or add cognitive load at the wrong moment.

8Reinforce brand and trust

After the concrete product decision, the brand can build trust, meaning and differentiation once more. This is about provenance, values, expertise, community, press, founder story, quality assurance or long-term credibility.

This part does not have to replace the actual purchase decision. It should reinforce it. Users should not only think “the product fits”, but also “this brand feels credible, relevant and trustworthy”.

Why this structure matters

Many online shops contain almost all the necessary information, but not in the right order, not in the right context, or not with the right visual weight. That leaves users to assemble the purchase decision themselves.

The job of a shop is to structure product understanding, desire, trust, choice, the act of buying and confirmation so that interest can turn into a confident decision to buy.

The quality dimensions

The five analysis areas

The eight steps say what a purchase decision needs, in order. The five areas say how well each individual step is served.

Decision Clarity

Whether users understand, at every point, what they are seeing, what it means for them and what the next sensible step is.

Trust & Decision Enablement

Whether the signals that secure a decision exist, and appear exactly where the doubt arises.

Semantic Integrity

Whether structure, language and markup say the same thing as the visible page, with no contradiction between content and code.

Machine Extractability

Whether search systems, recommendation engines and agents can reliably read out the relevant information and assign it correctly.

System Robustness

Whether the implementation holds under real conditions: loading behaviour, states, variants, error cases and updates.

People and systems

Your shop stopped being read only by people a while ago.

Search systems, recommendation engines and increasingly AI agents have a say in which products are visible and get bought. When structure, content and data do not line up cleanly, products are interpreted wrongly and lose visibility before a user has even clicked. The system makes sure your shop works not only for people, but is also correctly understandable to machines.

Relationship to the guidelines

Three levels

  1. Eight steps

    The sequence: what a purchase decision needs, in order

  2. Five areas

    The quality dimensions: how well each step is served

  3. 81 guidelines

    The concrete, checkable rulebook underneath, openly accessible

The guidelines are general and deliberately live on their own domain. The Semantic Performance System is the application of that rulebook to one specific shop.

Go to the Open E-Commerce Guidelines →

Two applications

How the standard gets used

Revenue from the traffic you already have

Your shop is running. We measure it against the standard and build in the changes with the greatest leverage.

Read more →

Delivery Assurance

You are building. The standard is set before the build, so the experience holds at launch.

Read more →

Getting started

Free assessment within 48 h

Send me your shop and you get a first assessment within 48 hours. Free, and with no strings attached.

No spam. Your details are sent via Formspree and used only to get back to you. How your data is handled