We publish the decisions that cost us something to learn.

01Buyer guide / Website redesign

Before a website redesign, decide what needs to work better.

A dated appearance can be a useful warning. To choose the right project, connect it to a specific problem: customers cannot understand the offer, find the right information or reach your team. This guide helps you write that brief and compare proposals.

02Industry guide / Automotive sales

A dealership website has to keep its stock and its sales conversation connected.

A car is a changing, individual item. The useful website helps a buyer identify it, understand what is known about it and start a conversation that retains that context. Here is how to plan that journey, using our Delux work and external examples.

03Industry guide / Professional equipment

Help the buyer identify the right equipment before asking them to request a quote.

A surgeon looking by specialty and a technician looking for a known handpiece arrive with different questions. An equipment website needs to serve both. Our DGM and Tiroj work shows why discovery, product identity and the contents of an offer need separate attention.

04Supplier website guide / Tiroj development example

Designing a dental and lab equipment website: from model comparison to a useful inquiry.

For equipment suppliers planning a website, the hard part is connecting the unit, handpiece and requested package without losing their meaning. Using Tiroj, which is still in development and is not live yet, and manufacturer sources, we show how to structure those comparisons and carry the choice into an inquiry.

05Multilingual websites

A language switcher is not a multilingual architecture.

A real multilingual website needs complete routes, clear language relationships, maintainable content, and one operational truth behind every version. The Delux project made those constraints concrete.

What we saw

Three language layers. Three content models. Three sources of truth. Every update cost translation, QA, and trust.

What we chose

One shared content architecture. Language became a view, not a system. One operation, three voices.

What you can use

Map your language layers. Collapse logic into one spine. Localize the edges, not the foundation.

06Internal software

Another subscription is not always the cheaper decision.

Buying software is sensible when the product fits. When the team keeps rebuilding the same workaround between disconnected tools, a focused internal system can become the smaller long-term choice.

What we saw

The same handoff happened over and over. Different tools. Same fixes. Time leaked in the loop, not the work.

What we chose

Buy what is stable. Connect what is necessary. Build only what is unique. Break the loop.

What you can use

Log the loop. Calculate the leak. Remove it before you commit to building.

07Motion and performance

Animation should explain the interface, not hold it hostage.

Motion-rich websites can remain usable and search-friendly when content survives without animation, mobile gets its own constraints, and performance is treated as part of the design decision.

What we saw

Motion was treated like decoration. Frames dropped. Weight added up. Performance was not a budget—it was luck.

What we chose

Set the frame budget. Design to it. Ship when every movement earns its cost.

What you can use

Define the budget. Profile the real path. Remove what does not move the outcome.

08Editorial rule

Write what the work taught us—not what a keyword tool requested.

Every article begins with a decision that appeared in real work. We publish when the explanation can help a founder, operator, designer, or engineer make a clearer choice.