There's a paradoxical observation worth making at the outset, because it overturns the way many blogs tell the user experience. The usual narrative presents UX as the art of retaining visitors — the more time they spend on the site, the more they "engage," the more they turn into loyal customers. The traditional metrics reinforce this idea: average time on site, number of pages viewed, depth of navigation.
But this idea, which had a foundation in a certain phase of the web, is progressively less useful for describing what makes a site truly effective for its users. A site that works well is often a site that lets the user do what they came to do in the shortest possible time, with the least possible friction, and then lets them leave satisfied. Artificially retaining the user, complicating the paths to increase pages viewed, hiding information behind tortuous flows to make them "explore" the site — these are all practices that produce apparently better metrics and actually worse experiences.
This change of perspective has practical consequences for design. A well-designed site responds quickly to the user's implicit question: "why am I here and how do I do what I need." A poorly designed site tries to retain the user, to direct them toward paths that serve the site more than them, to build an experience the site would like but that the user doesn't ask for.
It's worth articulating what it really means to design effective UX, distinguishing the practices that produce value for real users from those that produce value only for corporate metrics.
The starting point: understanding what the user really wants
The fundamental question of UX, before any design choice, is just one: what does the person who arrives on this site really want? The answer isn't obvious, and it's rarely the one the site owner assumes.
The user who arrives on an e-commerce site doesn't want to "navigate the purchasing experience" — they want to find a specific product (if they know what they're looking for) or explore to discover something interesting (if they're looking for inspiration), buy it, receive it, and get back to their life. The user who arrives on a services site doesn't want to "discover your value proposition" — they want to understand whether your service solves a specific problem of theirs and, if so, how to start working with you.
Understanding what the user really wants requires specific methodologies, and it's an activity many companies underestimate. The practices that produce results are well known: research with real users, usability testing on prototypes before developing the final site, analysis of actual behavior on the existing site, systematic collection of feedback. These are activities that require investment but that produce measurable results.
The common mistake is designing based on internal assumptions — what management thinks users want, what designers find elegant, what agencies propose as the trend of the moment. Almost always there's a significant gap between these assumptions and what users actually do and desire. The gap is reduced only with direct research, not with intuition.
Visual hierarchy, and what it really means
One of the central technical skills of UX is visual hierarchy — the organization of the elements of a page so that the user immediately understands what's most important, what comes first, what they can ignore.
Well-made visual hierarchy isn't a matter of making the site "aesthetically pleasing." It's a matter of guiding the user's eye through the page so that the important information is noticed first, the main actions are clearly available, the secondary elements don't distract from the priorities.
Concretely, this means specific choices: text sizes that actually reflect the importance of the content, color contrast that brings attention where needed, white space used to separate and give breathing room instead of filling every corner with elements, buttons and calls-to-action visible without being aggressive.
The common mistake is the overloaded page. A site that tries to show everything at the same time — all the products, all the value propositions, all the awards won, all the customers, all the possible actions that can be taken — ends up communicating nothing clear. The user who arrives doesn't immediately understand what's a priority, and in the absence of a clear visual guide tends to abandon.
The operating rule is to design by exclusion: what is it actually essential that the user sees right away? Everything else can be accessible through specific paths, but it must not compete for attention with the main elements.
Navigation, and why it's often too complex
A site's navigation is one of the elements most frequently handled badly. The natural tendency, when designing a site, is to organize everything into multiple categories with subcategories, build menus with many items, provide for specific paths for every possible user need. The result is often a complex navigation structure that users can't use well.
The navigations that work have some specific characteristics:
They're flat where possible. When the user can reach what they're looking for in two or three clicks instead of five or six, they do it more willingly. Deep hierarchical navigation structures discourage exploration.
They use the users' language. Categories are named with the terms users actually use to search for those things, not with the company's internal terminology. A company that sells "industrial automation solutions" might discover that its customers search for "packaging machines" — using the corporate language instead of the users' is one of the most frequent causes of low engagement.
They have a prominent search. For many users, especially on sites with broad catalogs, search is the preferred path over navigation by categories. A visible, fast search functionality capable of handling variants and typos is often more important than a perfectly structured navigation.
They keep the user oriented. Indications of where you are in the site (breadcrumbs), clear page titles, the possibility of going back without confusion — these are elements that reduce the feeling of being lost.
The operational approach that produces results is to simplify progressively. Start from a rich structure and then reduce it, testing with real users which level of simplification preserves functionality without losing its sense. Subtraction is almost always more effective than addition in navigation design.
Speed, a factor many companies underestimate
A site's loading speed is probably the UX element that has the most direct impact on commercial results, and it's also the one many companies treat as secondary compared to more visible aspects.
The data on the relationship between speed and conversions is robust and consistent. Pages that load in less than two seconds convert significantly more than pages that take five or six seconds. The effect is particularly marked on mobile devices, where the user's patience is further reduced. A slow site is a site that loses sales every day, invisibly but constantly.
Speed optimization requires serious technical attention. The main areas of intervention include compression and optimized formats for images, reduction of unnecessary code, the choice of a performant hosting platform, the use of content delivery networks, minimization of external dependencies that slow down loading, lazy loading for content that isn't needed immediately.
For Italian SMEs, the investment in speed is one of those with the most immediate and measurable return. Often the most relevant optimizations are obtained with relatively contained technical interventions — image compression, elimination of non-essential plugins and scripts, hosting upgrades. It's an investment that pays off in concrete commercial metrics in the short term.
Responsive design is no longer enough
For years the key concept in web design was responsive design — a site that automatically adapts to any screen. It has become an industry standard, and it's by now assumed as a starting point.
What's worth articulating is that responsive design is no longer enough. A responsive site works on all devices, but "working" doesn't mean "being optimal." A site designed as a desktop experience and then adapted responsive for mobile produces mediocre mobile experiences, even if they technically "work."
The contemporary approach that produces better results is mobile-first design — designing the experience for mobile first, then adapting it for larger screens. It's a conceptual inversion that has practical consequences: layouts designed for vertical navigation, information hierarchies that work in the limited space, interactions designed for touch, performance optimized for mobile connection conditions.
For most contemporary commercial sites, the share of mobile traffic is the absolute majority. Continuing to design desktop-first means prioritizing a minority of users against a majority, and the results reflect it.
Accessibility as a dimension of UX
A site's accessibility — the ability to be used by people with different disabilities — is progressively treated as an integral dimension of the user experience, not as a separate theme.
The European regulatory framework has reinforced this integration. The European Accessibility Act, which came into force in 2025, has extended accessibility obligations to a broad perimeter of companies and commercial sites. For Italian SMEs that aren't micro-enterprises, accessibility is today a compliance requirement, not a strategic option.
Beyond the regulatory dimension, accessibility has concrete commercial implications. The practices that make a site accessible for people with disabilities — well-structured code, clear hierarchies, adequate contrasts, alternative descriptions for images, keyboard navigability, contained loading times — are largely the same practices that improve the experience for all users and that are rewarded by search engines.
Designing an accessible site from the start is significantly less costly than adapting it later. For companies evaluating significant investments in their site, integrating accessibility into the starting requirements is an operational choice that produces advantages on multiple dimensions.
Content, where UX meets copywriting
A dimension of UX that's often treated as a separate theme — but that is in reality central — is the quality of the content.
A site can have the most elegant design in the world, but if the content is mediocre the user's experience will be mediocre. Content includes the texts that explain what the company does, the product descriptions, the calls-to-action, the error messages, the support pages, the transactional emails. These are all elements the user reads and that influence their perception and their decisions.
The characteristics of content that works well in UX are identifiable:
They're written for the user, not for the writer. They focus on what interests the reader, not on what the company wants to say. The difference between "We've been the industry leader since 1985 with our high-quality products" and "Find the right product for your specific problem" is the difference between a company-centered site and a user-centered one.
They're concise. More than ninety percent of web users don't read linearly — they scan, skip, look for specific points. Long, dense texts are texts that don't get read. Conciseness, paradoxically, communicates more than verbosity.
They use headings and subheadings to facilitate scanning. A well-structured page can be "read" in ten seconds to understand what it's about, leaving the user the choice of going deeper into the sections that interest them.
They have clear and specific calls-to-action. "Learn more" is a weak call-to-action. "See the products for pastry-making" is specific and functional. The specificity of the proposed actions is one of the factors that most influences conversion.
They're honest. Content that overstates, that promises too much, that uses aggressive promotional language, often generates resistance in the modern user. Honest specificity about what the company offers, about what its limits are, about who it's suited for, builds trust in ways that traditional marketing can't produce.
The metrics that really count
A question worth addressing is which UX metrics are useful to monitor. The traditional metrics — average time on site, pages viewed per session, bounce rate — are still widely used, but their meaning is less clear than is thought.
A long time on site can mean deep engagement or confused navigation. A high number of pages viewed can mean interest or difficulty finding what you're looking for. A low bounce rate can mean interest or a site structure that forces you to click for every piece of information.
The metrics that actually tell the quality of the UX are more specific and more results-oriented:
Task completion rate. How many users who start a task (buying, registering, contacting) complete it? This is probably the most important metric, because it directly tells the site's effectiveness in serving its users.
Time to complete key tasks. How much time does it take to do the main things the site exists for? If it decreases over time, the site is improving. If it increases, something is getting worse.
Error rate. How many users make errors during the main flows? Frequent errors indicate design problems worth solving.
Return-visit rate. How many users return to the site after the first visit? For many businesses, it's a metric more revealing of real value than absolute traffic.
Net Promoter Score or equivalents. How willing are users to recommend the site to others? It measures the overall perceived quality in a way that behavioral metrics don't capture.
For Italian SMEs, building a dashboard of meaningful UX metrics — rather than distractedly monitoring the standard metrics — is an investment that produces informed decisions on the improvements to make.
What AI tools have changed for UX
Several aspects of designing and managing the user experience have been transformed by AI tools in ways that deserve to be named.
Personalizing the experience. AI systems make it possible to adapt content, recommendations, navigation paths according to the single user with a sophistication that five years ago was reserved for the large platforms. For sites that have sufficient volumes, personalization produces significant improvements in conversion rates and in perceived satisfaction.
Behavior analysis. AI tools can analyze quantities of behavioral data that human work alone couldn't process. They identify patterns, anomalies, opportunities for improvement, suggest hypotheses to test. For companies that want to improve their site iteratively, it's a level of operational intelligence that amplifies the effectiveness of the investments.
Generating variants for testing. A/B tests require producing variants of pages, layouts, content to compare. AI tools significantly accelerate this production, making it possible to test more hypotheses in shorter times.
Integrated conversational interfaces. The chatbots and AI assistants integrated into sites are qualitatively different from those of a few years ago. They can actually understand requests in natural language, access site-specific information, handle complex conversations. For many sites, integrating a quality AI assistant means offering a level of support that significantly improves the experience.
Automatic optimization. Systems that automatically test variants, optimize elements based on behaviors, suggest continuous improvements. For companies that have limited internal resources dedicated to UX, it's a capability that makes it possible to keep the site optimized without continuous investments in dedicated staff.
AI doesn't replace the strategic design of UX. The fundamental decisions about what the site must do, for whom, how — remain human. But AI significantly amplifies the effectiveness of a well-designed site and reduces the chances of a mediocre one.
Effective user experience isn't an obscure art accessible only to specialists. It's an applied discipline that combines knowledge of users, technical competence, attention to detail, the ability to iterate continuously. Italian companies that invest seriously in UX build sites that actually serve their users and that produce consistent commercial results.
For SMEs that want to improve the UX of their site without necessarily rebuilding it from scratch, the practical thing to do is to begin with structured observation. What do users actually do on the current site? Where do they abandon? Which paths work and which don't? Which task-completion metrics can be improved with targeted interventions? Often the most relevant improvement opportunities emerge from specific observations, not from complete redesigns.
UX is one of those areas where perfectionism is the enemy of the good. A site that improves by five percent every quarter, through targeted interventions based on data, amply surpasses in the medium term a perfectly designed site that's then never touched again. The ability to observe, understand, improve iteratively is more important than the brilliance of the single initial project.