Mobile-first web design means the phone is the starting point of the design and not the check at the end. The typical mistakes happen when a page is designed on a big monitor and then squeezed down to phone width.
Then buttons are too small, menus respond only to the mouse, the header fills half the screen, and a slider hides the most important message. You can find all of these mistakes with a simple test on your own phone.
Mobile-first is recommended everywhere and still often misunderstood. It does not mean that you also take a look on the smartphone at the end. It means the narrow view comes first and everything else follows.
Why do so many still design on the big screen?
Because it is more comfortable. A large monitor leaves room for wide menu bars, big images and effects that appear when you move the mouse over them. While designing, that feels generous.
Afterwards the whole thing is "scaled down" for the phone, and that is exactly where the flawed thinking lies. A desktop layout squeezed into 375 pixels of width is not a mobile layout, it is a compressed desktop layout. The difference sounds academic, but it explains why so many phone pages feel wrong.
Starting with the narrow view, by contrast, forces you to decide from the beginning: what is really important, what comes first? A small screen forgives no decoration. I think that tightness helps, because it makes the page commit.
Which mistakes should you know in mobile-first web design?
The following four mistakes can be spotted without any expertise. How often they occur on your site, only the test further down will show.
1. Buttons you have to aim at instead of tap
A finger is not a mouse pointer. On web.dev, Google recommends a minimum size of about 48 device-independent pixels for touch targets and roughly 8 pixels of spacing to the next target (web.dev). The accessibility guideline WCAG 2.2 sets a lower limit of 24 by 24 CSS pixels at level AA, with exceptions such as links within running text (W3C). If the "Call now" button is much smaller and sits right next to a second link, the thumb easily misses. The phone number in particular should be large enough and stand free.
2. Hover menus without a touch alternative
On the desktop, a submenu opens when the mouse moves over it. On the phone there is no hovering, only tapping. If the main menu works only by hover, a phone user cannot reach the subpages. Mistakes like this stay undiscovered for a long time if everyone internally works on the desktop and nobody uses their own site on the phone.
3. Giant headers that eat half the screen
A fixed header with a logo, a slogan and three lines of navigation looks smart on the desktop. On the phone, a header like that can take up a large part of the visible area before the visitor sees any content at all, and it sticks while scrolling. Check how much room is left for the actual content after loading, and shrink the header while scrolling or let it scroll away with the page.
4. Important things end up outside the thumb zone
Anyone holding the phone in one hand and using the thumb reaches the lower part of the screen more easily than the top edge. If the only call to action sits at the very top right, you have to change your grip. A fixed call or enquiry button in the lower area is often the more comfortable solution on a phone.
Does a slider belong on the mobile homepage?
Usually not. An automatically rotating image slider at the very top pushes the messages onto slide two and three, and only those who wait see them. On the phone, the space at the top is scarce anyway.
An often-cited analysis comes from a web developer at the University of Notre Dame in 2013: on the university's homepage, about one percent of visitors clicked an element of the slider, and 84 percent of those clicks went to the first position (Erik Runyon). Those are old figures from a university website, not a rule for your business. In the same year, the Nielsen Norman Group described from a user test that auto-forwarding carousels annoy users and reduce the visibility of content (Nielsen Norman Group).
Then there is the technology. A slider often loads several large images and an extra script, in exactly the area that should be visible first. What belongs at the top instead:
- a fixed, meaningful image instead of changing slides,
- a headline that names the benefit and not just the company name,
- a clearly visible button with a clear goal: call, enquire or book an appointment.
If you have several messages, stack them one below the other instead of one behind the other in slides. Movement is not forbidden. It makes sense when the visitor controls it, for example by swiping through an image gallery further down, or when it underlines a fixed message instead of replacing it. Whether your slider is used shows in your analytics tool: if clicks land almost only on the first slide, the rest contributes little.
How many fonts and colours can a small screen take?
Fewer than you think on the desktop. On the phone, headline, body text and buttons lie close on top of each other, and every extra font or colour makes the picture busier. Our rule of thumb, not a study: two fonts and three colours are enough for most company sites.
One font for headlines, one for body text, or a single font in two weights. Plus a main colour for the brand, a calm text colour and an accent colour for buttons and links. The accent colour draws the eye to where it should go. If every button has a different colour, none stands out any more. Also make sure there is enough contrast between text and background; outside in sunlight, every weakness shows immediately.
Clutter rarely arises on purpose. It grows when different people add subpages and campaigns over the years, each time with whatever was at hand. A written-down mini rulebook helps: two fonts, three colour codes and a note on when which button is used.
A mobile-first test that non-technical people can manage
You do not have to be a developer to find the roughest mistakes yourself. Take your own phone and proceed like this:
- Test one-handed: hold the phone in one hand, as in everyday life. Can your thumb reach the phone number, the contact form and the menu?
- Do not zoom: forbid yourself from pinching with two fingers. If you have to zoom anywhere to read something, the text is too small or the layout too wide.
- Five seconds: open the homepage and count to five. Afterwards, do you know what the company does and how to get in touch?
- Poor reception: switch off Wi-Fi and load the page over mobile data, best of all on the move on the S-Bahn towards Munich. If it takes long, the page is too heavy.
- Click through: open the homepage, a services page and the contact page one after the other. Do the three look like the same website, with the same fonts and buttons?
Run through it on an iPhone and on an Android device if available, because the two display some things differently. And let someone try it who has never seen the page. You yourself know where everything is, after all.
An assumed example shows how quickly such a test finds something: a restaurant puts its menu online in two columns. On the monitor it looks tidy, on the phone the columns are squeezed into 375 pixels, and to read the price of the pork knuckle you have to zoom several times. The fix would be a single-column menu with larger type and categories that fold open.
Why does the phone view matter to Google too?
Because for indexing and ranking, Google uses the mobile version of a website, fetched with the smartphone crawler. Google calls this mobile-first indexing (Google Search Central). Content that is missing on the phone may not be seen by search at all.
How many of your visitors come on a phone, you can see in your own statistics, split by device type. If it is the majority, the phone view is your actual website and the desktop is the second version.
How can you have the website tested automatically?
You have the website tested automatically by having a script open the most important pages at phone width after every change, measure them and tell you when something has got worse. The manual test above remains useful; the automation only makes sure that no new photo, no new plugin and no new block of text quietly worsens the phone view.
Here is how to go about it:
- Decide which pages are checked: the homepage, the most important services page, contact.
- Have these pages measured with Lighthouse in the mobile view after every change or once a week. For continuous operation there is Lighthouse CI, a collection of tools from Google that runs measurements regularly, stores them and checks them against set thresholds (GitHub). What Lighthouse measures is explained in our glossary under Lighthouse.
- Add screenshots at phone width. A script can open each page in a browser without a window, save an image and compare it with the last version. If something shifts noticeably, you get both images side by side.
- Decide for yourself. The automation reports, the person judges. Not every deviation is a mistake.
Say a new reference photo goes onto the homepage at Muster Joinery. The weekly test then shows that the button to the enquiry form has slipped below the image on the phone, and sends the screenshot along. That is fixed in a few minutes, but only if someone notices.
Frequently asked questions
Do I have to rebuild my website completely for mobile-first?
Not necessarily. Buttons that are too small, a header that is too tall or a slider can usually be fixed one by one. A rebuild is more worthwhile if the basic framework of the page does not work on the phone.
Is it enough if the page is responsive?
Responsive only means the layout adapts to the screen width. Whether the page is easy to use on a phone only shows in the test with your thumb.
At bytebrise we build and rent out websites for small businesses and set up processes that take over recurring office work. The automation is new for us and does not yet run permanently at any business. Every process starts as a prototype at a fixed price from €1,500 with four weeks of testing in your business, followed by a monthly flat fee, cancellable monthly. The overview is on the bytebrise GO page under workflows. The rented website costs €490 one-off to start and then €149 a month, maintenance and updates included, cancellable at the end of any calendar month with no notice period (pricing).
We accessed all the linked sources on 25 September 2026.