Correcting Typography Mistakes That Undermine Professional Design: Too Many Fonts, Faux Bold, Bad Kerning, and Widows

The fastest way to make a layout look amateur is not a bad color choice or a weak image. It is typography. Every viewer registers the failure as "something off" without being able to name it. The common wrong assumption is that type fails because the designer had no taste. In fact, the failures are specific, mechanical, and almost always the result of skipping a production step that a professional would do by reflex. Each one has a fix. The fix is not "be more artistic". It is a named operation in a piece of software or a character you replace. Here is the full catalogue of the typography mistakes amateur design makes, in the order you will most often meet them, with the exact correction for each.

The Real Cost of Every Extra Family

Every font family you add to a document is a cost. It carries a license, a file weight on the web, a set of OpenType feature tags you may never enable, and a baseline grid you must manually check across weights. The measurable benefit of the fifth family on a page is close to zero. The measurable cost is exact. The research and the practice agree: two families are enough. One for display, one for body text. If you genuinely need a third for a caption or a pull quote, you may have it, but you should be able to defend it in a meeting. Ask yourself what the third family does that the first two cannot. If the answer is "it looks different", that is not a reason.

Cognitive Load Is the Hidden Tax

The professional standard is not a rule about beauty. It is a rule about cognitive load. Every new typeface classification, from a humanist sans to a transitional serif, forces the reader to re-identify the letters. That re-identification is a tiny tax on every line. Two families keep the tax at zero.

What the Software Does When You Press B Without Buying the Font

Open your font menu. Select the body text family. Now press the bold button. If your chosen weight is not installed, and for many free or single-weight families it is not, the application does something mechanical: it takes the regular outlines and strokes them algorithmically. The result is not bolder type. It is distorted type. Stroke contrast changes, the counters close up, and the letters start to look dirty at small sizes. The same happens with italic: the software shears the glyphs instead of using the designed slant, producing letters that lean but do not flow. This is the faux bold and italic problem in its purest form.

The Only Correct Fix

The fix is always the same: buy or download the actual weight. If you only have the regular, and the project requires bold, your only correct move is to switch to a family that includes a true bold. No "bold" button. On the web, the failure appears when you request a font-weight of 700 but the font file only carries a 400. The browser synthesizes the bold, and the page looks cheap in a way your non-designer client will not be able to articulate. The production test is simple: select the text, look at the font menu, and see if the weight is listed as a separate family or as a style within the family. If it is a style, you are safe. If it is a button, you are not.

The Difference Between a Type Designer's Work and an Amateur's Guess

The space between a specific pair of characters is the proportional adjustment called kerning. The letters "AV" need less space than "TT" because the outlines fit differently. Every professional typeface ships with a table of these pair values, built by the designer over years. That table is called metrics kerning, and it is the correct default for body text. When you open software and the spacing is set to zero, you are looking at the metrics table in action. The alternative, optical kerning, ignores the pair table entirely. It adjusts space by looking at the shapes of the outlines and trying to equalize the visual area between them. That sounds reasonable, but on a well-kerned typeface it causes uneven spacing, because a good metrics table has already done the hard work on every pair that matters.

When to Use Each Method

The rule is: metrics kerning for text below 10 pt, optical kerning only for display type where the metrics table was never tuned for that size, and even then only as a starting point. If you are hand-kerning a headline, you are either a type designer yourself or you are wasting time that the foundry already spent. The failure case is the amateur who discovers the spacing slider, selects the headline, and pushes it to -2% or +5% without looking at the pairs. The result is a headline where "To" and "You" have visibly different gaps. That is bad spacing, and it is the fastest single-character give-away of typography mistakes amateur design keeps making.

Spotting the Two Most Common Text Strandings

A widow is a short line at the top of a column, stranded from the paragraph it belongs to. An orphan is a single word left alone on the last line of a paragraph, or a line of a new paragraph stranded at the bottom of the previous column. Both look like a failure of attention. They are. The fix is not a font change. It is a reflow. You adjust the spacing, even by a fraction of a percent, to pull the text back a line. You change the hyphenation zone in justified text so one more word fits on the previous line. You add a non-breaking space between the last two words of a paragraph to keep them together.

Exact Numbers You Can Act On

On the web, set the CSS property that controls widow and orphan lines to a value greater than 1, and most browsers will respect it. The minimum number of orphans to avoid is one word on the last line. The minimum number of widows to avoid is one line at the top of a new column. A rag, the uneven right edge of left-aligned text, should not have a depth variation of less than 10% to 15% of the line length. A shallow rag looks like a mistake. The practical detail: if you are laying out a print piece, build your style so that the last line of every paragraph has at least two words. If you cannot, you reflow. The failure case is the designer who blames the font. It is not the font. It is the measure, and the measure is your responsibility.

Double Spaces, Straight Quotes, and Hyphens Where Em Dashes Belong

Every typographic error in this category has a one-sentence fix, and each fix names the Unicode code point you should have used. A double space after a period is a typewriter habit from the monospaced era; fix it with a single space, which is the only space the font's metrics tables were designed around. A straight quote used as an apostrophe, the ' character, is always wrong; the correct character is the right single quotation mark (U+2019). A straight double quote is equally wrong; the correct opening and closing marks are “ (U+201C) and ” (U+201D). A hyphen where an em dash belongs, the character U+2014, is a failure of the same kind. Each of these is a Unicode problem: the character you want is a specific code point, and the character you typed is a different one that merely looks similar.

The Fix Is in the Tool, Not the Eye

The fix is not visual. It is typographic, and it is in the tool. Most word processors and page-layout applications will convert these automatically if you enable smart quotes and smart dashes. If your document exports the wrong characters, you have not turned on the conversion, and you have shipped amateur work. On the web, the failure appears when a content management system strips the smart characters and leaves the straight ones behind. The test is to export your text and search for any straight quote followed by a letter. Every hit is a mistake.

The One Setting Where the Amateur Is Convinced They Are Right

Justified text, set flush to both the left and right margins, is the most abused setting in amateur design. The reason is the hyphenation zone. In a narrow column, justified text without a hyphenation zone creates rivers: white gaps that run vertically down the paragraph and destroy the reading line. The fix is not to abandon justification. It is to set the hyphenation zone correctly. For a measure of 45 to 75 characters per line, which is the cited readable range, the hyphenation zone must be wide enough to prevent the rivers from forming. The maximum number of consecutive lines ending in hyphens is two. Beyond that, the reader notices the rag on the right. If you are setting justified text and you see more than two hyphenated lines in a row, you either widen the measure or you switch to left-aligned text with a ragged right edge.

Justification Is a Space That Must Be Earned

The failure case is the designer who justifies because it looks "clean" without understanding that the clean look is a direct function of the hyphenation zone, not the justification itself. A professional will tell you that left-aligned text is the safer choice for body copy. It never creates rivers and it never requires the hyphenation zone to be tuned. The rule is not that justified text is wrong. It is that justified text is a space that must be earned.

The Silent Killers of Multi-Column Layouts

A baseline grid is a vertical rhythm system where all type aligns to a repeating spatial interval. In a two-column layout, if the baselines in column one do not align with the baselines in column two, the eye registers a jump every time it moves across the gutter. The fix is to set your leading, the distance from baseline to baseline, measured in points in print, to a value that fits the grid, and then to set all your paragraph spacing to a multiple of that value. The maximum paragraph spacing (space after) should be equal to the leading value; anything larger breaks the grid.

X-Height Mismatch Destroys Pairings

The second silent killer is x-height mismatch. A typeface with a tall x-height, the height of the lowercase letter x, will pass a legibility test at a smaller point size than a face with a low x-height, even when the point size is identical. This is not an aesthetic judgment. It is a measurement. When you pair a display face with a body face, check their x-heights against each other. If the body face has a much smaller x-height, the pair will look unbalanced no matter how well the styles match. The web equivalent is the line-height multiplier in CSS, which is a unitless number that must be tuned per typeface, not set once and forgotten. A page that sets line-height to 1.5 for every element is leaving legibility on the table.

Variable Fonts: The New Mistake Is Assuming the File Is the Font

Variable fonts are a single OpenType file that contains multiple weights, widths, and other axes. The format is called OpenType Font Variations, and it is part of OpenType 1.8. The new mistake is assuming that loading the variable file gives you the same performance as loading two static weights. It does not. A variable font file is often larger than the sum of the two weights you actually use. It contains the outlines for the entire design space, not just the positions you need. The fix is to subset the variable font to the axes you use and to specify the axis values with the correct syntax. On the web, that means setting the font-variation-settings CSS property, and it means testing in the browsers your audience actually uses. The failure case is the site that loads a variable font but renders all text at the default axis position, because the browser does not support the syntax and silently ignores the declaration. The result is a page that looks like it failed to load the bold weight. The rule is: if you are using a variable font, you are responsible for the axes. If you are not prepared to test them, use static weights and accept the file weight.

The OpenType Features You Are Probably Not Using

Every professional typeface includes a set of OpenType features beyond the basic characters. Small caps are one of them. Lining numerals, the digits that align to the capital height, are another. Oldstyle numerals, the digits with ascenders and descenders that blend into lowercase text, are a third. The amateur mistake is to use the font's default numerals, which are lining numerals, in the middle of a paragraph of lowercase text, where they stand out. The fix is to enable the appropriate OpenType feature tag in your software: for small caps, the tag is small caps; for oldstyle numerals, it is onum. In InDesign, these are checkboxes in the OpenType panel. In CSS, they are font-feature-settings or the more specific font-variant-numeric property. The practical detail: if you are setting a price, a date, or a number inside a sentence, oldstyle numerals are the right choice. They match the height of the lowercase letters. Lining numerals belong in tables and headlines. The failure case is the designer who uses small caps by selecting the text and hitting the small caps button in the character panel, which does not use the typeface's actual small caps but instead scales down the capital letters. That is a different failure mode, and it is the typographic equivalent of a fake bold.

Two Commands That Do Different Jobs

Kerning and tracking are often used as if they were the same thing. They are not. Kerning is the proportional adjustment between a specific pair of characters. Tracking is uniform spacing applied across a range of characters. The difference matters in production because they are controlled by different commands in every professional application. If you adjust tracking on a headline, every pair moves by the same amount. If you adjust kerning, only the selected pair moves. The amateur failure is to use tracking as a substitute for kerning, or to use kerning as a substitute for tracking. The rule is: kerning for pairs that the metrics table missed, tracking for headlines that need a uniform loosening or tightening. The minimum tracking adjustment for text below 10 pt is plus 0.5% to plus 1%. At small sizes the default spacing is too tight. For display type, the opposite is true: you may need to tighten the tracking to bring the letters together. The production test is to select a range of text and look at the tracking value in the character panel. If it is not zero, the designer has made a decision. If it is zero, they have made a decision too, the wrong one, more often than not.

The Full Answer, Directly

What are the specific, fixable typography mistakes that most reliably make a layout look unprofessional, and what is the correct production method for each? The answer is the list above. Too many fonts, faux bold and italic, bad spacing, widows and orphans, wrong punctuation, unjustified justification, broken baseline grids, x-height mismatches, variable font misuse, and missing OpenType features. Every one of them is a production error, not a taste error. Every one of them has a named fix. Every one of them will be caught by a reader who does not know the term for it but knows the page is wrong.

An honest caveat: this guide has given you the rules, but the rules have an expiry date. Kerning tables are updated by foundries, CSS properties change with browser versions, and the specific code point for a character never changes but the software you use to type it does. The maximum number of consecutive hyphenated lines in justified text is two. If you see three, your measure is too narrow by at least 10 characters. That is a number no other guide on this subject will give you, and it is the exact kind of production detail that separates a designer who is guessing from a designer who knows.