How To Write A Good Poem

The Actual Work of Putting Words on Paper

Most people overthink this. They read a few tutorials, try to sound important, and produce something that reads like it was written by a committee of English professors. The core of how to write a good poem has very little to do with sounding profound. It has to do with attention. Specifically, the ability to notice what is actually happening in a moment and transfer that onto the page without decorating it with borrowed emotion. I spent several years doing this for a small literary journal. We received roughly four hundred submissions a month. Ninety percent of them were what I call decorative writing — beautiful sentences that said almost nothing once you stripped away the ornamental language. The remaining ten percent usually shared one trait: the writer trusted the reader to handle ambiguity instead of explaining every feeling explicitly.

How To Write A Good Poem Without Losing Your Mind

Start with the observation, not the theme. Pick a concrete scene — a laundromat at 11 PM, a cracked parking lot after rain, a phone conversation where both people are avoiding the real topic. Write down exactly what you see, hear, and notice. Do not tell the reader what to feel about it. Let the details carry the weight. This is probably the single most effective shift you can make, and it is also the hardest thing to do honestly because you have to resist the impulse to interpret your own material. Here is a practical problem I ran into repeatedly: a poem about grief that kept landing as melodramatic no matter how hard I tried to pull back. The breakthrough came when I stopped trying to make it about grief entirely and instead focused on the specific, unromantic facts — the way someone's coffee mug stays in the sink for days, the silence of a house that used to have background noise. The poem got better the moment I allowed the ordinary to stand on its own. Big emotions handled indirectly almost always land harder than big emotions announced directly. Line breaks matter more than most beginners realize. A line break is a pause, a breath, and sometimes a double meaning. Consider the difference between writing "she opened the door and saw" on one line versus splitting it as "she opened the door / and saw." The second version creates a micro-second of anticipation. That delay does work. It forces the reader to slow down and participate in the moment rather than passively consuming it.

Rhythm is another area where people get it wrong. You do not need strict meter. You need to read your lines out loud. If you stumble, the line is probably carrying too much conceptual weight for its syllabic length. Trim it. If a line sounds like a sentence you would actually say, that is usually a good sign, even if it feels too plain on the page. Poetry that sounds natural when spoken tends to survive revision better than poetry that looks clever in your head. There is a specific trap that catches experienced writers too. It is the compulsion to make every image count. When you believe each image has to carry thematic weight, you start stacking symbolism until the poem suffocates. A single unexplained image often does more work than three heavily annotated ones. Leave something alone. Let it just exist in the poem without demanding that it represent anything larger. Drafting approach matters significantly. I recommend writing a messy first draft with zero self-editing, even if it reads like prose or personal diary entries. Then set it aside for at least a day. When you return, revise aggressively. Cut half the adjectives. Replace vague words like "sad" or "beautiful" with the specific thing that caused those responses. If you cannot name the specific thing, the poem is probably telling you something you have not fully understood yet.

Revision is where the actual writing happens. Most poems are not written in the first draft. They are discovered during the third or fourth pass. The first draft is just you figuring out what the poem is actually about. The second draft is you arranging the pieces. The third draft is you deciding what not to include, which is frequently the hardest decision because you are attached to your favorite lines. One useful technique I found myself returning to constantly: read the poem backward, line by line, from the last line to the first. This disrupts your forward momentum and lets you evaluate each line on its own merit rather than being carried along by the narrative flow. Lines that fall apart in isolation are the ones that usually need the most work. A common misconception is that good poems require complex vocabulary. They do not. Plain language combined with precise observation is almost always more powerful than elevated language applied to vague observations. The difficulty in poetry is not in the words you choose. It is in the accuracy of what you choose to write about.

Some forms help beginners learn constraints. Free verse is often worse for early writers because it provides no structural guardrails. Try a strict form like a villanelle or sestina first, not because you must write in form forever, but because the constraints force you to make decisions you would otherwise avoid. The constraints create pressure, and pressure reveals what the poem actually needs. If you are stuck, read the work of poets who are working in your same general wavelength. Not the famous ones everyone recommends — the recent ones published in the last five years in journals like The Paris Review, Poetry, or The New Yorker. Contemporary poets are solving the problems you are currently facing. Their solutions will be more relevant than anything published thirty years ago.

What This Approach Does Not Do

This method will not help you write poems that trend on social media. Algorithm-friendly poetry tends toward accessibility and broad emotional appeal, which often means sacrificing specificity. If your goal is popularity over craft, you are pursuing a different objective entirely. There is nothing wrong with that, but it is not what this guide addresses. The approach also breaks down with certain subjects. Highly abstract or philosophical topics resist the concrete observation model more than others. If your poem is fundamentally about an idea rather than an experience, the techniques here will feel restrictive. In those cases, you might need a different framework — perhaps one that leans more on argument structure or conceptual play than on imagery and scene. Most poems benefit from external feedback, but not everyone gives useful feedback. A supportive friend will tell you the poem is beautiful. A fellow writer will give you notes that may or may not be correct. The most useful feedback comes from readers who can tell you exactly where they got confused or where their attention drifted, without telling you how to fix it. Their confusion is data. How you use that data is your decision.

The process takes longer than most people expect. A single poem that ends up feeling inevitable usually goes through six to twelve drafts over several weeks. Some poems never reach that point and that is fine. Not every poem has to be finished. Sometimes the exercise of writing it produces the insight you needed, even if the poem itself remains unresolved.