You hit submit on your assignment, lean back, and wonder: what happens to my paper now? For millions of students, the answer is a Turnitin similarity report — a document your instructor opens that looks nothing like the single percentage number you imagine. It's a detailed, color-coded breakdown of your entire paper.
Most students never see this side of Turnitin. They hear their similarity score secondhand — "you got 14%" — without ever learning what that 14% actually contains. The truth is that the percentage is the least interesting part. What matters is where the matches come from, how long they are, and what patterns they form. Understanding this view changes how you write, paraphrase, and cite — and it removes a lot of unnecessary fear. Let's walk through exactly what your professor sees.
When your instructor opens your submission, the first thing they notice is a single number: the overall similarity percentage. This is the figure everyone talks about, and it represents the proportion of your paper's text that matches sources in Turnitin's databases — student papers, published journals, books, websites, and more.
But here's what students usually miss: this number is not a verdict. A 25% score on a literature review with heavy quoting and a bibliography is routine; a 12% score made of one unquoted paragraph lifted from a single website is a problem. Instructors are trained (or learn quickly) that the percentage alone tells them almost nothing definitive. It's a starting point — an invitation to look closer. Anything below roughly 20–30% often gets a quick glance, while higher numbers trigger a deeper review. The real investigation happens in the panels beneath that number.
Below the headline figure sits the Match Overview panel — a ranked list of every source your paper matched, ordered by overlap. Each entry shows the source name (a website, a journal article, a previously submitted student paper), the percentage of your paper it accounts for, and a color swatch.
This is where instructors form their first real judgment. One source contributing 15% of your paper looks very different from fifteen sources each contributing 1%. The first pattern suggests heavy reliance on — or copying from — a single source. The second usually means your paper is well-researched and engaging broadly with the literature, which is exactly what academic writing should do. A diverse, low-percentage spread across many credible sources reads as healthy to an instructor. A concentrated spike from one obscure source reads as suspicious.
Clicking any source in the Match Overview highlights every matching passage directly in your document in that source's assigned color. Your professor doesn't just see a list — they see your paper with rainbow patches showing exactly which sentences came from where.
This visual layer is remarkably revealing. Long continuous blocks of one color — three or four highlighted sentences in a row — scream copy-paste even before a word is read. Scattered single-word or short-phrase highlights in many colors, on the other hand, are usually just common terminology, standard phrases, or correctly used technical language, and instructors dismiss them in seconds. The pattern matters more than the pigment: blocky highlights suggest copied passages; sprinkled highlights suggest normal academic writing.
Beyond the overview, the Match Details view lets instructors click through to the actual source text side by side with your writing. They can see the original webpage or paper passage next to your version, highlighted word for word.
This is where paraphrasing quality gets judged. If your version sits next to the source and only a few words have been swapped — the classic "synonym shuffle" — it's instantly obvious, and it still counts as plagiarism if uncited. Genuinely reworked sentences with proper citation, by contrast, usually don't even appear here at all, or appear as tiny fragments. Instructors also check whether matches come from legitimate sources: matches to published journals and books are expected in research writing; matches to essay mills, homework sites, or another student's paper are red flags.
Here's a feature students rarely know exists: instructors can exclude categories of text from the score. The three standard filters are:
Not every instructor uses these filters, and settings vary by institution — which is exactly why two students with identical writing can get different reported scores. It also explains why a paper with extensive quoting might show 30% to you but be read as perfectly fine by a professor who filters quotes and bibliography. Before you panic about a number, remember that your instructor may be looking at a very different version of it.
Alongside similarity, most instructors can open a separate AI-writing report that estimates what percentage of your paper was likely generated by an AI tool. This appears as its own indicator — not mixed into the similarity score — and flags specific sections as likely AI-generated.
Two things matter here. First, the AI score and the similarity score are independent: a fully original essay can show low similarity and high AI probability at the same time. Second, like similarity, the AI percentage is a starting point, not proof — instructors are advised to use it as one signal among others, such as a mismatch with your usual writing voice or knowledge shown in class. Still, an elevated AI indicator will absolutely trigger a conversation you don't want to have. The safest position is simple: do your own writing, and disclose any permitted AI assistance exactly as your instructor requires.
After reviewing thousands of reports, instructors develop instincts. Here's the honest summary of what raises eyebrows and what earns a nod:
Notice that none of the "satisfied" items require a 0% score. Instructors expect some matching — it's evidence you researched. A clean-looking report isn't one with no color at all; it's one where the color tells a story of honest research.
The best time to see your report is before your professor does. A pre-submission check lets you review your own Match Overview with the same eyes your instructor will use. When you look at yours, run through this checklist:
Crucially, use a non-repository check for this — one that doesn't store your paper in Turnitin's student-paper database. If your draft gets saved to the repository, your final submission can match your own draft and flag as self-plagiarism. A quick pre-submit check with TurnPro AI gives you the official Turnitin similarity and AI report without adding your paper to the database, so you can revise with confidence and submit clean.
There is no universal "safe" number — it depends on your assignment type, your institution's policy, and which filters your instructor applies. As a rough guide, under 20% with matches spread across many sources and proper citations is rarely questioned. What matters far more than the number is the pattern: thin, scattered matches to credible sources are fine; concentrated blocks from one source are not.
They can see the AI-writing indicator if your institution has it enabled, which estimates the share of likely AI-generated text and highlights suspect passages. It works independently of the similarity score. It's a signal, not a verdict — but combined with other signs, it can lead to an academic integrity review. Disclose permitted AI use as your instructor requires.
In the raw score, yes — but most instructors exclude the bibliography and quoted material using Turnitin's filters before judging your paper. Properly formatted quotes and a complete reference list are legitimate, so don't strip them out to game the number. That said, a paper that is mostly quotations is still weak academic work even if the filtered score is low.
Instructors can apply different filter settings (excluding quotes, bibliography, or small matches), and assignment settings vary by course. Two instructors looking at similar papers can legitimately report different numbers. Focus on the quality of your matches — cited, diverse, and in your own words — rather than chasing someone else's percentage.
The Turnitin report your professor sees is not a single damning number — it's a visual map of how your writing relates to the world's published work. Instructors read it with nuance: they expect matches, filter out the legitimate ones, and judge patterns rather than percentages. Once you understand their view, the path to a clean report is straightforward — research broadly, paraphrase genuinely, cite everything, and quote properly.
And you don't have to guess how your paper looks. Reviewing your own similarity report before submission turns anxiety into a simple editing task. A pre-submit check through TurnPro AI — an official Turnitin similarity + AI report with non-repository checking — lets you spot block matches, single-source spikes, and citation gaps while there's still time to fix them. Submit knowing exactly what your professor will see.