Why PTE templates do not work, and why selling them is a scam

Search "PTE template" and you will find hundreds of pages promising a script that gets you 79+. Coaching centres sell them. WhatsApp groups pass them around. Students memorise them the night before.

They do not work. They have never worked the way they are sold, and since August 2025 they actively cost you marks.

This is not an opinion about study habits. It is a consequence of how the test is scored, and you can reason it out yourself in two minutes.

What PTE actually scores in a speaking answer

Every open speaking task — Describe Image, Retell Lecture, Respond to a Situation — is scored on three things:

| Component | What it measures | |---|---| | Content | How much of the thing in front of you your answer actually covered | | Oral fluency | Rhythm, pace, whether you stop and restart | | Pronunciation | Whether the sounds you made are the sounds of the words |

Note what Content is measured against: the source material. The specific image. That specific lecture. Not "a good answer in general" — this answer to this prompt.

The mechanism: a template cannot cover the source

Here is the whole argument, and it is almost embarrassingly simple.

A template is, by definition, the same words regardless of what appears on your screen. That is what makes it a template. You memorise it once and say it every time.

But Content is scored by comparing what you said against what was in front of you. Semantic overlap, key content words, named entities, the numbers on the chart.

So:

If your words are identical whether the chart shows unemployment in Brazil or rainfall in Norway, those words cannot be covering either of them.

A template does not score badly on Content. It scores nothing on Content — it is structurally incapable of doing otherwise. Every second you spend on "This graph provides valuable information about various categories and it is clear that there are noticeable differences" is a second producing zero content marks.

And you only get forty.

It costs you twice

1. It eats your time. Describe Image gives you 40 seconds. A twenty-second template preamble leaves twenty seconds for the only part that scores Content. Students who say less but name two actual numbers routinely beat students who deliver a polished, empty paragraph.

2. Since August 2025, it is detected. PTE moved to a hybrid model — AI plus human oversight — and template detection is part of it. Reported penalties run up to 30% of the score. The AI is described as flagging "rhythmic monotony": the flat, recited delivery of someone reading from memory.

Pearson does not publish its algorithm, so treat the exact 30% as reported rather than official. But note where that reporting comes from: the coaching sites selling template packs are the ones publishing the warning. They sell you the script on one page and tell you it gets detected on another.

That contradiction is the tell. Nobody who believed templates worked would publish the warning.

So why does everyone sell them?

Because a template is the easiest thing in test prep to sell.

It is a file. It can be delivered on WhatsApp, costs nothing to reproduce, and feels like a shortcut — which is exactly what a frightened student twelve hours before an exam wants to buy. Teaching someone to describe an unfamiliar chart in forty seconds takes weeks and cannot be forwarded to a friend.

The scam is not that templates exist. It is selling a memorised script as a score guarantee, to people who cannot verify the claim until they have already paid for the exam.

What actually helps — structure, not script

There is a version of this that works, and the difference is precise:

• A script is sentences. Same words every time. Scores nothing, risks detection.
• A structure is an order. Same five slots, filled with what is actually on your screen, every time different.

A workable structure for Describe Image is simply: what type of thing it is → what it measures → the highest value with its label → the lowest or a clear contrast → a one-line close.

Five slots. The words change completely with every image, because four of the five slots can only be filled by looking at it. You never freeze wondering what comes next, and every sentence you say is about the thing in front of you.

That is the version we teach in our Describe Image guide — and we are explicit there that the structure gets you through the 25 seconds of preparation, and nothing more. It does not earn marks. Naming two real numbers earns marks.

The part nobody can shortcut

Here is why templates stay popular despite all of this: most students have no idea what their pronunciation and fluency actually score. They cannot hear themselves the way a machine hears them. So they reach for the one thing they can control — the words — and hope.

That is a measurement problem, not a strategy problem.

Our PTE mock scores your recording the way the exam does and then shows you every word of your own answer coloured green, orange or red for pronunciation, plus per-question feedback on what you did and what to change. Not a band. Not "work on fluency". The actual words you mispronounced, in the answer you just gave.

Once you can see that, the appeal of memorising someone else's paragraph disappears — because you can finally see where your marks are really going.

FAQ

Do PTE templates still work in 2026?

Not as a way to earn marks. Content is scored against the specific image or lecture in front of you, and a template is the same words regardless — so it cannot cover the source. Since August 2025 PTE's hybrid AI-plus-human scoring also flags memorised responses, with reported penalties of up to 30%.

Is using any structure risky?

No. A fixed order of what to mention is not a template — the words change with every prompt because they describe the prompt. What triggers detection is word-for-word language that thousands of candidates deliver identically.

Why do coaching centres still sell templates?

Because they are trivially easy to sell: a file, sent on WhatsApp, that feels like a shortcut. Teaching someone to describe an unfamiliar chart in forty seconds is slow, personal work that cannot be forwarded.

What should I practise instead?

Reading two numbers off any chart inside 25 seconds, and speaking for 35 seconds without stopping. Those two skills transfer to every image. A memorised paragraph transfers to none.

How do I know my pronunciation is good enough?

You cannot hear it yourself — that is the actual problem templates are papering over. Get a per-word score on a recording of your own voice. Take a free PTE mock and look at which words come back orange or red.

The short version

A template is the same words whatever is on the screen, and Content is scored on covering what is on the screen. It cannot work. Since August 2025 it is also detected and penalised.

Learn the order, not the sentences — then find out what your pronunciation and fluency actually score, because that is the number the template was hiding from you.

Take a free PTE mock test — full length, per-word pronunciation on your own recording, and feedback on every question. No card.