Fifteen days might sound rushed, but you can still improve your TOEFL Speaking score significantly if you know what you are targeting.

The 2026 TOEFL Speaking section is a completely different animal from the old format, and most students are still prepping for the version that ETS retired in January. That gap right there is where scores are being left on the table.

This plan is built for you if your test is coming up fast and you want to boost your results in a focused, efficient way, without vague advice like ‘practice every day’ or ‘think in English.’

It is specific, it is grounded in how the 2026 Speaking section actually works, and it is structured to move your score in the right direction over two weeks.

Let’s get into it.

---

First, Know What You’re Being Tested On

The old TOEFL Speaking had four tasks — one independent task where you gave your opinion, plus three integrated tasks where you read, listened, and then gave a structured response. That whole setup is gone.

The 2026 TOEFL Speaking section now runs for about 8 minutes total and consists of exactly two task types:

TASK	WHAT GETS SCORED	RESPONSE TIME
Listen and Repeat (x7)	Pronunciation, rhythm, intonation, accuracy of repetition	8-12 seconds per sentence
Take an Interview (x4)	Fluency, coherence, idea development, grammar, vocabulary	45 seconds per question

No preparation time is given for either task. You hear the prompt, and you respond immediately. The section measures your ability to communicate clearly and effectively in spoken English in academic and campus life settings, with a focus on spontaneous, real-time communication.

That spontaneity is the key word here. On the 2026 test, every Speaking response is AI-scored — ETS’s own test blueprint places all 11 Speaking items in the AI-scored category. That is exactly why practising with SpeechRater® matters: it is the same engine ETS uses, so your feedback is measured against the criteria that actually decide your score.

For a full breakdown of how the 2026 format compares to what came before it, the 2026 TOEFL format guide on the Santa blog covers every change in detail.

---

The One Technique to Ace ‘Listen and Repeat’

The Listen and Repeat task catches a lot of students off guard because it tests something most people have never trained: auditory memory under speaking pressure. You hear a sentence once, it disappears, and you have 8 to 12 seconds to reproduce it as accurately as possible, including the rhythm, intonation, and phrasing.

The technique that directly trains this skill is shadowing: listening to a spoken sentence and repeating it back simultaneously or immediately after, without looking at any written text.

A systematic review published in Tandfonline (2025) covering 44 studies found that shadowing training consistently improves comprehensibility, intelligibility, fluency, and prosodic control in second-language speakers.

Separately, research published in NIH’s PubMed Central found that shadowing training also increases working memory capacity, which is exactly what Listen and Repeat demands.

Here is how to run a shadowing session to improve your TOEFL Speaking score:

1. Pick an audio source with a clear, native speaker at normal conversational speed. Podcasts, TED talks, or official TOEFL practice audio all work well.
2. Listen to one sentence all the way through without stopping.
3. Immediately repeat it aloud from memory, focusing on matching the rhythm and intonation, not just the words.
4. Record your response and play it back against the original. This is where most of the learning happens.
5. Move on to longer, more complex sentences once shorter ones sound natural. The Listen and Repeat task escalates in length across its 7 questions, so your practice should too.

Days 3 to 8 of your 15-day plan should be centered almost entirely on shadowing. Ten sentences a day, recorded and reviewed, is enough to build real auditory fluency in that window.

---

How to Stay Fluent in ‘Take an Interview’

The Take an Interview task format comes across as more casual than the old TOEFL Speaking tasks. Four questions, 45 seconds each, on a single topic like commuting habits, campus life, or daily routines.

There’s no reading or listening passage to reference. Just you, talking.

What ETS is scoring here is your fluency, coherence, idea development, grammar, and vocabulary, per the official 2026 scoring criteria. A high-scoring response at this task is not the longest response. It is a response where the main idea lands clearly in the first few seconds, gets supported with a specific example or detail, and finishes without trailing off or repeating itself.

🎙️The 45-Second Rule: Say Something Real, Then Expand It

Position (5 seconds): State your answer directly. ‘I prefer commuting by train’ beats ‘Well, it really depends, but I think maybe…’

Reason (15 seconds): Give one clear reason. One is enough. Two is fine if they are brief. Three is too many for 45 seconds.

Specify (20 seconds): Add a concrete example, comparison, or personal experience.

Conclude (5 seconds): End with a one-sentence conclusion that finishes the response cleanly.

On Days 9 to 11, practice this structure by picking a new topic each day, running through all four questions in a single 45-second-per-question block, and recording the whole set. The goal is not to sound like you rehearsed it, but to have a conversation.

Tip: Speak to the topic, not to the timer. Students who are clock-watching while they talk are the ones who lose coherence. Trust the structure, fill the 45 seconds naturally, and your pacing will sort itself out.

---

The 15-Day Plan to Level Up Your Speaking

This plan is designed for roughly 45 to 60 minutes of daily focused practice. It leads with diagnostic work so that everything after Day 2 is targeted, not generic. And it reserves the final two days for pattern correction based on the specific feedback you have already collected.

🗓️Here Is Every Day, Mapped Out

DAYS	FOCUS	DAILY TASK
Days 1-2	Diagnostic	Take a full Speaking simulation. Review your responses and identify where points drop.
Days 3-5	Listen and Repeat foundations	Shadow 10 sentences daily. Record yourself. Compare your rhythm and intonation to the original.
Days 6-8	Listen and Repeat fluency	Practice with longer, more complex sentences. Focus on auditory memory, not reading from text.
Days 9-11	Take an Interview fluency	Answer 4 interview questions per session on a new topic. Prioritize idea development and pacing.
Days 12-13	Full, timed simulation	Complete a full Speaking section under timed test conditions. Review with detailed AI feedback.
Days 14-15	Weak-area sprint	Use LUMI feedback to drill the 2-3 patterns that cost you the most points. Final mock session.

One rule to live by: Record every single Speaking response. Without recordings, you are flying blind. The gap between how you think you sound and how you sound is where most of the score improvement is hiding.

---

Recording Yourself Is the Cheat Code High Scorers Use

Here is something most students skip entirely: self-review through recording. Nobody likes listening to themselves. But the research on self-monitoring in language acquisition consistently points to recording as one of the highest-ROI habits for spoken fluency improvement.

When you play back your own responses, you catch things you do not notice while speaking:

• Filler words that eat into your 45 seconds.
• Sentences that made sense in your head but fell apart midway through.
• Intonation patterns that flatten everything out and make your response sound robotic.

None of this shows up to you in real time, but it is completely obvious on playback.

Thus, pair recording with structured feedback, and the improvement compounds fast. Which is exactly what AI-powered feedback tools like LUMI on Santa TOEFL are built to provide: specific, pattern-based analysis of where your pronunciation, fluency, and coherence are dropping, not just a score with no context.

---

Three Patterns That Might Drain Your Speaking Score

After 15 days of deliberate practice with recorded review, patterns become very visible. Here are the three most common ones that you can rectify to improve your TOEFL Speaking score:

1. Trailing intonation on repetitions

In the Listen and Repeat task, a lot of students start the sentence with good energy and then let the second half of the sentence flatten out. ETS’s scoring system specifically flags intelligibility, which means each part of the sentence needs to hold up. Practice sentences to the very last word, not just the beginning.

2. Starting the response with a hedge

Phrases like ‘That’s a really interesting question’ or ‘I think it depends on many factors’ eat into your 45 seconds and delay the core content of your response. The scorers are evaluating communication, and communication starts with your answer, not your preamble.

3. Running out of content before 30 seconds

Forty-five seconds is longer than it feels under pressure. Students who finish in 20 to 25 seconds are leaving points on the table. If your idea runs out early, that is a signal to go deeper on your example, add a comparison, or tie your answer back to the original question.

Santa TOEFL’s AI tutor LUMI flags all three of these patterns specifically, with feedback on which questions triggered them and what the fix looks like. Students who want to improve their TOEFL Speaking score results without that kind of granular feedback are essentially guessing at what to fix.

---

The Takeaway: 15 Days Are Enough If You Use Them Right

A successful 15-day Speaking sprint is not about grinding through as many questions as possible. It is about understanding the two tasks you are being scored on, practicing the specific skills each one tests, reviewing your responses with the right feedback, and adjusting based on what the data tells you.

Generic apps that give you a thumbs up or a rough pronunciation rating are not going to move your band. So you need the same scoring system the real exam uses.

Santa TOEFL is an official ETS partner. Its Speaking practice is scored using ETS’s own SpeechRater® technology, the same engine ETS uses to evaluate responses on the actual TOEFL iBT. That is why Santa’s predicted score aligns with real test results 99% of the time — a data-driven forecast built on the actual scoring criteria, not a rough guess from a generic app.

On top of that, Santa’s AI tutor LUMI does something most tools do not: it tells you exactly why your score came in where it did. Whether your rhythm dropped mid-sentence on a repetition task, your idea development stalled in the Interview, or your vocabulary variety pulled your grammar score down, LUMI surfaces the pattern and tells you what to do about it. That is the feedback loop that turns 15 days of practice into score improvement.

✨A Few More Santa Perks

• 21+ official ETS mock tests, all updated for the 2026 Speaking format, including the new task types
• 99%+ score prediction accuracy, so you know exactly where you stand before test day
• 4M+ downloads globally, with 600 million+ questions answered by learners on the platform
• Works on mobile, tablet, and desktop, so your 15-day plan fits whatever schedule you are working around

So, rest your speaking plan on shadowing for Listen and Repeat. Structured, spontaneous responses for Take an Interview. Recorded self-review every session. And AI feedback that matches ETS’s real scoring criteria.

Stick to it across 15 days, and you’ll be able to improve your TOEFL Speaking score surely.