10 min read

12 Daily Speaking Drills That Took Me From B1 to C1

SpeakShark founder Duy shares the exact 12 daily speaking drills (timing + reps) that broke his B1 plateau and reached C1 in 18 months. No fluff.

Quick answer: Going from B1 to C1 takes most adult learners 18-30 months of daily focused practice. The exact stack that worked for me as a non-native founder: one SpeakShark open-conversation session daily (drill #1 — generative output + real-time phoneme scoring), plus 11 supporting drills covering shadowing, voice memos, picture-prompts, reaction-speed Q&A, and CEFR-progressive monologues. Total daily commitment: 45-60 minutes, six days a week. No tutor required. Total cost: under $200 across 18 months.

I'm Duy. I built SpeakShark because I was the user who needed it. I sat at B1 for nearly three years before I figured out the stack below.

This isn't theory. These are the twelve drills I ran daily, with the timing and reps that actually moved my level. If you're stuck at B1 wondering why "more Netflix" and "more vocab apps" aren't working, you're about to find out why, and what to do instead.

Why B1 is the worst plateau (and why most people never escape it)

B1 is the "I can survive" level. You can order food, ask for directions, have small talk. Your brain registers this as success and stops being uncomfortable. Without discomfort, no growth.

The escape requires three things working in the same session:

  1. Forced generative output — you have to produce novel speech, not just imitate
  2. Immediate phoneme-level feedback — otherwise you ossify wrong sounds for life
  3. Progressive overload — turn length, abstraction, and topic difficulty must climb weekly

Most B1 learners do exactly one of these (usually #1, badly) and wonder why nothing changes. The twelve drills below cover all three. Drill #1 covers all three in a single session, which is why it's the foundation.

The 12 daily speaking drills (ranked by leverage)

I've ranked these by what gave me the biggest measurable jump per minute spent. The top drill is non-negotiable. The bottom three are optional polish for the C1 final mile.

Drill 1 (⭐ Editor's Pick foundation): SpeakShark open conversation, 20 minutes

Reps: 1 session daily, 6 days a week Difficulty curve: Start with familiar topics → abstract opinion → ethics/economics by month 6 Why it leads: This is the only drill that hits all three escape conditions simultaneously. You're producing novel speech, getting real-time phoneme-level pronunciation scoring, and you can progressively overload by changing teachers and topics.

I rotated all four native-accent teachers weekly:

  • Sarah (American) — Mondays/Thursdays, for the accent I hear most in tech
  • James (British) — Tuesdays/Fridays, for register and hedging vocabulary
  • Emily (Australian) — Wednesdays, for prosody variety
  • Liam (Canadian) — Saturdays, for casual register and idioms

The free tier (3 sessions/day, no card, no trial timer) was enough for my first four months. When I crossed into B2 territory and started needing longer sessions and harder topics daily, I upgraded to Pro at $100/year. Cheapest level-up in the history of language learning. See pricing for the current breakdown.

🦈 Try SpeakShark Free → — 3 sessions a day with native-accent AI teachers, real-time phoneme scoring, no credit card. This is drill #1. Start tonight.

Drill 2: Shadowing, 10-15 minutes

Reps: Same 30-90 second clip × 5-8 passes Material: TED talks at B1/B2, podcast intros at B2/C1, news anchors at C1 Why it works: Shadowing is the highest-leverage pronunciation and prosody drill ever invented. You lock onto a native speaker's rhythm, pitch, and chunking and force your mouth to match in real time.

Method: pick a 60-second clip. Listen once. Play it again and speak over the speaker, matching their speed and intonation exactly. Don't translate. Don't pause. Five passes per clip is the sweet spot. By pass three, your mouth knows shapes it didn't know existed.

What shadowing doesn't do: train generative output, give you feedback on your own errors, or build argument structure. That's why it's drill #2, not #1. Pair it with drill #1, never replace.

Drill 3: Voice memo monologue, 5 minutes

Reps: 1 recording daily, listen back same day Prompt: "Talk about your day for 3 minutes without stopping" Difficulty curve: Day → opinion → abstract argument → counter-argument by month 9

Open your phone's voice recorder. Hit record. Talk for three minutes without stopping. Then listen back the same evening.

The first time you do this you'll be horrified. That's the point. You hear every filler word, every grammar collapse, every mispronounced th. This is the cheapest feedback loop on earth. Cost: zero. Time: 5 minutes including playback.

Combine with drill #1 for compounding effect: after a SpeakShark session, record a voice memo summarizing what you just discussed. Forced retrieval + immediate self-review.

Drill 4: Picture-prompt monologue, 5 minutes

Reps: 1 prompt × 2 minute response, daily Source: Cambridge speaking exam practice books, Pinterest random image search

Open a random photo. Describe it for two minutes. Then speculate — who are these people, what happened before this photo, what happens next.

This drill is criminally underused. It trains the exact skill the Cambridge C1 Advanced speaking test grades you on (extended turn, speculation, comparison). It also trains the everyday skill of producing English under unpredictable cognitive load.

Once you can do this in your sleep at B2, raise the difficulty: pick two photos and compare them for 90 seconds. That's the actual C1 task.

Drill 5: Reaction-speed Q&A, 10 minutes

Reps: 15-20 rapid questions, single-sentence answers Tools: SpeakShark conversation set to "rapid mode" or a friend with a question list

Speed kills the B1 plateau. The reason you stall mid-sentence isn't grammar — it's retrieval speed. Reaction-speed drills force you to produce some answer in under 2 seconds, even if imperfect.

I'd hand my partner a printed list of 20 random questions ("describe your kitchen", "what do you think about remote work", "name three things you'd take to a desert island") and just rapid-fire through them. No editing. No do-overs. Move on.

Later in the journey I replaced the human partner with SpeakShark's open conversation in fast topic-switching mode. Same drill, available 24/7.

Drill 6: Chunk acquisition, 10 minutes

Reps: Learn 5 new chunks daily, drill into spoken use within 48 hours Source: Anki deck "Lexical Chunks B2-C1" or harvest from drill #2 shadowing

C1 speakers don't think in words. They think in chunks: to a certain extent, that said, given the circumstances, I'd argue that, not to put too fine a point on it.

Five new chunks a day. That's 1,825 a year. By the end of month 6 you'll find yourself reaching for them automatically mid-conversation. By month 12, you sound C1 even when your grammar slips, because chunks carry register.

The trick: a chunk doesn't count until you've spoken it in a real session within 48 hours of learning it. Otherwise it stays passive forever.

Drill 7: Read-aloud with prosody focus, 10 minutes

Reps: 1 paragraph × 3 reads (cold → analyzed → expressive) Source: New Yorker articles, podcast transcripts, novel excerpts

Pick a paragraph of native-written English. Read it cold once. Then analyze: where would a native pause? Which words carry stress? Where does pitch rise and fall? Mark it up. Then read it a third time, expressive, like you're recording an audiobook.

This drill is the bridge between shadowing (someone else's voice) and your own monologue (your own voice). It teaches you to self-direct prosody.

Drill 8: Self-explanation in English, 5 minutes

Reps: Anything you read or learn → re-explain in English out loud

You read a tweet thread, watch a YouTube video, finish a chapter. Before doing anything else, you stand up and explain the content out loud, in English, to nobody. Two minutes. Free recall.

This drill compounds with everything else you consume during the day. Cost: zero extra time, because you're using existing content. It's the closest thing to "passive practice that actually works."

Drill 9: Substitution drills, 5 minutes

Reps: 1 base sentence × 10 variations Example: "If I had known, I would have called" → swap tense, swap subject, swap verb

This is the only "old-school grammar drill" that still earns its place in 2026. It builds automaticity on structures you understand intellectually but stall on under pressure (especially conditionals, reported speech, and modal perfects).

Five minutes daily on whichever structure you keep flubbing in drill #1. The next SpeakShark session, you'll deploy it without thinking. That's the goal.

Drill 10: Debate-yourself, 10 minutes

Reps: 1 topic × 90 sec FOR + 90 sec AGAINST + 90 sec synthesis Prompts: Daily news headline, ethics dilemma, business decision

You take a topic. You argue FOR it for 90 seconds. Then AGAINST it for 90 seconds. Then synthesize a balanced view in 90 seconds. Record the whole thing.

This drill is the single biggest accelerator at the B2→C1 transition. C1 isn't about knowing more vocabulary. It's about handling complexity — qualification, concession, hedging, synthesis. Debate-yourself trains exactly that.

I did this every Sunday for an hour (six topics × 10 minutes), and within 90 days my SpeakShark session scores on opinion topics jumped from low 70s to mid 80s.

Drill 11: Pronunciation minimal pairs, 5 minutes

Reps: 10 pairs daily, focus on YOUR weak sounds Tool: SpeakShark phoneme scores (it tells you which phonemes you fail most) + YouGlish for examples

Every L1 has signature L2 weaknesses. Vietnamese speakers fail final consonants and th. Spanish speakers fail vowel length distinctions. Mandarin speakers fail r/l and final clusters. You know yours by now.

Five minutes a day on minimal pairs (ship/sheep, think/sink, full/fool) won't fix everything, but compounded over 18 months it eliminates the accent markers that keep recruiters and exam graders from rating you C1.

This drill is only worth doing if you have phoneme-level feedback. Otherwise you're just reinforcing the wrong sound. SpeakShark's scoring tells you per-phoneme where you actually fail — see how it works for the breakdown.

Drill 12: Dream-language journaling, 5 minutes (optional)

Reps: First thing in the morning, out loud Why: Forces L2 retrieval before L1 wakes up

Optional but high-leverage. The first 60 seconds after waking, you describe out loud (in English) what you remember from your dreams, or what you plan to do today. Five minutes.

This trains unbidden English retrieval — the kind of speech that comes out before your brain has fully booted L1 censorship. It's also the closest you'll get to thinking-in-English without living abroad.

The full daily schedule (45-60 min, 6 days/week)

Here's how I stacked the drills. You don't need to do all 12 daily. The top 5 are the engine. The rest cycle in based on what's weak that month.

Time slot Drill(s) Minutes Notes
Morning (right after coffee) #1 SpeakShark conversation 20 Non-negotiable. The keystone.
Mid-morning #12 Dream journal (optional) 5 While brushing teeth
Lunch break #2 Shadowing + #6 Chunks 15 Headphones, anywhere
Afternoon #8 Self-explanation 5 Whenever you consume content
Evening #3 Voice memo OR #4 Picture-prompt OR #5 Reaction Q&A (rotate) 10 Phone-only, no setup
Weekly Sunday #10 Debate-yourself (6 topics) 60 Replaces evening drill that day

That's it. 50 minutes weekdays. One longer Sunday block. Nothing on Saturdays (rest is part of the protocol — your brain consolidates).

The CEFR progression schedule

Don't run the same difficulty for 18 months. Climb deliberately.

Months 1-4 (B1 → B1+):

  • SpeakShark topics: daily life, hobbies, opinions on familiar topics
  • Shadowing: B1 TED-Ed, slow podcasts
  • Goal: 3-minute uninterrupted turns, phoneme scores ≥75 on familiar topics

Months 5-10 (B1+ → B2):

  • SpeakShark topics: opinion on unfamiliar topics, hypotheticals, comparisons
  • Shadowing: regular-speed TED talks, podcast intros
  • Add: Drill #10 debate-yourself weekly
  • Goal: 4-minute turns, phoneme scores ≥80, comfortable hedging vocabulary

Months 11-16 (B2 → C1):

  • SpeakShark topics: abstract argument, ethics, economics, identity
  • Shadowing: fast podcasts, news anchors
  • Add: Drill #7 read-aloud with The New Yorker, drill #9 advanced conditionals
  • Goal: 5-minute turns under cognitive load, phoneme scores ≥85, self-correction without flow-break

Months 17-18 (C1 consolidation):

  • SpeakShark topics: free, very long sessions, multiple teachers per week
  • Drop drills you've over-trained (#9 substitution, #11 minimal pairs if your phoneme scores are stable ≥85)
  • Optional: one mock exam with a CELTA tutor for calibration

🦈 Try SpeakShark Free → — Run the keystone drill #1 daily, free, no card. Three sessions a day with phoneme scoring is plenty to break the B1 plateau in the first 90 days.

What I cut from my stack (and why)

I want to be honest about what didn't work, because the internet is full of advice that wastes B1 learners' time.

  • Duolingo at this level: Useless. It tops out near A2. Don't waste a minute past month one.
  • Watching Netflix with subtitles: Pure input. Built zero of my speaking ability. Fine for vocabulary recognition, irrelevant for production.
  • Reading grammar books: I read one (Murphy's). That's enough. More grammar reading at B1+ is procrastination.
  • Generic "speak with random people online" apps: Too inconsistent, no feedback on errors, often awkward. SpeakShark's structured open conversation with real-time scoring replaced these entirely. See SpeakShark vs ELSA Speak for why open conversation beats sentence-repeat drilling at this level.
  • IELTS/Cambridge prep books before C1: Useless until you're already B2+. Doing them at B1 just teaches you to fail at C1 tasks.

How SpeakShark fits the full stack (and why I built it)

Quick honesty: I built SpeakShark because the keystone drill — daily generative conversation with real-time phoneme feedback — didn't exist in a form a non-native learner could actually afford and stick with.

The closest alternatives were either too narrow (ELSA Speak does pronunciation drilling, not open conversation), too generic (chatbots with no pronunciation scoring), or too expensive (iTalki at $25 a session burns out the budget by month two).

SpeakShark's design choices map directly to the drills above:

  • Open conversation, not sentence-repeat → covers drill #1, the only drill that builds generative C1 output
  • Real-time phoneme-level scoring → covers drill #11 inside drill #1, no separate session needed
  • Four native-accent teachers → built-in accent rotation, no extra app required
  • 3 free sessions/day, no card → enough to run the keystone drill daily without paying anything for months
  • $12/mo or $100/yr Pro → cheapest path to unlimited daily practice when you outgrow the free tier

If you're comparing options before committing, the honest breakdown is in best ELSA Speak alternatives. For most B1-to-C1 learners, SpeakShark is the keystone and 11 other free/cheap drills fill in around it.

What to do tonight

Don't read another B1-to-C1 article. Run the drill instead.

  1. Sign up for SpeakShark free (takes 30 seconds)
  2. Pick Sarah or James, pick a topic you have an opinion on
  3. Talk for 15 minutes
  4. Tomorrow morning, do it again
  5. Day 30, look back at your scores

That's how I started. That's how every B1-to-C1 jump I've seen actually started. The other 11 drills layer in once the keystone is locked.

The plateau breaks when you stop reading about breaking it and start outputting daily under feedback. Twelve drills, 50 minutes, 18 months. See you at C1.

Frequently Asked Questions

How long does it actually take to go from B1 to C1 in speaking?

For adult learners doing focused daily practice, 18 to 30 months is realistic. The CEFR's own research suggests roughly 350-400 guided learning hours between B1 and B2, and another 500-600 between B2 and C1. That's about 900-1000 hours of productive practice, not passive Netflix time. If you do 45-60 minutes of structured speaking drills six days a week, you'll hit 1000 hours in roughly 30 months. I did it in 18 because I front-loaded conversation reps using SpeakShark's free tier (3 sessions/day) plus shadowing and voice memos.

Why do so many learners get stuck at B1 forever?

The B1 plateau is real and it has a specific cause: at B1 you can survive most conversations with simple structures, so your brain stops being uncomfortable. Without discomfort, no growth. Most B1 learners also stop outputting and switch to passive input thinking it'll level them up. It won't. The escape requires daily forced output, immediate phoneme-level feedback, and progressive overload on turn length and topic complexity. SpeakShark's open conversation covers all three in one session.

Is shadowing alone enough to reach C1?

No. Shadowing is the best drill for pronunciation and prosody but it's an input-imitation drill, not a generative one. C1 demands you produce novel speech under pressure. Use shadowing as drill #5 for 10-15 minutes daily, but pair it with generative drills like SpeakShark conversations, picture-prompt monologues, and reaction-speed Q&A.

Do I need a tutor or can I self-study from B1 to C1?

You can absolutely self-study to C1 in 2026. The tooling has changed. SpeakShark now solves daily conversation practice and pronunciation correction for $0 (free tier) or $12/month. A human tutor at $25/hour costs $1,500 for 60 sessions; SpeakShark Pro gives unlimited sessions for $100/year. I went B1 to C1 with zero tutors and total spend under $200.

What's the single most effective drill if I only have 15 minutes a day?

Open conversation with immediate feedback. One SpeakShark session with Sarah or James on a topic that requires opinion and argument. The combination of forced generative output, real-time phoneme scoring, and a teacher pushing follow-up questions hits more skills per minute than any other drill.

How do I know when I've actually reached C1 vs just feeling fluent?

Real C1 has measurable markers: 4+ minute turns on abstract topics without filler stalls, real-time self-correction without breaking flow, natural hedging and discourse markers, and pronunciation that stays consistent under cognitive load. The cleanest test: take SpeakShark's longest open conversation on a hard topic (ethics, economics) and check your phoneme scores stay above 85 across the whole session.

Can I do these drills if I'm working full-time?

Yes. The total daily commitment is 45-60 minutes split into three blocks: 20 minutes morning, 15-20 minutes lunch, 10-15 minutes evening. SpeakShark works on mobile, shadowing works on a commute, voice memos work walking the dog. The non-negotiable is daily consistency — six days a week of 45 minutes beats one weekend of 6 hours by a factor of about 4x.