EnglishBlitz

TOEIC Link Part 3 Strategy — The Three-Speaker Conversation: How to Track Who Wants What When a Third Voice Joins

A handful of Part 3 sets swap the usual two speakers for three, and the added voice is where careless test-takers lose points. This strategy shows you how to tag each speaker by role the moment they speak, follow the disagreement instead of the words, and answer the "what will the second man do next" question without re-listening.

EnglishBlitz Editorial Team·

TOEIC Link Part 3 Strategy — The Three-Speaker Conversation: How to Track Who Wants What When a Third Voice Joins

Most Part 3 conversations are a tidy back-and-forth between two people. A few sets, though, add a third speaker, and that single change is where a surprising number of points disappear. The words are no harder. What breaks is attribution: the questions ask what a specific speaker decided, offered, or will do next, and when three voices are trading lines quickly, it is easy to hang the right action on the wrong person. This strategy is about keeping the three roles straight so the answer is a matter of recall, not reconstruction.

Why the third speaker is a trap, not just more words

A two-person conversation has a built-in safety net: if you miss who said something, there are only two candidates, and the voices usually alternate. Add a third speaker and both supports vanish. Two of the three may be the same gender, so voice alone will not separate them. The turns stop alternating predictably — one person can speak twice in a row while another stays quiet. And crucially, the answer choices are built to exploit exactly this. A question like "What will the woman do next?" will offer an action that a different speaker proposed, banking on the fact that you remember the action but not who owns it.

The fix is to stop listening for content first and start listening for role. If you know that the woman is the client, the first man is the vendor, and the second man is the vendor's technical colleague, then every action naturally files itself under the right person, and the attribution traps collapse.

Tag each speaker by role in the first ten seconds

Three-speaker sets almost always establish who everyone is at the very start, because the conversation needs it too. Listen to the opening lines for the relationship, not the details:

  • "Thanks for coming in — I'm hoping you can walk our team through the proposal." → this speaker is the host / client.
  • "Of course. I've brought my colleague from engineering to answer the technical questions." → speaker two is the lead / salesperson, speaker three is the specialist.

That is your whole map: client, lead, specialist. From that point on, do not try to remember names or voices. When the specialist says the install will take two weeks, you file two-week install under "specialist," not under "the second man" or "the one with the deeper voice." Roles are sticky; voices and names are slippery.

This role-first habit is the same discipline that makes the Part 2 tag-question confirmation responses manageable — in both cases you are listening for the function of the line (who is confirming, who is asking) rather than trying to hold every word.

Follow the disagreement, because that is where the question lives

When three people talk, the interesting questions cluster around friction: someone proposes, someone objects, and someone resolves. TOEIC loves to ask about that arc.

  • The proposal: "Why don't we schedule the training for next Friday?"
  • The objection: "Friday's tight — the new hires don't start until Monday."
  • The resolution: "Then let's do it the following Wednesday, once everyone's onboarded."

If a question asks "What do the speakers decide to do?", the answer is the resolution — the following Wednesday — not the first proposal, even though Friday was said first and louder. Train yourself to hear the shape "someone pushed back, then they landed somewhere else," and mark the landing point. The wrong answers are almost always the discarded proposal.

Pre-read the questions for the name tags

You get a few seconds before each set. In a three-speaker set, spend them scanning the three questions for who each one is about. If you see "What does the woman suggest?", "What is the second man concerned about?", and "What will they do next?", you now know before the audio starts that you must separately track the woman's suggestion and the second man's concern. That advance warning is what lets you file the right lines under the right person in real time instead of scrambling afterward.

A quick note on the ubiquitous "What will [speaker] do next?" question: the answer is nearly always in the last few lines, and it is an action that specific speaker commits to about themselves — "I'll send over the revised quote this afternoon." Do not accept an action that someone else assigns to them unless that speaker agrees to it. "Could you email the quote?" — "Sure, I'll do that now" only counts because the speaker took it on.

A 40-second practice loop

Take any three-speaker Part 3 transcript and run this drill without the audio:

  1. Read only the first two lines and write the three roles: client / lead / specialist (or whatever fits).
  2. Read the body and, next to each proposed action, write the role that owns it.
  3. Find the one point of disagreement and circle the resolution.
  4. Answer the three questions using only your role tags and the circled resolution.

If you get an attribution question wrong, it will almost always be because you skipped step 1 and tried to track voices instead of roles. Do the tagging first, every time, and the third speaker stops being a threat and becomes just one more clearly labeled column in your head. For the vocabulary that fills these workplace conversations, keep the conference registration and badge pickup cluster drilled alongside this strategy, since the settings overlap constantly.