# The Hormozi content research report

500 collected Reels. 431 in the performance comparison. Snapshot: 21 September 2026.

Created by Artem Novitckii with Creator Lab. Resource: https://novitckii.com/resources/creator-lab/

These are patterns worth testing, not proven causes of reach or guaranteed winning formulas.

## Start with a useful sequence of ideas.

**613K median plays · 31 Reels · 1.71× the sample median**

The steps label had the highest median plays among script structures with at least 15 examples. Its median like response was 26.1 per 1,000 plays, versus 17.6 overall. In the reviewed examples, this includes both ordered instructions and compact lists of lessons. A repeatable format to test is a clear promise followed by several distinct, usable points.

**What held up:** The signal stayed above the period baseline in both April-June (853K; 12 Reels) and July-September (453K; 19 Reels). Among structure labels with confidence of at least 0.70, steps still had a 613K median across 21 Reels. That is a consistency check, not a significance test.

**What limits it:** One high-performing personal reveal was labeled steps. Removing it leaves 30 Reels and a 584K median. Within marketing alone, steps had 330K median plays across eight Reels, below 345K for the other marketing structures. The overall result does not transfer equally to every topic.

**Try it:** Turn one topic you know well into 3-5 concrete decisions or actions. Give each point a different job. Cut any point that only restates the promise.

## A specific dilemma gives a story somewhere to go.

**494K median plays · 33 story-structured Reels · 1.38× baseline**

Story structure ranked second among structures with at least 15 examples. The co-founder example below moves from an awkward, recognizable problem to a personal parallel, then a decision and its cost. That progression gives the viewer a reason to stay beyond the opening sentence.

**What held up:** Story-structured business posts had a 536K median across 12 examples, versus 330K for the other 200 business posts. The subgroup is small. Story hooks are a separate label: those 35 Reels had 427K median plays, versus 357K for curiosity hooks.

**What limits it:** A story is not automatically a winner. The lower-performing interview example below also contains a personal lesson. These are different subjects, speakers and publication dates, not a controlled comparison.

**Try it:** Use a real situation: what someone wanted, what made it difficult, what they chose, and what changed. Bring your own experience instead of borrowing Hormozi’s life story.

## Human stakes reached beyond the core business topic.

**714K median plays · 20 relationship Reels · 24.9 likes / 1K**

Relationships had the highest topic median among categories with at least 15 examples. Habits followed at 495K across 18 Reels. Mindset offered a larger sample: 92 Reels, 408K median plays and 24.0 likes per 1,000. Business made up 212 Reels, with 338K median plays and 14.6 likes per 1,000.

**What held up:** After removing later posts flagged for similar scripts, relationships still had a 781K median across 18 examples. This category includes both advice and personal or emotional material, so it is not a single repeatable format.

**What limits it:** Broader subjects may attract a broader audience. These figures do not show leads, customers or revenue. Money had an even higher raw median, but only seven examples, so it is not a reliable topic recommendation from this sample.

**Try it:** Connect your expertise to a real human tradeoff: time with family, embarrassment, uncertainty, commitment or a difficult decision. Keep the connection relevant to the people you want to help.

## A direct instruction is worth testing, not crowning.

**511K median plays · 19 instruction openings · 26.9 likes / 1K**

Among opening forms with at least 15 examples, instructions led the raw median. Story openings had 424K across 42 Reels; claims had 352K across 287; questions had 331K across 77. The useful creative move is to give a particular viewer something concrete to do immediately.

**What held up:** The instruction median was 459K across the 15 examples with opening confidence of at least 0.70. Several high-performing instruction posts address broad money or mindset concerns, so subject matter may account for part of the difference.

**What limits it:** Opening form and hook mechanism are different. A question can create curiosity or recognition; an instruction can promise a result. Most Reels here started with a claim, which describes frequency, not the best choice for your next post.

**Try it:** Try: ‘Before you [specific action], check [specific thing].’ Deliver the check immediately, then show why it changes the decision.

## There is no universal winning hook in this archive.

**Story hooks: 427K · result hooks: 418K · curiosity: 357K**

Story and result hooks led median plays across the seven hook categories. Direct hooks had the highest median like response at 24.8 per 1,000, but lower median plays than story or result hooks. The best-looking option depends on whether you care about reach or response.

**What held up:** Hook rankings changed under a confidence filter. At confidence of at least 0.70, result hooks fell to a 355K median across 28 examples, while curiosity rose to 423K across 36. Story retained 533K but only seven examples. Business + story had 424K across 18 Reels; mindset + curiosity had 400K across 24.

**What limits it:** These are overlapping topic, length, format and hook differences. Jev’s label confidence is not a probability of success. The data does not support ‘use this hook and go viral.’

**Try it:** Pick two hook mechanisms for the same topic family. Publish several genuinely different ideas in each, then compare at the same post age.

## Longer Reels were competitive here.

**90+ seconds: 427K median plays · 89 Reels · 22.7 likes / 1K**

The 90+ second group led median plays and median like response among the four duration buckets. Under-30-second Reels had 278K median plays and 20.4 likes per 1,000. The 30-59 second group had 382K; the 60-89 second group had 338K.

**What held up:** The older and newer groups are not evenly distributed: median post age was 53 days for the 90+ second bucket and 132 days for the under-30 bucket. A length comparison cannot isolate editing, topic or audience effects.

**What limits it:** We do not have watch time, completion rate or retention curves. Nothing here proves that making a Reel longer improves it.

**Try it:** Let the explanation earn its length. Use extra time for a concrete example, an objection or a demonstration, rather than repeating the same advice.

## Personal experience is a useful ingredient, not proof by itself.

**414K median plays · 85 personal-evidence Reels**

Personal-evidence posts had 21.1 median likes per 1,000 plays, versus 17.2 for reasoning-led posts. The reviewed co-founder story makes its advice concrete by showing an actual conversation and decision. That is a more useful writing lesson than adding an unsupported number to an opening.

**What held up:** Personal evidence retained a 399K median across 52 high-confidence labels. The numbers category had 430K across only 13 Reels, and only three high-confidence labels. The apparent winner ‘no evidence’ had just five examples.

**What limits it:** An anecdote can illustrate a lesson without proving a universal rule. Emotional categories were also noisy: amusement had 526K across 24 labels, but only two met the 0.70 confidence threshold.

**Try it:** Show an experience you can honestly describe, then separate what happened from what you think it means. Let the viewer see the decision or before-and-after.

## A spoken CTA was rare. That says little about conversion.

**419 of 431 Reels had no detected spoken CTA**

About 97% of the comparison sample had no spoken call to action. Comment, buy, follow and visit categories were too small for a sensible performance ranking. This archive mostly shows how content delivers an idea, not how different CTAs convert.

**What held up:** CTA labels came from speech. A CTA in a caption, graphic or profile would not necessarily appear in these transcripts. Shares, saves, profile visits and sales are absent.

**What limits it:** Do not interpret ‘no spoken CTA’ as ‘no offer’ or ‘never use a CTA.’ Public plays and likes cannot tell you which posts produced business results.

**Try it:** Finish the promised lesson first. Add one relevant next step when there is a real resource, action or offer that helps the viewer use it.

## All eight dimensions

Medians, not pooled rates. Groups under 15 are exploratory. Confidence-filtered columns use labels >= 0.70.

### topic

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| money | 7 | 850771 | 24.83 | 0.368 | 59.2 | 80.0 | 6 | 905215 |
| relationships | 20 | 713590 | 24.88 | 0.353 | 50.9 | 84.5 | 14 | 780742 |
| health | 2 | 702604 | 17.32 | 0.206 | 44.6 | 50.1 | 2 | 702604 |
| habits | 18 | 495092 | 21.92 | 0.309 | 49.2 | 71.5 | 14 | 495092 |
| other | 5 | 477349 | 18.46 | 0.384 | 56.0 | 124.7 | 1 | 2894088 |
| mindset | 92 | 407942 | 23.99 | 0.335 | 49.8 | 79.3 | 74 | 402576 |
| marketing | 69 | 345285 | 17.07 | 0.273 | 60.6 | 76.7 | 62 | 345398 |
| business | 212 | 337826 | 14.56 | 0.231 | 72.2 | 92.0 | 181 | 346110 |
| technology | 6 | 279661 | 15.82 | 0.242 | 32.0 | 147.7 | 4 | 276515 |

### mechanism

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| story | 35 | 427443 | 17.72 | 0.262 | 76.1 | 94.8 | 7 | 533429 |
| result | 61 | 418288 | 18.61 | 0.293 | 74.1 | 107.7 | 28 | 354912 |
| direct | 51 | 375478 | 24.83 | 0.293 | 47.0 | 76.7 | 9 | 231989 |
| curiosity | 127 | 356516 | 16.68 | 0.247 | 58.5 | 92.3 | 36 | 423129 |
| recognition | 105 | 326549 | 17.13 | 0.255 | 71.1 | 71.4 | 45 | 344257 |
| contradiction | 40 | 322162 | 18.85 | 0.315 | 50.0 | 81.6 | 16 | 322162 |
| mistake | 12 | 314709 | 14.60 | 0.233 | 67.4 | 88.4 | 6 | 314709 |

### opening

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| dialogue | 6 | 743596 | 16.21 | 0.363 | 85.4 | 73.1 | 2 | 422514 |
| instruction | 19 | 510802 | 26.93 | 0.291 | 51.2 | 83.0 | 15 | 458610 |
| story | 42 | 423998 | 21.65 | 0.260 | 77.3 | 90.6 | 25 | 413607 |
| claim | 287 | 351874 | 16.68 | 0.259 | 65.7 | 90.9 | 244 | 343147 |
| question | 77 | 330788 | 17.91 | 0.285 | 53.9 | 65.4 | 68 | 333156 |

### structure

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| steps | 31 | 612524 | 26.09 | 0.276 | 79.8 | 76.7 | 21 | 612524 |
| story | 33 | 494122 | 18.99 | 0.263 | 62.2 | 83.7 | 27 | 508456 |
| qa | 76 | 397799 | 15.93 | 0.247 | 65.0 | 93.0 | 35 | 447619 |
| comparison | 21 | 373677 | 17.62 | 0.266 | 59.8 | 63.7 | 12 | 332533 |
| unclear | 4 | 367826 | 10.52 | 0.339 | 36.4 | 66.4 | 0 | Unavailable |
| opinion | 53 | 356450 | 22.81 | 0.354 | 30.7 | 66.7 | 26 | 355518 |
| explanation | 41 | 349234 | 22.54 | 0.292 | 45.8 | 86.5 | 13 | 430176 |
| problem_solution | 172 | 314988 | 15.77 | 0.246 | 73.5 | 86.1 | 88 | 302149 |

### evidence

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| none | 5 | 1203753 | 39.44 | 0.445 | 31.0 | 32.7 | 1 | 1203753 |
| numbers | 13 | 430176 | 12.15 | 0.311 | 68.1 | 86.5 | 3 | 219941 |
| personal | 85 | 414089 | 21.05 | 0.292 | 60.5 | 86.4 | 52 | 398988 |
| reasoning | 262 | 344666 | 17.21 | 0.262 | 65.6 | 92.0 | 182 | 336179 |
| example | 62 | 342458 | 16.99 | 0.244 | 60.7 | 74.8 | 32 | 294530 |
| source | 4 | 324659 | 15.12 | 0.246 | 45.5 | 31.9 | 1 | 352044 |

### emotion

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| amusement | 24 | 525999 | 17.01 | 0.290 | 58.8 | 56.5 | 2 | 527262 |
| surprise | 11 | 415256 | 12.33 | 0.284 | 62.4 | 69.6 | 5 | 415256 |
| concern | 131 | 362067 | 17.27 | 0.272 | 73.7 | 83.7 | 69 | 346110 |
| aspiration | 213 | 352044 | 17.91 | 0.259 | 63.5 | 86.6 | 129 | 356516 |
| relief | 30 | 321562 | 21.73 | 0.279 | 54.5 | 106.2 | 12 | 277207 |
| neutral | 22 | 242568 | 16.14 | 0.239 | 44.3 | 75.2 | 4 | 307835 |

### specificity

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| unclear | 1 | 758821 | 16.99 | 0.395 | 57.7 | 13.9 | 0 | Unavailable |
| none | 13 | 751471 | 19.82 | 0.356 | 53.0 | 86.5 | 5 | 525411 |
| sequence | 73 | 374500 | 17.13 | 0.251 | 82.6 | 84.7 | 44 | 390786 |
| principle | 156 | 356085 | 17.14 | 0.277 | 49.6 | 83.8 | 96 | 382038 |
| action | 188 | 347138 | 18.08 | 0.266 | 67.9 | 81.6 | 83 | 346520 |

### cta

| Label | n | Median plays | Likes / 1K | Comments / 1K | Median seconds | Median age days | Confident n | Confident median plays |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| comment | 3 | 510315 | 24.35 | 5.155 | 84.8 | 84.4 | 2 | 400958 |
| follow | 1 | 502344 | 26.96 | 0.330 | 91.7 | 7.6 | 0 | Unavailable |
| buy | 7 | 458610 | 18.87 | 0.234 | 92.5 | 114.6 | 5 | 661917 |
| none | 419 | 356516 | 17.61 | 0.266 | 63.7 | 83.7 | 414 | 356483 |
| visit | 1 | 313715 | 24.71 | 0.274 | 41.9 | 9.7 | 0 | Unavailable |

## Topic + hook combinations (n >= 15)

| Group | n | Median plays | Likes / 1K | Comments / 1K |
|---|---:|---:|---:|---:|
| business + story | 18 | 423998 | 18.07 | 0.229 |
| mindset + curiosity | 24 | 399566 | 21.16 | 0.329 |
| mindset + recognition | 24 | 375844 | 23.14 | 0.331 |
| mindset + contradiction | 16 | 369396 | 22.76 | 0.346 |
| business + result | 41 | 362067 | 14.88 | 0.236 |
| marketing + curiosity | 21 | 355935 | 17.98 | 0.226 |
| business + direct | 19 | 352044 | 15.85 | 0.205 |
| marketing + recognition | 21 | 337144 | 17.13 | 0.259 |
| business + curiosity | 63 | 331408 | 13.56 | 0.220 |
| business + contradiction | 15 | 285042 | 13.80 | 0.269 |
| business + recognition | 50 | 263523 | 14.37 | 0.226 |

## Duration

| Group | n | Median plays | Likes / 1K | Comments / 1K |
|---|---:|---:|---:|---:|
| Under 30 seconds | 52 | 277882 | 20.37 | 0.340 |
| 30–59 seconds | 146 | 382038 | 17.00 | 0.270 |
| 60–89 seconds | 144 | 337982 | 15.83 | 0.233 |
| 90+ seconds | 89 | 427443 | 22.74 | 0.270 |

## Annotated examples

Transcript-led readings. Snapshot performance does not establish cause.

### Compress experience into distinct lessons

https://www.instagram.com/p/DZL3PYSuGT1/

1,868,492 plays · 38.2 likes / 1K · 47 seconds · 2026-06-05

> and I'm going to compress 13 years of business advice into this one video.

- Promise: a large body of experience in one short piece.
- Body: a sequence of separate decisions about offers, visibility, pricing and focus.
- Payoff: the viewer leaves with several things to reconsider, without waiting through a long setup.

The transferable move is the density of usable distinctions. Make the scope honest: three lessons from one project can work as a premise without pretending to have 13 years of experience.

A second business list, DaeUZqsu86B, reached 1.13M plays. That makes the format worth studying, but it does not establish that lists caused either result.

### A second example of the compact lesson list

https://www.instagram.com/p/DaeUZqsu86B/

1,127,474 plays · 35.1 likes / 1K · 53 seconds · 2026-07-07

> Here's 14 years of business advice in 60 seconds.

- Promise: experience plus a time boundary.
- Body: short, distinct judgments about priorities, hiring, conversations and growth.
- Payoff: an accumulation of takeaways rather than a single reveal at the end.

Each sentence introduces another decision. Use this as a structure for your own earned lessons; do not copy the claims or borrow the authority.

Its saved duration is about 53 seconds. The spoken ‘60 seconds’ is the creator’s framing, not a timing measurement for our analysis.

### The awkward co-founder conversation

https://www.instagram.com/p/DY7xLtSPfhQ/

1,269,425 plays · 24.0 likes / 1K · 74 seconds · 2026-05-29

> I made the mistake of telling a friend about a business idea earlier.

- Problem: a friend assumes a casual idea means a partnership.
- Bridge: Hormozi describes a similar moment early in his business.
- Turn: he chooses to work alone despite a threat of competition.
- Payoff: avoiding one hard conversation can carry a long-term cost.

The story answers a specific dilemma. Start with a situation the viewer can picture, then make the decision and tradeoff clear.

This is a recognition hook with a story body. It demonstrates why hook mechanism and script structure need separate labels.

### Business advice with a human cost

https://www.instagram.com/p/DbrDjD4y-F9/

1,041,281 plays · 21.7 likes / 1K · 114 seconds · 2026-08-05

> I'm 22, my girlfriend is annoyed that I'm working so much.

- Conflict: the viewer feels forced to choose between work and a relationship.
- Body: personal experience, then the idea of explicit agreements.
- Payoff: turn a vague conflict into concrete expectations.

The opening makes a business ambition matter at home. The content lesson is to show the human stake in your topic. We are not validating the relationship advice or research claims in the source.

This sits in relationships, not business, despite discussing work. Topic labels are interpretations, not objective boundaries.

### A useful sequence that still reached fewer people

https://www.instagram.com/p/DbrgGQ6Bfmb/

126,122 plays · 7.8 likes / 1K · 88 seconds · 2026-08-06

> Hey Alex, if you can pay for specific knowledge, skills, and wanna learn one-on-one, how would you approach someone

- Question: how to learn from people who might see you as competition.
- Body: approach vendors and pay to observe their process.
- Payoff: a concrete way to frame the request and explain the value.

This is practical, detailed advice, yet its saved plays and like response sit below the sample medians. Useful content and broad reach are different outcomes.

Same business topic and steps label as the compact lists above, but a longer question-led explanation. Different audience appeal, dates and delivery prevent a causal conclusion. Its structure confidence is only 0.38.

### A personal story is not enough on its own

https://www.instagram.com/p/Dau_sXbOTsd/

141,383 plays · 11.6 likes / 1K · 78 seconds · 2026-07-13

> What are the unintended consequences and maybe some of the sacrifices that you still feel like you're making

- Setup: a broad interview question about the sacrifices of working hard.
- Body: reflection on preferred work and advice from a mentor.
- Payoff: notice what separates a good day from a bad day and adjust.

There is a recognizable lesson, but the opening takes longer to establish a specific situation. That is an editorial observation from the transcript, not an explanation proven by the metrics.

It had 141K plays, compared with 1.27M for the co-founder story. Both are story-structured, but their topics and dates differ. Keep the weaker example in view when studying the format.

## Original templates to adapt

### Compact lessons

Three things I learned after [real experience] that changed how I [specific task].

Give one distinct action or decision per point. Use only experience you actually have.

### Specific dilemma

You [ordinary action]. Now [unexpected complication]. Here is how I would handle it.

Make the situation concrete, explain the tradeoff, then give a usable next step.

### Story with a decision

I wanted [goal], but [obstacle]. The decision that changed it was [choice].

Show what the choice cost and what happened. Do not invent a clean success story.

### Instruction first

Before you [task], check [specific detail].

Demonstrate the check and explain the consequence of getting it wrong.

### Human stakes

The hardest part of [your topic] is explaining [real tradeoff] to [person affected].

Connect your expertise to an authentic situation, then help resolve one part of it.

### Concrete process

If [specific situation], try this three-step process for [bounded task].

State the scope, show the steps and give the viewer a way to check the result.

### Result with explanation

We changed [one thing]. Here is what happened and what else might explain it.

Use a documented result. Separate the observation from the cause you suspect.

### Honest comparison

These two approaches solve [problem] differently. Here is when I would use each.

Compare the same decision using concrete examples, including the downside of each option.

## Run your own experiment

1. Choose one narrow topic family and one goal: reach, response, or a business action you can actually measure. Do not combine them into an invented success score.
2. Pick two formats from the findings, such as a compact lesson list and a specific dilemma story. Write three original ideas for each. Keep the subject and intended viewer comparable.
3. Use a real example, a clear decision and a delivered payoff in every script. Choose a length that serves the idea. The opening should match what the body delivers.
4. Publish at your normal sustainable pace. Record date, topic, format, opening and length before looking at performance. Avoid changing the label to fit the result.
5. Read metrics at a consistent age, such as seven days after each post. Compare median plays and median likes per 1,000. Track qualified enquiries or resource signups separately if those matter to you.
6. Review a stronger and a weaker example from each format. Six posts are a pilot, not a verdict. Keep the useful observations, collect a larger batch, and change one meaningful creative choice at a time.

## Methodology and limitations

1. Collection snapshot: 21 September 2026, 11:12 UTC. The 500 collected posts span 1 April to 21 September 2026. These are saved figures, not current Instagram counts and not the entire account.
2. Coverage: 447 completed classifications, 45 no-audio posts, seven with too little speech, and one failed media item. Four of the completed posts were already marked as duplicate scripts. Removing those leaves 443. Removing 12 posts less than seven days old at collection leaves the 431-post comparison sample. All 431 have positive plays and known likes and comments.
3. Plays are the available metric. The views field is absent in this sample, so we do not rename plays as unique reach or views. One person can generate more than one play. The baseline is the median of the 431 eligible posts: 358,306 plays, 17.64 likes per 1,000 plays and 0.267 comments per 1,000 plays.
4. For each label, we report the median post, not the average and not the pooled engagement rate. Likes per 1,000 = likes / plays × 1,000, calculated per post before taking the median. Multiples compare a category median with the overall sample median, which includes that category; they are descriptive, not incremental lift.
5. All eight dimensions use the stored Jev labels: topic, hook mechanism, opening form, structure, evidence, emotion, specificity and spoken CTA. Confidence of at least 0.70 is a sensitivity filter, not proof the label is correct. Missing confidence fails that filter. We reviewed the complete stored transcripts of the six featured examples and additional candidate and weaker examples. We did not independently verify every line against video or reclassify all 431 scripts.
6. Groups with fewer than 15 posts are shown but marked as small samples. Fifteen is an editorial screening rule, not a statistical guarantee. No significance tests, causal estimates or predictive validation were performed. We checked selected patterns by topic, publication period, label confidence and repeated-script sensitivity.
7. A second duplicate check found five later posts with substantial script overlap beyond the existing duplicate flags. It compared unique five-word sequences, required at least 35 such sequences in each script and flagged overlap of at least 75% of the smaller set. Removing the later candidates gives 426 posts; steps retain a 627,228 median across 30. These are sensitivity results, not a new primary sample or five manually confirmed duplicates.
8. The personal reveal DZu7FiJhmNU was labeled steps, illustrating category noise. Removing only that post gives 584K median plays across the other 30 steps posts. Lists of advice and sequential instructions also share the steps label. Read the examples before treating it as a precise creative format.
9. This is an observational sample from one large established account. Posts differ in age, topic, duration, speakers, format and likely distribution. Seven days is a minimum-age filter, not equal observation time. We do not have retention, watch time, saves, shares, follower growth attributable to each post, paid distribution, leads or sales. The results describe associations, not why a Reel succeeded or what will work on another account.
10. Example quotations are short excerpts from machine transcripts and may contain transcription errors. Open the original before quoting or adapting. The templates are newly written suggestions, not Hormozi quotations. Public CSV exports contain source links, snapshot metrics, labels, confidence and short openings; they omit credentials, signed media URLs, raw provider responses and full transcripts.