Choose the analysis rubric that matches the talk you gave

A sermon, academic lecture, tutorial, TED-style talk, political address, and general keynote create different obligations for the speaker and different expectations for the audience.

AskSpeech does not force them into one interchangeable checklist.

The report pipeline first identifies the main sustained speech inside the video, then applies the category whose labels, evidence rules, and grading expectations best fit the talk.

That means a political speech can receive Actionability, a TED-style talk can receive Originality and Practicality, an educational presentation can receive Delivery, and a religious message can use a sermon-specific rubric with categories such as Structure, Biblical Accuracy, and Conviction.

Many videos include introductions, announcements, sponsorships, interviews, audience questions, or post-talk discussion.

AskSpeech first identifies the main sustained message.

When a clear boundary can be established, material outside the talk is excluded from the structure map and score so the report reflects the speech itself rather than the surrounding video.

Quick Verdict

A concise editorial judgment of the message's most consequential strengths and limitations.

Intended Audience

A description of who the talk appears to address and why the topic matters to those listeners.

Key Quotes

Up to six transcript-grounded quotations showing the thesis, strongest material, central weakness, turning point, practical implication, or requested response.

Strongest Elements

Up to three meaningful strengths explaining what the speech is doing well, why the moment works, and which category it most directly affects.

Most Important Weaknesses

Up to three material weaknesses explaining what is going wrong, where it appears, why it matters, and how it affects the listener.

Direct evidence rules

The report uses transcript-grounded wording.

It does not invent quotations, improve language, or treat the same underlying issue as several separate findings.

Current Spine

Every category includes a Current Spine that maps the message as it was delivered.

The structure varies by category, but the spine always shows the central claim, the major sections or movements, whether each section is strong, weak, or drifting, the likely listener effect, and a supporting quotation.

Re-Spined

Re-Spined appears only when a material rebuild would improve the talk.

It preserves the strongest existing material, proposes a clearer order, and is ignored when grading the original speech.

AskSpeech grades the focused transcript in the category set that matches the talk.

Categories by talk type

• Religious: Biblical Accuracy, or the tradition-specific text and doctrine label; Depth; Conviction; Structure; Clarity; Practicality; Engagement.
• Academic: Evidence / Accuracy / Truthfulness; Depth; Structure; Clarity; Practicality; Engagement; Credibility.
• Educational: Evidence / Accuracy / Truthfulness; Clarity; Structure; Practicality; Engagement; Depth; Delivery.
• TED-style: Originality; Clarity; Structure; Practicality; Engagement; Depth; Credibility.
• Political: Originality; Clarity; Structure; Actionability; Engagement; Depth; Credibility.
• General: Evidence / Accuracy / Truthfulness; Depth; Structure; Clarity; Practicality; Engagement; Credibility.

The 1-10 score scale

Scores are displayed out of 10, with decimals when the judgment falls between two levels.

Read the scale this way:

• 1: Not functioning in this category; the transcript gives almost no usable evidence of the quality being measured.
• 2: Severe breakdown; there are fragments of intent, but the category mostly fails.
• 3: Weak; some relevant material is present, but major gaps or confusion dominate.
• 4: Underdeveloped; the category works in places but remains unreliable, thin, or poorly connected.
• 5: Mixed; the speech shows basic competence and clear limitations in about equal measure.
• 6: Adequate; the category generally works, but with noticeable unevenness, missing support, or missed opportunities.
• 7: Solid; the speech is effective in this category with specific areas to strengthen.
• 8: Strong; the category is consistently effective, with refinements needed more than repairs.
• 9: Excellent; the speech handles this category with unusual clarity, development, and control.
• 10: Exceptional; the category is handled with rare precision and effectiveness across the focused transcript.

A 0.0 can appear only when the transcript gives effectively no usable evidence for a category. The scale describes the public score; the report's reasoning explains why that speech landed there.

Overall score

The overall score is a holistic judgment of how well the speech accomplishes its category-specific task across the full focused transcript.

It is informed by the seven category scores and by the severity of the transcript evidence behind them, but it is not calculated by a hidden weighted formula.

Grading sequence

AskSpeech grades the focused transcript before writing the qualitative report. The category scores, reasoning, and overall score are selected from the original speech as delivered, then locked while the report explains them.

Re-Spined is produced only after grading. Its suggested rebuild can help revision, but it does not raise or lower the original scores.

AskSpeech does not grade on a curve or spread scores to match a preferred pattern. Two speeches with the same number can earn it for different reasons, which is why every score is paired with transcript-based reasoning.

AskSpeech evaluates the spoken message through the transcript.

That means it can analyze claims, reasoning, structure, transitions, stories, examples, audience orientation, repetition, and conclusion quality.

It does not directly measure eye contact, gesture, facial expression, vocal tone, physical stage presence, slide design, or visual delivery.