Point an AI at your own Instagram.

Copy thisCopy

You have access to my Instagram account. Run a competitor teardown for me.

STEP 1 — Profile me
Look at my account: bio, last ~20 posts, formats I use (Reels/carousels/statics),
topics, hooks, posting cadence, and which of my posts over- and under-performed
relative to my own median. State my niche and target audience in one line.

STEP 2 — Find competitors
Identify 5 accounts in that same niche and audience — not the biggest accounts in
the category, but ones 1–10x my size doing the same job I'm doing. Skip accounts
that are famous for something other than this content. List each with follower
count and a one-line description of their angle.

STEP 3 — Analyze their last 30 days
For each competitor, look at recent posts and separate outliers from the pack:
- Top 3 posts by engagement relative to that account's own median

For each outlier, record: format, hook (first line / on-screen text), topic,
length, CTA, and your best guess at WHY it worked or didn't. Ignore raw like
counts across accounts — only compare a post to its own account's baseline.

STEP 4 — Find the patterns
Across all five accounts, answer:
- Which hook styles consistently outperform?
- Which formats are winning right now in this niche, and which are decaying?
- Which topics are saturated (everyone posting, mediocre performance) vs.
  underserved (rare, but performs when someone does it)?
- What does the pack all do that nobody's outliers actually reward?

STEP 5 — Tactical advice for me
Compare those patterns against my own content from Step 1 and give me:
- 3 things my competitors do that I should start doing, with specifics
- 2 things I'm doing that the data says to stop or cut back
- 5 concrete post ideas for my next two weeks, each with a proposed hook,
  format, and the competitor evidence it's based on
- 1 gap nobody in this niche is covering that I'm positioned to own

Rules: cite specific posts (account + description + date) as evidence for every
claim. If the data is thin or you're inferring, say so. No generic advice —
if it would apply to any account in any niche, cut it.

Copy thisCopy

You have access to my Instagram account, @davecto. Analyze my last 20 posts
and give me a tactical audit. Be specific and evidence-based, not encouraging.

CONTEXT
[Give some context about your account, goals, who you are. Having Claude or
ChatGPT write this paragraph for you is fine.]

WHAT TO DO
1. Pull the actual numbers for each of the 20 posts: views, reach, likes,
   comments, shares, saves, follows-from-post, and average watch time or
   retention if available. Include the date and format (Reel, carousel, static).
2. Rank the posts by shares and saves first, then follows. Treat views as a
   secondary signal — I care about what makes people keep and forward, not scroll.
3. For the top 5 and bottom 5, transcribe or describe the first 3 seconds:
   the spoken hook, the on-screen text, and the opening visual.
4. If you see the same or very similar video most likely it's a trial reel.
   Keep that in mind for your statistics.

WHAT TO TELL ME
- Hook patterns: what do my top performers open with that the bottom performers
  don't? Name the pattern, quote the actual hooks, cite the numbers.
- Format: which format wins per pillar, and where retention drops off in my
  worst Reels.
- Pillar performance: which of my four pillars over- and under-performs, and
  whether that's the pillar itself or my execution of it.
- Mechanics: length, caption structure, CTA placement, posting time, hashtags —
  only where the data actually shows a difference. Say so if it doesn't.
- One adjacent format or topic I'm not doing that my top-performing posts suggest
  would land, with the reasoning tied to specific posts.

RULES
- Every claim needs a post and a number attached to it. If you can't retrieve a
  metric, say "not available" — do not estimate or infer it
- 20 posts is a small sample. Flag when a pattern rests on one or two posts.
- Skip praise and general best practices I could get anywhere. I want what's
  true about my account specifically.
- End with the 3 highest-leverage changes, ranked, each phrased as something I
  do differently on my next post.

Copy thisCopy

You have access to my Instagram account. I want a forensic analysis of how I END my videos.

CONTEXT ON MY ACCOUNT
[Give some context about your account, goals, who you are. Having Claude or
ChatGPT write this paragraph for you is fine.]

I care about retention and engagement, not follower vanity.

STEP 1 — PULL THE DATA
Look at my last 30 video posts. For each one, record:
- Post date, topic, and a one-line description of the content
- Total views
- Shares / Comments / Likes and View Count

If you cannot access any of these metrics, say so explicitly and tell me
which ones are missing. Do not estimate or infer numbers you don't have.
I would rather have a thin analysis of real data than a rich analysis of
invented data.

STEP 2 — CLASSIFY THE ENDINGS
Transcribe or describe the final 5-8 seconds of each video. Then classify
each ending into a type. Build the categories from what you actually see
in my content rather than a preset list — but examples of the kind of
distinction I mean: keyword CTA, follow-ask, open loop/cliffhanger, hard
summary, emotional landing line, abrupt cut, question to the audience,
teaser for a next post.

Also note for each: how long the ending runs in seconds, whether the
on-screen text at the end matches or extends the spoken line, and whether
the video loops cleanly back to the first frame.

STEP 3 — FIND THE PATTERN
Rank the ending types by performance. For each type, give me:
- Which posts used it (name them by date and topic)
- The average end-completion rate and engagement rate for that group
- The single best and single worst performer in that group, and your
  read on why they diverged

Then answer these specifically:
- Do my keyword-CTA endings out- or under-perform my non-CTA endings on
  retention? On saves and shares?
- Is there a length threshold where my endings start losing people? At
  what second mark does the drop-off usually begin?
- Do my worst-performing endings share a structural trait — a wind-down,
  a summary, a slowdown in pacing, a change in energy?
- Which endings produced comments that were substantive versus one-word
  keyword drops? Are those different endings?

STEP 4 — GIVE ME RULES, NOT VIBES
Output 5-7 tactical rules I can apply to my next video today. Each rule
must be:
- Stated as an instruction, not an observation
- Backed by a specific post from my account, named, with its numbers
- Concrete enough that a person editing a video knows exactly what to do

Then give me three rewritten endings: take my three worst-performing
endings from these 30 posts, quote what I actually said, and rewrite each
one applying your rules.

CONSTRAINTS
- Cite the specific post for every claim. If you can't cite one, delete
  the claim.
- Do not give me advice that would apply to any Instagram account. I want
  advice that only makes sense for mine.
- If the data contradicts conventional short-form advice, say so and side
  with the data.
- Flag anything where the sample is too small to conclude from.