A hook testing system is three things: a log where every post's opening line is recorded with a label, a weekly slot where you test one hook idea against your normal baseline, and a rule for what counts as a win. You can build it in a spreadsheet in about an hour. Without it, "hook testing" is just posting different openings and remembering the ones that felt good.
TLDR: Log every hook with a type label, test one hook type per week with everything else held steady, and score each post against your own median. In Mydrop's data a normal Instagram post earns between 1.1% and 6.7%, so one post proves nothing and the log is what finds the pattern.
Why a system beats "try some hooks"
Post-to-post variation is large. In Mydrop's September 2026 data, the middle half of Instagram posts ran between 1.1% and 6.7% engagement, and the median was 3.1%. A single post that beats your last one by a point can be luck.
That is why one-off tests mislead. What works is the same test repeated until a pattern shows, and that needs a record. If you want the single-test method, read how to run one clean hook test. This post is about what you build around it.
Step 1: set up the log (10 columns)
One row per post. Use a spreadsheet, a Notion table or any tool you will actually open.
| Column | Example |
|---|---|
| Date and platform | 14 Oct, Instagram Reel |
| Hook text | "I deleted 40 posts and my reach went up" |
| Hook type | Result, Question, Mistake, Number, Story, Contrarian |
| Format | Reel, carousel, photo |
| Topic | Pricing, behind the scenes, tips |
| Hour posted | 9 am (audience's local time) |
| Reach | 1,240 |
| Engagement rate | 3.8% |
| Score | Rate divided by your median rate |
| Notes | Weak second line? Strong thumbnail? |
The Score column is the one that matters. A 3.8% post on an account whose median is 3.1% scores 1.2. Anything above 1.2 is a candidate winner and anything below 0.8 is a candidate miss.
Step 2: label hooks by type, not by wording
You will never post the exact same sentence twice, so compare types. Six are enough to start:
- Result: "How we cut approval time to one day."
- Question: "Do you check your analytics on Fridays?"
- Mistake: "The caption mistake that hides your posts."
- Number: "7 posts I would delete tomorrow."
- Story: "Last week a client sent back a draft with one word circled."
- Contrarian: "Posting daily is not the answer."
Tag each row when you log it. After 20 to 30 posts, filter by type and look at the average Score for each.
Step 3: run one test a week
The test is the same every week:
- Pick one hook type to try this week (for example Mistake).
- Keep everything else the same: your usual format, topic mix and posting hour.
- Publish two or three posts with that hook type and log them.
- Compare their average Score to your recent posts using other hook types.
Change only one variable. Our timing data shows why: posting between midnight and 6 am local time ran 36% below the median on Instagram, so a hook that "failed" at 3 am may simply have been posted at the wrong hour.
Step 4: decide with a simple rule
Pick the rule before you look at results, so you do not argue yourself into a win.
- Keep it: a hook type with an average Score above 1.1 over at least 5 posts.
- Drop it: an average below 0.9 over at least 5 posts.
- Anything in between: not enough signal, keep testing.
Winning hooks go into a hook library: a second tab with the hook text, type and the Score it earned. Next time you are stuck, start from the library instead of a blank page.
Where Mydrop saves time
The log is only as good as the numbers you put in it. Mydrop's performance view shows each post against that account's own median, which is the Score column done for you. You can plan your weekly test posts in the calendar so the hour stays fixed, and AI captions can draft several hook variants for you to pick from. If you run several accounts or clients, workspaces keep each account's log separate.
Start the log tonight with your last 15 posts
Create the log tonight with the ten columns above and fill in your last 15 posts. Pick one hook type for next week and write it at the top of the sheet. After four weeks you will have a first honest answer about what stops your audience, and Mydrop is free to start if you want the scoring built in.
Google review
Trustpilot review
Product Hunt comment
