A single look at a competitor’s profile gives you totals: followers, views, likes. Totals do not tell you how those numbers arrived. Did the video peak on day two or build for a week? Did it get a second life on day six? Did the account’s follower count move when the posting style changed?
Daily snapshots answer those questions. ContHunt’s tracking, an experimental feature, lets you follow public accounts and selected videos and stores daily cumulative data, so you can come back later and read growth as a curve. This guide covers what to track, which metrics to derive and how to review them.
What to track#
Choose a small, deliberate set.
- 3 to 5 direct competitors. Brands selling to your buyer.
- 1 or 2 category leaders. They show where the category is heading.
- 1 breakout small account. Often the best source of new formats.
- Selected videos, not everything. Add individual videos that are launches, outliers or tests. Following an account does not mean enrolling every video it posts.
For accounts, the data covers followers and totals where the platform provides them. For videos, it covers views, likes, comments, shares and saves, where the platform exposes them. Some platforms hide or rename a metric, so read blanks as “not available”, not as zero.
Public data only. Tracking shows what anyone can see on a public profile. It does not reveal private analytics, ad spend or sales.
From cumulative numbers to daily change#
Snapshots store cumulative totals. The useful figure is the difference between two days.
Illustrative snapshots for one competitor video:
| Day | Cumulative views | Daily gain |
|---|---|---|
| D1 | 12,000 | 12,000 |
| D2 | 31,000 | 19,000 |
| D3 | 40,000 | 9,000 |
| D4 | 44,000 | 4,000 |
| D5 | 45,500 | 1,500 |
| D6 | 70,500 | 25,000 |
| D7 | 74,000 | 3,500 |
Three things show up that the D7 total of 74,000 hides:
- The first peak was D2 (19,000 gained).
- Half-life. Define it as the first day the daily gain falls under half the first peak. Half of 19,000 is 9,500, and D3 gained 9,000, so the half-life is 3 days.
- A second spike on D6 (25,000 gained, bigger than the first peak). Likely causes include a repost by a larger account, a trend-sound pickup or a paid push. Treat the cause as a hypothesis and check the comments and the account’s other posts that day.
The daily metrics#
| Metric | Formula | Question it answers | Watch-out |
|---|---|---|---|
| Daily gain | Views today − views yesterday | Is this video still being pushed? | Platforms revise counts, so small negatives happen |
| Day-3 and day-7 views | Cumulative at those days | How fast does this account’s content spread? | Compare videos of the same age only |
| Half-life | First day gain < 50% of the first peak | How long does the video stay alive? | Needs at least 5 days of data |
| Share rate | Shares ÷ views | Is the content passed on? | Not exposed everywhere |
| Save rate | Saves ÷ views | Is it useful enough to keep? | High on tutorials and products |
| Engagement rate | (Likes + comments + shares + saves) ÷ views | Overall response per view | Mixes strong and weak signals |
| Follower gain per post | Follower change over the 48 hours after posting | Did a video earn followers? | Several posts in a window blur it |
| Posting cadence | Posts per week | Is the account scaling or slowing? | Check for batch posting |
| Format mix | Share of posts by format | Is a new format being tested? | Needs tagging (see below) |
The ratios help you compare across accounts of different size. Reading virality explains why ratios travel better than totals.
Tag the content as well as the numbers#
Numbers tell you which videos worked. Tags tell you why you might care. For every tracked video you keep, record:
- Hook pattern (use the hook taxonomy)
- Format (demo, review, comparison, comment reply)
- Product or offer featured
- Creator or voice (founder, employee, hired creator)
- Call to action
After a month, you can answer questions like “What hook mix does Competitor A use for launches?” and “Do their posts with a product visible by second 3 hold their daily gain longer?”
Alerts worth setting#
Alerts only help if they trigger a decision. Three are enough:
- Outlier. A tracked video passes 3x its account’s median by day 3.
- Cadence change. Posting frequency doubles or halves for a week.
- New format repeats. A format you have not seen from them appears three times in seven days.
When an alert fires, spend 15 minutes. Watch the video, read 50 comments, tag it and decide whether it changes your plan.
The weekly review (20 minutes)#
Ask the same five questions each week:
- Which competitor video had the biggest outlier ratio, and what did it open with?
- Which format appeared more often than last week?
- Did any account change cadence or product focus?
- Did a launch behave like last month’s launches (similar day-3 views, half-life)?
- What is one thing we will test because of this, and one thing we will ignore?
Write the answers down. Over a quarter, the notes become a record of how competitors move, which memory and screenshots cannot give you.
Limits to respect#
- Daily means daily. A spike and a drop within one day are invisible.
- Public counts can include inflated numbers. The FTC’s final rule on fake reviews and testimonials covers buying or selling fake social media indicators such as bot-driven followers or views. A spike with no matching comments, shares or follower change is worth doubting.
- One hit is one data point. If a rival’s video worked, the cause may be the creator, the audience or timing. Test the pattern yourself with a hook test before copying it.
- Do not copy. Use the structure, the angle and the proof beat. Reshoot with your own product and voice.
Starting in ContHunt#
A sensible first week:
- Run a Deep Search to find the accounts that dominate your niche’s top videos. Use a prompt such as: “Which accounts account for the most top-performing videos in the last 30 days for sensitive-skin skincare? Group by brand versus creator.”
- Choose three to five brand accounts and one breakout creator for tracking.
- Add the launch videos or outliers you care about.
- Return in seven days, read the curves and run the weekly review.
ContHunt stores the snapshots, so the history is there when you return. A weekly trend radar pairs well with this, since tracking shows how known accounts behave and the radar shows what is arriving.
Get started by finding the accounts that matter in your niche, then set up your first tracked set.
