X analytics pattern: what daily posting + replies actually did¶
Real numbers from the maintainer's X account (indie dev, Mac apps and games, X Premium), X analytics screenshot taken 2026-10-11, 28-day window (Sep 14 → Oct 11). Totals are as X shows them. Per-day values are read off X's bar charts, so they're approximate (±10%). Use this as a pattern to check your own account against, not as a promise.
The 28-day totals¶
| Metric | Value | What it says |
|---|---|---|
| Impressions | 2K | ~80% of them came in the last 3 days |
| Engagement rate | 1.6% | 33 engagements / ~2K impressions |
| Engagements | 33 | |
| Likes | 18 | Weakest signal for ranking (03-how-x-ranks-posts.md) |
| Replies received | 6 | |
| Bookmarks | 2 | |
| Reposts · Shares | 0 · 0 | Nothing yet that people want to send on |
| Profile visits | 7 | 0.35% of impressions: the bottleneck |
| Verified followers | 55 / 1.5K | |
| Active followers (24 h) | 352 / 1.5K | ~23% of followers were active that day: an audience is there |
Three phases¶
| Phase | Days | Activity | Impressions / day | Follows |
|---|---|---|---|---|
| A · Silent | Sep 14 → Sep 30 | ~2 posts, ~2 replies in 17 days | ~0–10 (one ~30 spike on the day with 1 post + 2 replies) | ~+2 |
| B · Light | Oct 1 → Oct 8 | ~1–2 posts most days, a few replies | ~15–95 (avg ~45) | ~0 net |
| C · Daily routine | Oct 9 → Oct 11 | 3 → 5 → 6 posts, 5 → 7 → 9 replies a day | ~195 → ~855 → ~580 | ~+21, then ~+7 (a few unfollows) |
activity / day ▏ 0–1 ▏ 1–2 ▏ 8 → 15
impressions / day ▏ ~5 ▏ ~45 ▏ ~200 → ~850
A silent B light C posts + replies
What the pattern says¶
- Nothing in, nothing out. Phase A: no posts, no reach. Followers alone don't bring impressions.
- Replies are the multiplier. Phase C had roughly 4× the actions of phase B and about 10–15× the impressions per day. The days with the most replies were the days with the most reach. This matches the Premium reply lever in 06-reply-playbook.md.
- Reach isn't the problem now, the profile is. 2K impressions → 7 profile visits. Before doing more volume, fix the bio, banner and pinned post so a visit turns into a click (social-x-profile-funnel).
- Zero shares, zero reposts. Shares weigh far more than likes. Add Wednesday posts people want to send to someone: checklists, tools, numbers (05-post-formats.md).
- Speed has a price. On the same day as the jump, X put a temporary reach label on the account (removed about 2 hours later after review), and a profile photo change removed the checkmark until re-review. Ramp up over weeks, and never change photo / name / handle in a push week (08-rules-and-risks.md, 02-profile-setup.md).
The routine to keep (from phase C, made safe)¶
| Daily | Target |
|---|---|
| Original posts | 1–3 (one good one beats five) |
| Replies to target accounts | 5–10, by hand, each adds a fact, number or screenshot |
| Answer replies on your own post | All, within 1 hour |
| Links | Bio or first reply only |
| Follows | Only people you'd actually reply to; no bursts |
Daily steps: daily-flow.md.
The one number to watch every day¶
Profile visits, last 7 days (X analytics → range 7D → Profile visits). Read it every morning, write it down.
Why this one: it sits in the middle of the chain (post → profile visit → link click → trial → paid). Impressions can rise from posts nobody cares about; likes barely move ranking; sales are too rare to read daily. A profile visit means someone wanted to know who made the post, and it's the number stuck at 7 in the baseline.
| Week | Profile visits, last 7 days (target) | ≈ per day |
|---|---|---|
| 1 · Oct 12–18 | ≥ 5 | 0.7 |
| 2 · Oct 19–25 | ≥ 8 | 1.1 |
| 3 · Oct 26 – Nov 1 | ≥ 11 | 1.6 |
| 4 · Nov 2–8 | ≥ 14 | 2 (→ ~50 in 28 days) |
Goals set from the baseline, not a forecast.
How to read it:
| What you see | Means | Do |
|---|---|---|
| Visits up | Posts and replies reach the right people | Keep going; on Sunday check link clicks (utm_source=x) |
| Impressions up, visits flat | Reach, but the wrong people or nothing makes them curious | Reply to buyer accounts, not growth accounts; end posts with you ("I build Mac apps…") |
| Both flat 3 days in a row | Not enough replies, or the replies add nothing | Back to 5–7 replies right after post 1, each with a fact or screenshot |
| Visits drop sharply in a day | Possible reach label | Check notifications / Under the Hood; drop to week-1 volume |
Once a week (Sunday), add one money check: link clicks on utm_source=x and installs / trials from them.
Next 4 weeks: targets, not results¶
These are goals to work toward, set from the baseline above. They are not results and not a forecast.
| Metric | Baseline (28 d to 2026-10-11) | Target (28 d to 2026-11-08) |
|---|---|---|
| Impressions | 2K | 10K |
| Profile visits | 7 | 50 |
| Profile visits ÷ impressions | 0.35% | 0.5% |
| Bookmarks + shares + reposts | 2 | 15 |
Link clicks → landing (utm_source=x) |
not tracked | tracked every week |
| Temporary labels | 1 | 0 |
Log (add a row every 4 weeks)¶
Copy the numbers from X analytics on the same weekday each time, so windows compare.
| 28 days to | Posts / day | Replies / day | Impressions | Eng. rate | Profile visits | Bookmarks + shares | Followers | Note |
|---|---|---|---|---|---|---|---|---|
| 2026-10-11 | ~1 (C: ~5) | ~1 (C: ~7) | 2K | 1.6% | 7 | 2 | 1.5K | Baseline. Started the daily routine Oct 9; label + checkmark review Oct 11 |
| 2026-11-08 |
Weekly numbers go in templates/weekly-numbers.md; what to do when a number is stuck: 07-metrics-and-stop-rules.md.