The first 29 days since the account's first Twitter Together post on Sep 7, set against the 60 silent days before it. Prepared for the ETC marketing working group. It opens with an executive summary and six KPIs, each tied to a decision. The detail behind them follows.
The relaunch worked and the message is landing. What's holding the account back is that it only talks to its own circle. The next gains come from replying to and quote-posting accounts outside the ETC space.
The ask: approve the reply amendment, or a 30-day reply pilot, so @eth_classic can take part in conversations outside its own circle. Judge it on the scorecard after 30 days by conversation per post, follows per 1,000 impressions and active followers.
The numbers to manage the account by. Values cover Sep 7 – Oct 5. Targets are proposals for the working group to confirm.
Operating check, not scored: 8.2 original posts a week against a 5–6 plan. Pillar posts ran 1.3 Predictability, 2.3 Convergence and 1.3 Longevity a week, against a plan of 2, 1 and none. 11 of 34 posts went out in the 6–10am Pacific window.
The posts are working. What's missing is conversation. For four weeks @eth_classic mostly announced things. Its interactions with other accounts came to 2 replies, 5 quote posts and 1 repost, all with the same three accounts, and the numbers that depend on conversation fell every week. The framework's next phase (weeks 5–8) is judged on replies and quote-posts, so this is the most important decision in front of the working group.
Posting isn't the gap. The account ran 9–11 posts a week, but impressions per post fell from 4,674 to 2,095 and conversation per post from 5.4 to 1.8.
Ten of the 34 posts were notices: 9 community-call posts and a CoinGecko ticket. Community posts drew reach but had the lowest engagement rate of any group (2.1%).
The 5 quote posts all quoted those same three accounts, and they ran below average: 2.4% engagement against 3.0% for other posts. Quoting our own circle doesn't reach anyone new.
When the account speaks in its own voice, people respond. Four of the five highest-engagement posts were short identity lines like "Still Alive" and "Don't Trust, Verify."
Only about 5,000 followers are active. An account that only broadcasts keeps reaching the same small group, and replies are how it reaches new people.
Every benchmark account that sustained momentum posted into live conversation. None held its position on scheduled posts alone.
Quote-posts were the common way peer accounts borrowed conversation, using other people's work as proof instead of making every claim themselves.
Official handles act as a distribution surface for other people's proof: named apps, operators and builders.
The v3 strategy already says posting is only half of it. Replying, quote-posting builders and turning up in threads we didn't start show that the account is attended, not scheduled.
The current workflow handles planned posts well, but a reply needs a full review cycle, so the account can't take part in conversations as they happen.
Every reply follows the Fact Sheet rules: no price talk, no hostility, sourced claims. Each one is logged in the reply list so there's a full record.
Quote real work from builders and the priority accounts, adding one factual line, within the boundaries the working group sets.
Look at conversation per post, follows per 1,000 impressions and active followers. If they don't move, we rethink it.
Daily impressions over the full 89 days in the export. The gray bars are the silent stretch before Sep 7.
Daily follows and unfollows come from X. The running total comes from Fedica, whose history starts Sep 23.
Sep 7 – Oct 5. The shaded band is relaunch week. Hover or use the arrow keys for each day's values.
Running total of follows minus unfollows since Sep 7 (X).
Fedica's snapshot at the start of each day.
X counts 371 follows and 86 unfollows, net +285. Fedica puts the total at 27,348 on Oct 6, up 75 from Sep 23. Working back from that, the account started the period at about 27,100.
Fedica counted 43 unfollows on Oct 5. 36 of them were accounts that were removed, suspended or deactivated. The other 7 were real unfollows, the same number X reports for that day.
The Oct 5 drop was an account cleanup, not a reaction to a post. Fedica counts deleted accounts as unfollows and X doesn't, which likely explains the rest of the gap between them (118 against 48 over Sep 23 – Oct 5). X's unfollow count is the one to track.
Impressions and posts share one timeline, so each day's reach reads against what went out that day.
Replies and reposts aren't counted. A gap means no original post that day.
Posting more on a day didn't cut reach per post. In weeks 3–4, days with two to four posts drew about 2,500 impressions per post, the same as single-post days, and about 5,800 in total against 2,500.
The fall in impressions per post follows time more than volume: the relaunch bump fading, and newer posts having had less time to collect impressions. Cutting volume would cost total reach.
X counts six actions as engagements: likes, reposts, replies, bookmarks, shares and profile visits. Engagement rate is their total divided by impressions.
The two lowest days, Sep 12 and Oct 3, were both Saturdays with no post.
All 2,687 engagements, Sep 7 – Oct 5.
Likes and reposts are 70% of engagement. Replies are 2.3%, about 2 a day.
Bookmarks are 1.2% of engagement, 32 in total.
Profile visits are a quarter of engagement, so posts are sending people to the profile. The bio and pinned post are worth keeping current.
All 34 original posts, tagged with the narrative framework's pillars, using the tweet bank's tag where the post came from the bank. Results come from X's post export. The two Oct 5 posts are under a day old and are left out of the averages.
Lifetime impressions per post as of Oct 6.
Six engagement actions ÷ impressions, all posts in the group.
| Pillar | Posts | Avg impressions | Engagement rate | Follows per 1K impressions | Bookmarks a post | Replies + quotes a post |
|---|---|---|---|---|---|---|
| Predictability | 4 | 2,364 | 3.3% | 0.32 | 0.25 | 1.8 |
| Convergence | 8 | 2,749 | 3.3% | 0.45 | 1.50 | 2.2 |
| Longevity | 4 | 3,322 | 2.8% | 0.23 | 0.25 | 4.2 |
| All narrative pillars | 16 | 2,796 | 3.2% | 0.36 | 0.88 | 2.6 |
| Community | 11 | 3,672 | 2.1% | 0.47 | 1.00 | 2.5 |
| Utility / Dev | 1 | 1,821 | 2.1% | 0.00 | 0.00 | 2.0 |
| Brand teaser | 2 | 2,483 | 4.5% | 0.20 | 0.50 | 4.0 |
| Relaunch | 1 | 5,830 | 5.2% | 1.54 | 3.00 | 12.0 |
| Admin | 1 | 552 | 1.3% | 0.00 | 0.00 | 0.0 |
| All other posts | 16 | 3,348 | 2.6% | 0.54 | 0.94 | 3.1 |
Predictability and Convergence both ran a 3.3% engagement rate, and Longevity 2.8%. Community posts reached the most people per post (3,672) but had the lowest rate (2.1%).
Convergence is the pillar people save: 12 of the 29 bookmarks, 1.5 a post against 0.25 for the other two pillars.
X credits only 45 of the 371 new follows to a specific post. The rest came through the profile, search and recommendations, so follow counts per post (0–9) are too small to rank pillars yet.
Against the weekday plan, Convergence ran at 2.3 posts a week (plan: 1) and Predictability at 1.3 (plan: 2). Utility / Dev ran once.
Predictability engages as well as Convergence and is below plan, so it's the place to add posts. Convergence earns the saves, so hold it at one to two a week rather than cutting it to one.
The framework's weekday slots against the three strategy weeks.
| Pillar | Plan a week | Actual a weekSep 14 – Oct 4 | Against plan |
|---|---|---|---|
| PredictabilityMon + Thu slots | 2 | 1.3 | ▼ Below plan |
| ConvergenceWed slot | 1 | 2.3 | ▲ Above plan |
| CommunityTue slot | 1 | 2.7 | ▲ Above plan |
| Utility / DevFri slot | 1 | 0.3 | ▼ Below plan |
| LongevityNo fixed slot | – | 1.3 | Not slotted |
| Brand teaser + adminNot in the plan | – | 1.0 | Not slotted |
| Total | 5 | 9.0 |
What the account did beyond its own posts, Sep 7 – Oct 5.
| Post | Impressions | Eng. rate |
|---|---|---|
| Still Alive. Still EVM Compatible. Still Proof of Work. Still…Sep 21 · Longevity | 1,855 | 5.3% |
| It's been too quiet around these parts for far too long.... That…Sep 7 · Relaunch | 5,830 | 5.2% |
| The Future is Coming on.Sep 21 · Brand teaser | 2,608 | 4.7% |
| #EthereumClassic = Programmable Digital GoldSep 10 · Convergence | 3,463 | 4.3% |
| Don't Trust, Verify.Sep 16 · Brand teaser | 2,358 | 4.2% |
| Post | Impressions | New follows |
|---|---|---|
| It's been too quiet around these parts for far too long.... That…Sep 7 · Relaunch | 5,830 | 9 |
| Attention all ETC builders, miners, & stakeholders 📢 This Weeks $ETC…Sep 7 · Community | 6,593 | 7 |
| ETC is Programmable Digital Gold. #ETH programmability. #BTC…Sep 17 · Convergence | 3,395 | 6 |
| ETC Community Call #60 Now Live! 🔥 · Codeaholic’s first ML-DSA gas…Sep 21 · Community | 5,125 | 4 |
| ETC Community Call #60 🎙️ Mark your calendars! Same place. New time.…Sep 16 · Community | 5,303 | 2 |
Monday-to-Sunday weeks. Week 1 is relaunch week; weeks 2–4 ran on the content strategy. The full weekly table is in the appendix.
Shaded band: 5–6 a week target
Weekly average
Engagements ÷ impressions
Follows minus unfollows
Per-day averages, so periods of different lengths compare fairly.
| Per day | Before relaunchJul 9 – Sep 6 · 60 d | Relaunch weekSep 7 – 13 · 7 d | Content strategySep 14 – Oct 5 · 22 d | Strategy vs. relaunch |
|---|---|---|---|---|
| Impressions | 149 | 3,216 | 3,992 | ▲ 24% |
| Engagement rate | 0.61% | 2.79% | 2.35% | ▼ 0.44 pts |
| Likes | 0.5 | 38.9 | 53.1 | ▲ 37% |
| Reposts | 0.0 | 12.0 | 16.0 | ▲ 33% |
| New follows | 3.2 | 14.7 | 12.2 | ▼ 17% |
| Unfollows | 2.4 | 1.9 | 3.3 | ▲ 79% |
| Net new followers | 0.8 | 12.9 | 8.9 | ▼ 31% |
| Posts per week | 0 | 5.0 | 9.2 | ▲ 85% |
Relaunch week's engagement rate is lifted by Sep 7 itself, when the comeback day ran 4.1%. Without Sep 7, relaunch week ran 2.33%, level with the strategy period. Follows peaked in relaunch week, consistent with returning followers finding the account again. The strategy period added reach and amplification on top of that.
Fedica's estimates from follower profiles. Activity, follower size and location cover nearly all 27,348 followers. Time zones cover 7,222 of them (26%).
When each follower last posted or engaged.
How many followers each follower has.
54% of followers haven't been active in over a year. 18%, about 5,000 accounts, were active in the last 30 days. 87% of follower accounts are more than two years old.
74% of followers have fewer than 100 followers of their own. 210 have 10,000 or more, and 163 of those were active in the last 30 days. After dropping off-topic, protected and paid-promotion accounts, 79 are worth replying to, 22 of them first.
This is a legacy audience that has mostly gone quiet, not one padded with new accounts. Active followers are the number to grow, and the 22 priority accounts on the reply list are where replies should start.
Fedica's estimated location for 27,242 followers.
The 7,222 followers placed by time zone, and their combined reach (11.0M).
The US holds 27% of followers, then Nigeria (7%), India (6%) and the UK (4%). Fedica groups 25% of followers under Science & Technology and 17% under Business & Finance.
Reach is concentrated. US East and UK/GMT hold 16% of placed followers but 49% of their reach.
Fedica's reach by hour of the day, in Pacific time, read from its best-time view. It varies day to day; Wednesday shows only the hours that were read.
Original posts by hour published, Pacific time, Sep 7 – Oct 5.
Fedica's reach peaks at 8am Pacific (14.8% on Tuesday, 15.1% on Wednesday, its best hour of the week) and holds above 14% from 6 to 10am. From 4 to 11pm it sits at 9.2–10.3%, so the morning runs about 45% higher.
11 of the 34 posts went out between 6 and 10am. Another 11 went out between 5 and 10pm, Fedica's weakest stretch.
Scheduling only started around Sep 21. Before that, posts went out whenever they were approved, which is why so many landed in the evening. Most of those evening posts came in weeks 1–2, when every post drew more reach, so our own results can't show the timing effect yet.
The time-zone model built from follower locations points to the same 6–10am window, so two independent sources agree. The next two weeks of scheduled posts are the first clean test.
Average impressions by weekday, with the original posts made on that weekday. Each weekday has only 4–5 days of data.
Fedica rates Tuesday the best day to post. In our data Tuesday averaged 4,596 impressions, second to Monday (5,221), and Monday's average includes the relaunch day and the Sep 21 peak. Saturday and Sunday were the two weakest days.
Approve the reply amendment or a 30-day reply pilot, then reply 3–5 times a week, starting with the 22 accounts below. This is the biggest lever in the data.
Sixteen reposts of our own posts against one repost and five quote posts of others, all with the same three accounts, is a closed loop. Quote-post real work from builders and outside accounts with one factual line, within the working group's boundaries.
Close most pillar posts with "Programmable Digital Gold", whatever the pillar. Pillar posts carrying a brand line ran 3.4% against 2.4% without one, and the line showed no fatigue after seven uses.
Keep 8–10 original posts a week as long as the bank holds quality. Schedule pillar posts for 6–10am Pacific, with the strongest on Tuesday or Wednesday at 8am. Add one on Saturday and avoid evenings. More posts added total reach, so the gain is in when they go out, not how many.
Post more short identity lines, which were four of the top five posts. Bring Predictability back to two a week using drafts #14, #22 and #30, hold Convergence at one to two, and cut community calls from three posts to two.
Grow the 5,000 active followers rather than the 27K headline, and rerun this scorecard each month from the same exports.
Followers with 10K+ followers of their own who were active in the last 30 days and post about crypto, from Fedica's follower export on Oct 6. The full list of 79 is in a separate spreadsheet with status tracking. Groups come from keyword matching on bios, so check each profile before replying.
| Account | Followers | Why reply |
|---|---|---|
| ETC community · 3 | ||
| @ETCGrantsDaoETCGrantsDao | 44,662 | Already in the ETC community |
| @CobraCryptoCobra Crypto | 13,769 | Already in the ETC community |
| @Crypto_Sketch_CryptoSketch.ETC | 10,375 | Already in the ETC community |
| Mining and proof-of-work · 6 | ||
| @CryptoMiners_CoCrypto Miners | 108,230 | Proof-of-work and mining audience |
| @SoloSatoshiSolo Satoshi 🇺🇲 | 37,383 | Proof-of-work and mining audience |
| @LitecoinDotComLitecoin | 24,844 | Proof-of-work and mining audience |
| @HobbyistMinerTheHobbyistMiner | 24,281 | Proof-of-work and mining audience |
| @MiningRabidRabid Mining | 12,113 | Proof-of-work and mining audience |
| @AssetKeenJC Keen | 10,676 | Proof-of-work and mining audience |
| Crypto media and research · 6 | ||
| @WuBlockchainWu Blockchain | 559,057 | Crypto media or research; reach beyond our base |
| @JoshuwaJoshuwa Roomsburg | 271,967 | Crypto media or research; reach beyond our base |
| @CryptoLogicHQCryptologicFlow | 39,532 | Crypto media or research; reach beyond our base |
| @VladcosteaVLAD HOSTS THE BEST PODCAST IN BITCOIN | 26,010 | Crypto media or research; reach beyond our base |
| @zenbirbZenbirb | 12,576 | Crypto media or research; reach beyond our base |
| @SynopsisDataSynopsis Data | 11,534 | Crypto media or research; reach beyond our base |
| Picked by hand · 7 | ||
| @BarrySilbertBarry Silbert | 849,087 | Founder of DCG, chairman of Grayscale; industry-wide reach |
| @qtumQtum | 271,616 | Smart contract platform forked from Bitcoin Core and Ethereum; a peer chain |
| @EdgeWalletEdge | 38,184 | Wallet focused on privacy, security and usability |
| @BitsBeTrippinBitsbetrippin | 29,465 | Bitcoin and crypto tech education |
| @iamtextureTexture | 23,158 | Ethereum commentator; fits Convergence debates |
| @GivethGiveth | 19,341 | Ethereum public-goods donations project |
| @BobSummerwillBob Summerwill | 18,990 | Ethereum historian, formerly at the ETC Cooperative; Longevity pillar |
The detail behind the charts above.
From X's post export, with quotes from Fedica. Gray rows are less than a day old. Post text links to the post on X.