Skip to content

manifoldmcp/marketing-skills / skills/growth-strategy

growth-strategy

Verified

When the user wants a growth plan, a choice of marketing channels, a go-to-market (GTM) plan, a launch plan, content ideas, a content calendar, or to repurpose a video or blog post as social posts. Holds dated, sourced findings a model lacks (how the channel mix shifted in 2025-2026, what a Product Hunt launch yields now, cadence and repurposing evidence, the attribution gap, which benchmarks to distrust) and how to take a cheap cross-channel baseline. Also use when the user mentions how do I grow my startup, a growth strategy or marketing plan, more traffic, signups, leads or customers with no channel named, traction, which channels should we be on, channel mix, a GTM plan, ICP and positioning for a new product, a launch plan or checklist, Product Hunt, Show HN, a waitlist, launch reactions, a posting schedule, content pillars, what to post, or turning a video or podcast into posts. One named channel goes to its own topic (seo, ai-search, reddit, linkedin, tiktok, paid-ads, outbound, creators, digital-pr).

references/benchmarks.md

View on GitHub
# Channel benchmarks

Numbers users ask for ("what's a good CAC", "is 3% engagement good") with the method behind each, so the agent quotes them with the right weight. Most channel benchmarks are vendor data from their own customers; few disclose a sample size. Use them for orders of magnitude, then replace them with the user's own numbers within 90 days. Checked October 2026.

## Conversion and acquisition

| Benchmark | Figure | Source and method | Weight |
|---|---|---|---|
| AI referrals against organic search, conversion | 4.87% against 4.60%, not significant (p = 0.79); B2B 2.03% against 1.68% in a 33-site subset, also not significant | Amsive, September 2025: 54 sites, six months of GA4, audited macro conversions, last touch | Strongest multi-site test found; sessions, not customers |
| AI referrals, one site | 0.5% of visits, 12.1% of signups, "23x" organic's conversion | Ahrefs' own site, 30 days, June 2025 | One site, one month; never generalise |
| ChatGPT referrals in ecommerce | Converted worse than every established channel except paid social (organic and paid search, email, affiliate) on rate and revenue per session | 973 ecommerce sites, August 2024 to July 2025, last click (Kaiser and Schulze, Marketing Science, 2026) | Large; last click; early channel |
| Landing page conversion, median | 6.6% across industries (range 3.8% to 12.3%); email traffic averaged 19.3%, paid search 10.9%, paid social 12% | Unbounce 2024: 41,000+ pages, 464M visitors, 57M conversions | Large; its own customers' pages; 2023 data |
| Google and Microsoft search ads, all industries | CTR 6.64%, CPC $5.42, conversion rate 8.18%, cost per lead $66.69; CPC up slightly, cost per lead down for the first time in five years | LocaliQ and WordStream, June 2026, "thousands" of customer campaigns | No count, no medians; directional |
| B2B CAC | B2B SaaS organic $205, paid $341, blended $239 | First Page Sage, January 2026: its own clients 2022 to 2025, no sample size, weighted 75% organic | Weak; the most-quoted table has no method; no per-channel split for SaaS |
| B2B ad return | LinkedIn ROAS 121% (279% at the 75th percentile), Google Search 67%, Meta 51%; cost per company influenced €70 LinkedIn, €110 Google Search, €129 Meta | Dreamdata, March 2026, its customers' 2025 journeys | Vendor that sells B2B attribution; model not stated |

## Organic social engagement

Two large datasets disagree by an order of magnitude because they divide by different things. Never set one vendor's rate against another's.

| Platform | Buffer median, 2025 (engagements ÷ reach) | Socialinsider average, 2025 (interactions ÷ followers) |
|---|---|---|
| LinkedIn | about 6.1% (6.4% in 2024) | not covered |
| Instagram | about 5.4% (7.3%) | 0.48% (0.50%) |
| Facebook | about 5.6% (5.0%) | 0.15% (0.15%) |
| X | about 2.8% (2.0%) | 0.12% (0.15%) |
| TikTok | about 4.5% (4.4%) | 3.70% (2.50%) |
| Threads | about 3.6% (4.4%) | not covered |

Buffer: 52M+ posts by Buffer users, 2024 to 2025, accounts with 10+ posts; Buffer warns that a change can reflect who posts rather than the platform. Socialinsider: about 70M brand posts, January 2024 to December 2025; its page has internal inconsistencies (TikTok 3.70% or 3.73%), and Facebook brands halved their posting from 47 to 24 posts a month.

Use these to sanity-check an account, not to grade it: compare an account with its own median over time, and a post with the median of the account's recent posts.

## What not to quote

- Visitors or signups "per Product Hunt launch": consultant estimates without method ([launch](launch.md#what-a-launch-yields-now)).
- "AI traffic converts 4x to 23x better": built on single sites.
- Any CAC by channel without the number of companies behind it.
- Organic reach percentages ("Facebook 1-2%", "Instagram 3.5%") from secondary blogs: the original measurements could not be traced.
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 196kUpdated Claude CodeCodexVerifiedInstall