theStacc link strategy: reworked recommendations

4 November 2026. Supersedes the "30 priority actions" in priority-30.csv and report.html. Those files stay as the raw inventory.

Nothing here has been submitted, emailed or posted.

1. Diagnosis: links are now the bottleneck

The August SEO master plan could not measure backlinks because it had no data. The on-site work it prescribed has largely shipped since: deletes, redirect fixes, titles and metas on 4,672 pages, and an internal-linking release. Commercial pages were ranking at positions 15–45. The question it left open was whether authority holds them there.

Common Crawl's web graph (Aug–Oct 2026, 135M domains, 1.8B links; free) now answers it:

theStacc ranks 28th of the 34 domains measured. It has 3–6x fewer referring domains than products at the same stage. File: authority/targets.csv.

Caveats: Common Crawl counts every link it crawled, nofollow included, and its coverage differs from Ahrefs. The comparison between sites is fair; the absolute numbers are not Ahrefs numbers.

2. What changed from the earlier recommendations

3. The plan, in order

Week 1: fix what already exists (hours)

1. Shopify noindex. Find out why apps.shopify.com/thestacc is noindex in the Partner dashboard; I haven't confirmed the cause. Add /integrations/shopify/ on thestacc.com, linking to the listing and back.
2. Third-party facts. Capterra shows $126/$199 plans, SourceForge "from $49 + 3-day trial", GoodFirms "start for $1". Current pricing: Blog $99, Local $49, Social from $49, managed from $749. Correct every listing.
3. Reviews. Trustpilot, Capterra and Shopify all show 0 reviews. A Trustpilot profile with 0 reviews is noindexed. Ask real customers for reviews and offer them no incentive. The listing sites' own rules matter here, and so do the no-fabrication rules.
4. Zero-click study. Correct the date range, add a data download and a "how to cite" line.
5. npm @thestacc/mcp-server. It is v0.3.0 and its repository points at a private repo that returns 404. Republish 0.6.0 with correct metadata.
6. Black Friday deal lists (deadline this month). GetResponse's SaaS Black Friday list and ContentStudio's deals list both link to vendors with followed links and are being updated now. Submitting needs a real deal, which is your call.

Weeks 1–3: MCP directories (1–2 days of work)

The code is ready: OAuth 2.1 + PKCE, dynamic client registration, Streamable HTTP, and annotations on all 55 tools. What's missing: - an MCP/AI-clients section in /legal/privacy/ (it currently says nothing about MCP) - a server icon - a reviewer account - server.json for the official registry

Then submit to the Claude connectors directory and Smithery (both followed links), the official MCP registry, and the ChatGPT apps directory. Also fix the tool count: the site says 56 tools, the code has 55. Skip mcp.so and mcpservers.org's $39 "dofollow" upgrades. Buying links that pass ranking credit breaks Google's spam policies.

Weeks 2–10: roundup inclusion (the main lever)

outreach/pitch-list.csv has 44 tier-1 pages. Each one: - is a legit publisher or agency - fits a theStacc product - links competitors with followed links - has a free route in - names an author or contact where one exists

Fit split: AI SEO/blog content 36, local SEO 18, white label 6, reviews 5, free tools 3. Pages fit more than one category.

Strongest first: SEO agencies' tool roundups (Searchbloom, Siege Media, Search Influence, TripleDart, WebFX), AgencyAnalytics' blog, Local SEO Guide's tools directory (Andrew Shotland), mktg.tech (free application, "not for sale"), Swetrix's "SEO autopilot software" list (closest product match), Screpy's awesome-seo-tools, Small Business Trends, Fit Small Business, DreamHost, Zapier's blog.

Prepare before pitching: - one page per fit category with true claims only - a reviewer account - screenshots - the affiliate program (20% recurring), because many of these lists are affiliate-funded

Pitch only where theStacc genuinely fits the list. "Best AI visibility tracking tools" is the most common list type in the data, but theStacc has no tracking product, only a free checker, so those lists are tier 3 unless a free-tools angle fits. Another 84 tier-2 pages are lower-fit, nofollow or harder.

Expect most pitches to get no reply. Even so, this is how the peer set built its link profiles, and these pages are also what AI answers draw on when asked for "best X tools". That last point is my judgment, not a measured fact.

Weeks 2–6: cheap product listings with followed links

• Ghost integrations (followed; hours, though Ghost staff choose who gets listed)
• n8n verified community node (followed; about 2–3 days)
• Webflow Apps (followed) once the deprecated OAuth flow is revived (days)

Not for links: Zapier (its listing has no vendor link; build it only if customers want it), WordPress.org (nofollow ugc), G2, Product Hunt, Capterra, AlternativeTo (all nofollow).

Months 2–3: the AI Lockout Index (biggest upside, one blocker)

A study of 140,000 US local business sites is the kind of asset that earns links from Search Engine Journal, Search Engine Land, The Register and small-business press. Comparable studies (BuzzStream, the SEL piece on WP Engine) got exactly that coverage. Targets: digital-pr/outlet-targets.csv (44 rows).

Blocker: the crawler sends AI-bot user-agent names from our own IP. CDNs block "fake bots" whose IP doesn't match the real bot. So some of the 15% "silent lockout" may be caused by the test itself. The 10.6% "Googlebot blocked" figure is the warning sign. The study needs a control before any pitch: compare against what the vendors document, and check a sample of site owners' real logs or settings. Wix's and Square's documentation contradicts the provisional figures, so both companies need a right of reply. This is Ritik's workstream (/tmp/lockout-rehearsal).

Ongoing: publishing

LinkedIn for founder reach (no link value), Substack if there's a recurring publication (followed links), Medium imports with a canonical (reach only).

4. Don't

• Mass directory submissions: most of the 215 are nofollow.
• Paid "dofollow" listings or paid placements on listicle farms (gomega, gitnux, wifitalents, zipdo, worldmetrics, the local-service-near-me network).
• Building integrations solely for a listing link.

5. Measuring it

• Re-run ~/cc-webgraph/analyze.py on each new Common Crawl graph release (free, about 30 minutes) and compare referring domains against the peer set.
• Track GSC positions for commercial pages, the ones that sat at 15–45.
• DataForSEO's balance is −$0.01. Topping it up would give monthly backlink data; that's your call.

Files

• authority/: peer comparison, our referrers, 26,473-domain link intersect, top-300 candidates
• outreach/pitch-list.csv: 163 checked pages, tiered (outreach-pages-all.csv = all 300 domains)
• verification/link-verification.csv: rel/indexability evidence for 41 platforms, raw HTML in raw/
• digital-pr/: outlet targets and comparable studies

Pitch list (interactive)

Tier 1 = legit publisher, fits a theStacc product, followed competitor links, free route. Tier 2 = lower fit, nofollow or harder. Tier 3 = weak fit or paid.

Data files

• pitch-list.csv — 163 checked roundup pages, tiered
• outreach-pages-all.csv — All 300 link-intersect domains, page-level
• targets.csv — Referring domains: theStacc vs 33 peers
• intersect.csv — 26,473 domains linking to 2+ peers but not theStacc
• thestacc-referrers.csv — theStacc's 686 referring domains
• link-verification.csv — Link rel / indexability evidence, 41 platforms
• outlet-targets.csv — Digital PR outlet targets
• comparable-studies.csv — Comparable studies and their coverage
• backlink-prospect-master.csv — Original 215-prospect inventory (superseded)

Research only. Nothing was submitted, emailed or posted. Generated 4 Nov 2026.