Next 30 days. The actual things to do this week.
Week-by-week task list pulled from the strategy critique, the wedge offer spec, and the sales system. Each task = action + thing + acceptance check.
Week 1
Foundations + Wedge LiveGoal: AI/LLM Security Review offer page live. Sample report drafted. First outreach batch out. Insurance quote requested.
- Pick the domain. tagwercher.io or tagwercher.com. Register the missing one. Set the other to redirect.Done when: both domains resolve, one is canonical, no mixed-branding docs go out again.
- Publish the AI/LLM Security Review offer page on the canonical domain. Use the verbatim offer block from
ai-llm-wedge-offer.md. Carrd, Astro, or plain HTML is fine.Done when:/ai-llm-security-reviewis live with scope, price, and a contact form. - Draft the sanitised Acme AI sample report. 8 to 12 pages. OWASP LLM Top 10 structure. 3-4 realistic findings.Done when: the markdown is written. PDF render happens Day 5.
- Rewrite the LinkedIn headline + About. Headline leads with "AI/LLM Security Reviews for SaaS founders, OWASP LLM Top 10, 3-day turnaround." About section: thesis + business background + offer link in Featured.Done when: any visitor lands and reads the positioning in 5 seconds.
- Build the first outreach list of 15 prospects. 3 from Product Hunt new AI launches, 3 from YC W25/S25 directory, 3 from Indie Hackers, 3 from AI Tinkerers recaps, 3 from DACH LinkedIn. Sheet schema from
sales-system.md.Done when: 15 rows in the sheet with Name, Company, URL, LinkedIn, AI feature URL, source channel. - Free-find 10 of the 15 prospects. 12 minutes per prospect using the recon script (prompt injection probe, exposed system prompt, rate-limit test, output sanitisation check).Done when: 10 findings written in the sheet, each with one sentence + reproduction + business impact.
- Send 10 free-finding emails by hand. Tuesday or Thursday. Personal Gmail or Proton. Use Hook 1 with real observations swapped in.Done when: 10 sent, no automation, no spintax, no Instantly.
- Request cyber liability + E&O insurance quotes. Hiscox (UK/EU) or Embroker (US). Online quote takes 20 minutes.Done when: 2 quotes in inbox, ready to bind in Week 2.
Week 2
Send rhythm + insurance boundGoal: Sample report PDF live. Insurance bound. Send rhythm locked in. First replies coming back.
- Render the sample report to PDF. Host at
/sample-report.pdfon the canonical domain. No email gate.Done when: a stranger can download it in one click. - Bind cyber liability + E&O insurance. Pick Hiscox or Embroker based on invoicing entity. Bind, request COI PDF, store in 1Password.Done when: COI PDF in hand, effective-date earlier than any SOW signature date.
- Source the second batch of 15 prospects on Monday using the 5-channel split.Done when: 15 new rows in the sheet, source channels balanced.
- Send Day-3 bump to last week's batch. Subject "Re:" the original. Short. Offer the other 3 findings or the full review.Done when: 10 bumps sent, sequence rule respected (3 touches max).
- Send 15 fresh free-finding emails this week. Tuesday batch + Thursday batch split.Done when: weekly send target hit, every email had real recon behind it.
- Send 15 LinkedIn connection requests in parallel with the email batch. 110 characters max. Reference where you found them.Done when: connection notes are specific, not generic "add to network".
- Reply to every responder within 2 business hours. Use the Reply 1-5 templates from
sales-system.md.Done when: zero responders left waiting more than 2 business hours. - Publish LinkedIn post #1. Topic: a thesis insight or a recent prompt-injection write-up. 250-400 words.Done when: live, tagged, anchors the positioning.
Week 3
Scoping calls + content cadenceGoal: First scoping call booked. Content cadence at 2-3 posts/week. Source channels showing which ones convert.
- Source 15 new prospects. Monday block, same 5-channel split. Update Sheet.Done when: 45 cumulative prospects in the sheet by end of Week 3.
- Free-find + send 15 new emails. Same rhythm: Tuesday recon, Wednesday/Thursday send.Done when: weekly send hit, recon time still ~12 min per prospect.
- Send Day-7 case-study angle to Week 1 batch. Last touch. Subject leads with consequence of not fixing.Done when: Week 1 sequence is now closed (3 touches done, prospects move to Replied or Dormant).
- Send Day-3 bumps to Week 2 batch.Done when: 15 bumps sent.
- Book first scoping call. When a qualified reply lands, offer the 20-minute scoping call link (Cal.com free tier).Done when: at least one calendar invite accepted.
- Publish LinkedIn posts #2 and #3. Mix: one tactical ("what I look for in a 30-min audit"), one CVE breakdown or recent attack analysis.Done when: live, 50+ impressions each, profile views ticking up.
- Run the Friday pipeline review. 30 minutes. Update Stage column, Next Action column, log what is working.Done when: sheet is sorted by Next action date, today's work is at the top.
Week 4
First close + curriculum on-trackGoal: First paid engagement signed. Sample report replaced with sanitised real client work next month. Curriculum Phase 1 complete.
- Source 15 new prospects + send 15 new free-finding emails. Keep the rhythm. Volume compounds.Done when: 60 cumulative prospects by end of Week 4.
- Run the scoping call(s) from Week 3. 15-20 minutes. Confirm yes-go signals (live AI feature, founder-led decision, pre-Series-B, specific trigger event).Done when: scoping data captured for each qualified prospect.
- Send the first SOW. Use Template 1 from
contracts-pack.mdwith AI/LLM Security Review scope. Find-replace placeholders, attach sample report PDF.Done when: first proposal in client's inbox. - Require the signed Authorization Letter before any testing traffic. Template 4 from
contracts-pack.md. Signer must be VP or higher.Done when: signed PDF stored before first test request goes out. - Collect 50% deposit on signed SOW. Stripe or Wise invoice link. Non-refundable.Done when: payment hit before kickoff call.
- Deliver the first AI/LLM Security Review. 3 days. 12 billable hours. Day 1 LLM01-05, Day 2 LLM06-10, Day 3 report + remediation call.Done when: report delivered, 1-hour call held, client has the final PDF.
- Publish LinkedIn posts #4 and #5. Keep cadence at 2-3 per week. One should be a sanitised observation from the live engagement.Done when: 5 posts live cumulative, profile views compounding.
- Friday review. Look at source-channel data across 60 prospects. Re-balance the 15/week split for Month 2 based on which channels actually replied.Done when: Month 2 channel allocation updated in the sheet.
sales-system.md Mistake 1-3 and look at finding quality first.The weekly rhythm (sustainable past Week 4)
One operator, about 10 hours per week on sales, split across 5 days. This is what makes 15 prospects per week sustainable without burning out or starving delivery time.
- Monday AM (90 min). Source 15 prospects from the 5 channels. Update sheet.
- Tuesday AM (3 hrs). Free-find. One finding per prospect, ~12 min each.
- Wednesday AM (90 min). Send batch 1. 8 emails + 8 LinkedIn requests.
- Thursday AM (90 min). Send batch 2. 7 emails + 7 LinkedIn requests.
- Friday AM (2 hrs). Reply to responders. Send Day-3 bumps to last week's batch. Send Day-7 case-study angles to the batch before.
- Friday PM (30 min). Pipeline review. Advance stages. Log what is working.
If a day gets blown out, the rule is to send fewer prospects that week, never to skip recon and mass-blast generic openers.
Sources: strategy-critique.md (week-by-week 6-month grid), ai-llm-wedge-offer.md (Week 1 TODO + decisions), sales-system.md (Mon source / Tue free-find / Wed-Thu send / Fri reply rhythm).