Delivery Plan · jennyselc.com.au · SEO & Performance Remediation

Weekly Delivery Plan & Estimates

Senior-engineer scoped · AI-assisted execution · 4 hrs/day capacity · Prepared 8 May 2026 · Earliest start Mon 11 May 2026 · Target completion Fri 5 Jun 2026

1. At a glance

4 cal. weeks

11 May → 5 Jun 2026

~74 hrs

Senior dev effort

~10 hrs

Total client time

9 releases

Across 4 sprints

Approach. One senior engineer working 4 hours per day, Monday to Friday (≈ 20 hrs/week), with AI-assisted tooling (code-gen, bulk transforms, diagnostic scripting). Estimates already account for the productivity uplift on repetitive work (bulk redirect generation, meta-tag application, content variants). Tasks dominated by infrastructure setup or external services (CDN, plugin audits, image pipeline) are estimated at conventional rates because AI doesn’t materially compress them.

Engagement schedule (calendar dates)

Calendar week	Sprint coverage	Key releases	Dev hrs
Mon 11 – Fri 15 May	Sprint W0 (kick-off, day 1) + Sprint W1 part 1 (SEO foundations)	R1.A redirects + internal links (Thu 14 May)	20
Mon 18 – Fri 22 May	Sprint W1 part 2 + Sprint W2 (bug fix & content)	R1.B titles, metas, keywords (Tue 19 May) · R2.A gallery-500 fix (Thu 21 May)	20
Mon 25 – Fri 29 May	End of Sprint W2 + Sprint W3 part 1 (performance)	R2.B sitemap + duplicates (Mon 25 May) · R3.A CDN + cache (Wed 27 May) · R3.B image pipeline (Fri 29 May)	20
Mon 1 – Fri 5 Jun	Sprint W3 part 2 + Sprint W4 (verify & close)	R3.C render-blocking + fonts (Tue 2 Jun) · R3.D plugin/script audit + lock-in (Wed 3 Jun) · Final report & handover (Fri 5 Jun)	14

Earliest start hinges on plan approval. This document is being shared on Fri 8 May for agent + client review. Mon 11 May 2026 is the earliest sprint-start date assuming approval and access land before then. If approval slips, every milestone above shifts by the same number of working days.

Sprint-by-sprint summary (thematic)

Sprint	Theme	Headline deliverables	Dev hrs	Client hrs
W0 Mon 11 May	Kick-off & access	Approval workshop · access provisioned · staging clone · diagnostics-ready	4	3
W1 Tue 12 – Tue 19 May	SEO foundations & quick wins	102 × 301 redirects · 17 titles · 22 metas · 16 keyword tweaks · gallery-500 root cause identified	22	3
W2 Wed 20 – Mon 25 May	Bug fix & content cleanup	Gallery-500 fix deployed · duplicate content rewrites · orphan pages resolved · sitemap reconciled · GSC resubmitted	14	2
W3 Tue 26 May – Wed 3 Jun	Performance sprint	CDN & page cache · image pipeline (WebP) · render-blocking deferred · fonts optimised · plugin audit · 9 URLs hit target PageSpeed	28	1
W4 Thu 4 – Fri 5 Jun	Verify & close	DS verification crawl · all rows closed in tracker · monitoring report · handover	6	1

2. Working assumptions

Senior engineer commits 4 hours per day, Mon–Fri (≈ 20 hrs/week).
Earliest sprint start: Mon 11 May 2026; target completion Fri 5 Jun 2026.
Client returns each approval batch within 48 hours of receipt.
Access provisioned by end of Day 1 (Mon 11 May): WP admin, cPanel/SSH, Google Search Console, GA4 (read), staging environment.
Site is on WordPress with shell/SFTP access available (matches what’s implied in the audit).
AI-assisted workflow (Claude Code / Copilot for bulk transforms, codegen, diagnostics).
Digital Spotlight remains responsible for verification (Screaming Frog re-crawl) at the end of Sprint W4.
No new feature work in scope — remediation only, against the existing tracker.
No public holidays fall inside the window (verify against AU/VIC calendar before kick-off).

3. Detailed weekly batches

Kick-off & access provisioning

Mon 11 May 2026 · 1 day · 4 hrs

4 dev hrs

3 client hrs

Goal: Every approval captured, every access granted, every diagnostic ready to run on Monday morning of Week 1.

#	Task	Owner	Hrs
0.1	60–90 min decision workshop — walk Q1–Q12 from the decision sheet, capture answers in writing	Dev + Client	1.5
0.2	Access provisioning: WP admin user, cPanel/SSH credentials, GSC + GA4 invites, staging URL	Client	1.0
0.3	Clone production to staging; enable `WP_DEBUG_LOG`; confirm we can reproduce the 500 there AI-assist	Dev	1.5
0.4	Run baseline Lighthouse on all 9 PageSpeed URLs; archive scores for before/after comparison AI-assist	Dev	1.0
0.5	Set up release branches / change log / shared status doc	Dev	0.5

Acceptance criteria.

Decision sheet returned with all 12 answers logged.
Dev can SSH into hosting and log into WP as admin.
Staging URL responds and matches production schema.
Baseline Lighthouse JSON saved for 9 URLs.

Dependencies. All work in Weeks 1–4 hinges on this. If access slips by >2 days, every subsequent week shifts.

SEO foundations & quick wins

Tue 12 May – Tue 19 May 2026 · 5.5 working days · 22 hrs

22 dev hrs

3 client hrs

Goal: Clear ~145 of the ~360 open items (40% of the backlog) and identify the gallery-500 root cause — all in five working days.

Releases this sprint

Release 1.A — Thu 14 May: redirects + internal-link fixes
Release 1.B — Tue 19 May: titles + metas + keyword copy
Diagnostic finding — Tue 19 May: documented gallery-500 root cause + fix patch ready for Sprint W2

#	Task	Owner	Hrs
1.1	Generate the 102 × 404→301 redirect map from the Crawl Error tab; produce both `.htaccess` rules and a Redirection-plugin CSV import AI-assist	Dev	2.0
1.2	Apply redirects on staging; smoke-test 20 representative URLs; deploy to production (Release 1.A)	Dev	2.0
1.3	Fix the 3 redirect chains/loops to single-hop 301s	Dev	1.0
1.4	Update 29 incorrect internal links via WP search-and-replace (WP-CLI `wp search-replace`) AI-assist	Dev	1.5
1.5	Apply 17 title tags via Yoast/Rank Math bulk editor; verify on staging AI-assist	Dev	1.5
1.6	Apply 22 meta descriptions via Yoast/Rank Math bulk editor AI-assist	Dev	1.5
1.7	Apply 16 “Keywords (Old)” copy edits (titles, H1s, body text) for centre pages	Dev	3.0
1.8	Resolve 6 orphan pages per Wave 1 decision (link in or 301 out)	Dev	1.5
1.9	Deploy Release 1.B to production; verify in browser + GSC URL inspection	Dev	1.0
1.10	Gallery-500 diagnostic. Reproduce on staging; capture stack trace from `debug.log`; bisect plugins via Health Check Troubleshooting Mode; isolate root cause; draft fix patch AI-assist	Dev	4.0
1.11	Mid-week status update + EOW status report to client	Dev	1.0
1.12	Buffer for unexpected approval clarifications	Dev	3.0

Acceptance criteria.

All 102 legacy 404 URLs return HTTP 301 to their approved targets (verified by re-crawl on a sample of 30).
17 titles + 22 metas live on production; visible in view-source.
16 centre-page copy edits live; H1 / title / body match the approved drafts.
29 internal links no longer point to http:// or 404 targets.
Gallery-500 root cause identified with a written diagnostic note + a tested patch on staging.

Client time needed (~3 hrs): 90 min workshop on Mon 11 May (W0), 60 min review of staging on Thu 14 May, 30 min sign-off on Tue 19 May.

Risk. If the gallery-500 diagnostic uncovers a deeply embedded plugin conflict (e.g. paid plugin no longer maintained), the fix may need a plugin replacement — could spill into Sprint W2 by 2–4 hrs. Mitigation: 2 hrs of contingency budgeted in Sprint W2.

Bug fix & content cleanup

Wed 20 May – Mon 25 May 2026 · 3.5 working days · 14 hrs

14 dev hrs

2 client hrs

Goal: Close out all remaining content / structural items so going into Week 3 the only open category is performance.

Releases this sprint

Release 2.A — Thu 21 May: gallery-500 fix live
Release 2.B — Mon 25 May: duplicate content + sitemap regenerated + GSC resubmitted

#	Task	Owner	Hrs
2.1	Apply gallery-500 patch on staging; confirm all 7 URLs return 200 + content; deploy to production (Release 2.A)	Dev	2.0
2.2	Apply 4 duplicate-content rewrites (gallery-category pages) using DS-drafted copy	Dev	2.0
2.3	Sitemap reconciliation: regenerate sitemap, verify the 112 listed URLs are removed, document closed rows AI-assist	Dev	2.0
2.4	Resubmit sitemap in GSC; request reindex for top 20 fixed URLs	Dev	1.0
2.5	Deploy Release 2.B; verify GSC index coverage report 24 hrs later	Dev	1.0
2.6	Wave 3 prep: server / hosting capability audit (PHP version, OPcache, HTTP/2, gzip/brotli, current asset weight) AI-assist	Dev	1.5
2.7	Wave 3 prep: image inventory across the 9 slow URLs; identify LCP candidates per page AI-assist	Dev	1.5
2.8	Wave 3 prep: render-blocking inventory (JS/CSS files, third-party domains, font files) AI-assist	Dev	1.0
2.9	EOW status report; confirm with client we’re green-lit to begin Wave 3	Dev	1.0
2.10	Contingency for Week 1 spillover (gallery-500 complications)	Dev	2.0

Acceptance criteria.

All 7 gallery URLs return HTTP 200 with rendered content (verified via curl -I + browser).
The 4 duplicate-content pages have unique copy ≥200 words, no longer flagged as near-duplicates by Screaming Frog.
Sitemap contains only indexable URLs; GSC accepts the resubmission with zero errors.
Wave 3 prep document delivered with prioritised performance work plan + estimated impact per URL.

Client time needed (~2 hrs): 30 min Thu 21 May sign-off on gallery fix, 60 min Fri 22 May review of duplicate content, 30 min Mon 25 May green-light to begin Sprint W3.

Risk. Sitemap-removal verification may surface URLs DS marked “to remove” that are actually legitimate. Mitigation: any URL that resolves with content is excluded from removal pending confirmation.

Performance sprint

Tue 26 May – Wed 3 Jun 2026 · 7 working days · 28 hrs

28 dev hrs

1 client hr

Goal: Move all 9 PageSpeed URLs from 45–73 mobile to ≥85 mobile. Single biggest week of value, sequenced biggest-win-first.

Releases this sprint

Release 3.A — Wed 27 May: CDN + full-page cache live
Release 3.B — Fri 29 May: image pipeline migrated to WebP + lazy-loading
Release 3.C — Tue 2 Jun: render-blocking JS/CSS deferred + fonts optimised
Release 3.D — Wed 3 Jun: plugin/script cleanup + locked-in PageSpeed scores

#	Task	Owner	Hrs
3.1	Install + configure WP Rocket (or LiteSpeed Cache); enable page cache, browser cache, GZIP/Brotli; preload key URLs	Dev	2.5
3.2	Front the site with Cloudflare; configure DNS, SSL, edge cache rules, Auto Minify, HTTP/3	Dev	2.5
3.3	Smoke-test caching: TTFB benchmarking before/after; deploy Release 3.A	Dev	1.0
3.4	Convert hero/gallery imagery to WebP (and AVIF where supported); regenerate WP thumbnails AI-assist	Dev	3.0
3.5	Enforce `width`/`height` on all images; native lazy-loading below the fold; `preload` on the LCP image of each centre page	Dev	3.0
3.6	Generate responsive `srcset` for hero images; verify on staging across viewports; deploy Release 3.B	Dev	2.0
3.7	Critical CSS extraction + inline; defer non-critical CSS	Dev	2.5
3.8	Defer non-essential JS; remove jQuery dependencies if theme allows; remove unused CSS via PurgeCSS	Dev	2.5
3.9	Font loading: `font-display: swap`, self-host or preconnect Google Fonts, subset to Latin AI-assist	Dev	1.5
3.10	Deploy Release 3.C; spot-check Lighthouse on 3 worst URLs	Dev	1.0
3.11	Plugin audit: deactivate unused plugins; replace any heavy plugins with leaner alternatives where safe	Dev	2.0
3.12	Third-party script audit: defer analytics, lazy-load chat widgets, conditionally load social embeds	Dev	2.0
3.13	Lock-in pass: re-run Lighthouse on all 9 URLs; iterate on any URL still <85 AI-assist	Dev	2.5
3.14	Deploy Release 3.D; produce before/after PageSpeed comparison report AI-assist	Dev	1.5
3.15	Mid-week + EOW status updates	Dev	0.5

Acceptance criteria.

All 9 audited URLs score ≥85 mobile on PageSpeed Insights (lab data).
LCP < 2.5s on all 9 URLs.
CLS < 0.1 on all 9 URLs.
TTFB ≤ 600ms (mobile) globally with Cloudflare cache hit.
No visual or functional regression on production (verified by 30-min UAT walkthrough).

Client time needed (~1 hr): 60 min UAT walkthrough on Wed 3 Jun afternoon after Release 3.D.

Risks.

Cloudflare DNS migration — brief propagation window. Mitigation: schedule the cutover for low-traffic hour and keep TTL low for 48 hrs prior.
Cache vs. logged-in admin behaviour — Mitigation: bypass cache for logged-in cookies; test admin flows on staging first.
WebP browser support on legacy iOS — Mitigation: use <picture> with JPEG fallback.
One URL stubbornly below 85 — Mitigation: 2.5 hr buffer in 3.13 specifically for iteration.

Verify, monitor & close

Thu 4 – Fri 5 Jun 2026 · 1.5 working days · 6 hrs

6 dev hrs

1 client hr

Goal: External verification by Digital Spotlight, tracker zeroed out, monitoring in place, project handed back to BAU.

#	Task	Owner	Hrs
4.1	Coordinate with Digital Spotlight for Screaming Frog re-crawl; provide redirect map + change log	Dev	1.0
4.2	Triage any leftover items DS flags from the verification crawl (small — budgeted 2 hrs)	Dev	2.0
4.3	Set up monitoring: PageSpeed regression check (weekly automated), GSC alert subscription, uptime monitor on the 9 URLs AI-assist	Dev	1.0
4.4	Final report: before/after PageSpeed scores, list of every closed row, list of any deferred items, monitoring dashboard URL AI-assist	Dev	1.5
4.5	30-min handover call with client & DS	Dev + Client + DS	0.5

Acceptance criteria.

Digital Spotlight tracker shows every previously-open row either Closed or formally deferred with rationale.
Final report delivered (PDF + tracked-changes spreadsheet).
Monitoring is live and alerts the team if mobile PageSpeed drops below 80 on any of the 9 URLs.

4. Estimate & resourcing

Phase	Dates	Dev hrs	Notes
Sprint W0 — Kick-off & access	Mon 11 May	4	Workshop, access, baseline benchmarks
Sprint W1 — SEO foundations	Tue 12 – Tue 19 May	22	Includes 3 hr buffer + diagnostic
Sprint W2 — Bug fix & content	Wed 20 – Mon 25 May	14	Includes 2 hr W1 spillover contingency
Sprint W3 — Performance	Tue 26 May – Wed 3 Jun	28	Largest sprint; 4 production releases
Sprint W4 — Verify & close	Thu 4 – Fri 5 Jun	6	External verification + monitoring + handover
Total	11 May → 5 Jun (4 cal. weeks · 18.5 working days @ 4 hrs/day)	74	Single senior engineer, AI-assisted

How AI assistance was factored in. The AI-assist tag marks tasks where Claude Code / Copilot meaningfully compresses the work — bulk redirect-map generation, meta tag application, sitemap reconciliation scripts, image inventory, font subsetting, monitoring scripts, and the final report. Conventional rates apply elsewhere (CDN setup, plugin audits, image pipeline, diagnostics that need real-system inspection). For comparison, the same scope without AI assistance would estimate at roughly ~110 hrs.

5. Top risks & mitigations

#	Risk	Likelihood	Mitigation
R1	Approval slippage by client (any week)	Medium	Decision sheet pre-loaded at Week 0; weekly 30-min cadence; written 48-hour SLA agreed up-front
R2	Gallery-500 fix needs plugin replacement (paid/abandoned plugin)	Low–Medium	4 hr contingency budgeted in Week 2; alternative plugin shortlist drafted in Week 1
R3	One PageSpeed URL stuck below 85 after Wave 3	Medium	2.5 hr iteration buffer in task 3.13; can extend into Week 4 contingency if needed
R4	Cloudflare DNS misconfiguration causes brief outage	Low	Lower TTL 48 hrs in advance; cutover during low-traffic window; rollback plan documented
R5	Cache invalidates editor / form behaviour	Low	Bypass-on-cookie rules tested on staging first; admin smoke test before each release
R6	Hosting can’t support modern stack (PHP < 8.0, no SSH)	Low	Capability audit in W2 task 2.6; if upgrade needed, raise immediately and re-baseline

6. Communication cadence

Daily — written EOD update in shared status doc (3 lines: shipped / in progress / blockers).
Twice-weekly — release notifications when Releases A and B drop, with verification URLs.
Weekly — 30-min Friday review with the client lead; written status report immediately after.
Ad-hoc — same-day response on anything blocking; escalate to phone after 4 hrs.

7. To start Sprint W0 on Mon 11 May 2026, we need by EOD Fri 8 May

Decision sheet (Q1–Q12) returned signed off.
WP admin login (Editor + Redirects + Yoast permission).
Hosting access (cPanel and/or SSH).
Google Search Console + GA4 (read-only) invite to our team email.
Confirmation that staging exists, or permission to clone production to staging.
Single named approver and confirmation of 48-hour approval SLA.

Anything not in by EOD Fri 8 May pushes the sprint start by the same number of working days — the 5-Jun completion date moves with it.