Shipping 12 SEO pages in 14 days (and what I cut)

Ezra needed inbound traffic. Paid ads on affiliate-tooling keywords run $5 to $50 per click. SEO is the only channel that scales for B2B tools without burning a marketing budget I do not have.

The plan was 14 days from "no SEO content" to "12 new pages submitted to Google Search Console." I hit the date. Most of the pages are now indexed and starting to rank on long-tail.

This is the workflow.

The 14-day plan

• Days 1-2: Keyword research. Page-type decisions. Slug list.
• Days 3-5: Shared components for the page types.
• Days 6-12: Content. Roughly one page per day.
• Days 13-14: GSC submission, sitemap, final QA.

I cut a lot to hit the date. More on that below.

Page types and counts

• 4 versus pages: "Ezra vs [Competitor]" for the four named competitors
• 3 guides: "How to set up an affiliate program," "How to vet affiliates," "How to handle payouts"
• 5 blog posts: industry insights and how-to content for top-funnel keywords

Total: 12 new pages. With existing marketing-site pages, the sitemap reached 38 URLs by day 14.

Each page type targets a different intent. Versus pages capture comparison-shopping keywords (high intent, lower volume). Guides capture informational keywords (lower intent, higher volume). Blog posts capture broader interest.

The template-first approach

Versus pages share a structure: hero, feature comparison table, pricing comparison, "Why Ezra" section, CTA. So do alternatives pages and guides.

I built two shared React components:

• ComparisonPage renders a versus or alternatives layout from a JSON config
• GuidePage renders a long-form guide with H2 sidebar nav and CTAs

Each page is a JSON config (titles, comparison fields, content blocks) plus prose. Adding the fourth versus page took 90 minutes. The first took two days because I was still building the component.

This compounds. If I write three more versus pages next month, each is a single-afternoon job.

MDX over a CMS

I used MDX. Each page is a .mdx file in the repo. Vercel rebuilds on push. No separate CMS to maintain, no headless API to call, no editor account to manage.

The tradeoff: the writer (me) edits Markdown directly. If a non-technical person needs to edit pages without touching code, MDX is the wrong choice. For solo work, MDX is faster and freer than any CMS I have tried.

What I cut to ship fast

The list of things I deliberately did NOT do in the 14-day window:

• Per-page hero imagery. Used a single repeating header pattern across all pages. Came back later and customized the top 3 pages once they showed traffic.
• Custom Open Graph images. Used the default site OG. Added per-page OG generation in week 4 (took 4 hours total).
• Schema.org markup. Added in week 3 once pages were ranking enough to matter for rich results.
• Internal linking matrix. Did it manually with a script that crawled the sitemap and suggested links. Hand-curated the output. Three hours total.
• Author bylines and structured author data. Skipped entirely. Adding later if an author program makes sense.
• A/B testing infrastructure. One variant only. Will A/B once traffic warrants it.

None of these were free skips. All of them were deferred deliberately. The first job was crawl status. The second job was rich results.

What I shipped instead

The minimum I needed for solid initial indexing:

• Clean title tags and meta descriptions per page
• Clean URL structure: /vs/competitor-name, /guides/how-to-X, /blog/post-name
• Internal linking by topic cluster, hand-curated
• Sitemap auto-generated from filesystem
• Robots.txt with explicit allow rules
• GSC verified, sitemap submitted, manual URL submission for the most important pages

This is the 80 percent that drives initial indexing. The other 20 percent (schema, OG images, author data) adds quality but does not affect crawl status.

The numbers

By day 14:

• 38 URLs in sitemap
• GSC sitemap submitted
• All 12 new pages submitted for indexing

By week 4:

• Most pages "Discovered" in GSC
• A handful of pages ranking on long-tail terms
• First inbound organic-search trial signups arriving

This is not viral SEO. It is the slow compounding baseline. The point is that getting from zero to "indexed and starting to rank" in a month is achievable solo, with no marketing team and no paid promotion.

What I would do differently

• Schema markup on day 1. Adding it later is fine. Adding it on day 1 is cheaper because the templates are still being built.
• AI-generated draft alt text. I wrote alt text manually. Claude would generate first drafts and save 30 minutes per page.
• Internal linking as code. The hand-curated approach worked at 12 pages. It will not scale to 50.
• Author program from day 1. I avoided this because it felt like overhead. Looking back, an author page lends authority that is hard to retrofit.

The takeaway

For B2B tools, 12 quality SEO pages in the first month is a low-effort moat. The work is templating and content. Most founders skip this because it does not feel like engineering.

If you have a B2B tool with no organic traffic, the highest-leverage thing you can do in the next two weeks is template, write, and ship. Not marketing campaigns. Not paid ads. Just pages, fast, with the discipline to ship the 80 percent and defer the rest.