The full playbook: 400 to 4,800 monthly organic visits in 12 months.

SEO that works in B2B services is rarely about tricks. It's about being the most useful source in your niche — and making sure search engines can read you cleanly. Below is the exact 12-month sequence we ran with Northspark, in the order we ran it.

The headline numbers came from the system, not from any single move. This article is about the system — what we shipped, when, and why each piece was sequenced where it was. If you skim it you'll think "this is just SEO 101 done in order." That's exactly the point. The compounding magic of SEO is in the sequencing, not the cleverness.

Month 0 — The audit nobody wants to do

Before we wrote a single line of code or content, we ran a full audit. Most SEO retainers skip this and start producing immediately because production feels like progress. It isn't. Producing content into a broken site is one of the most expensive mistakes in marketing.

Our Month 0 audit answers three questions:

• What's already working? Pages with even moderate traffic — what are they ranking for, and are they aligned with revenue intent?
• What's broken that's invisible? Indexing issues, canonical misfires, orphaned pages, dead redirects. These are the silent killers.
• What's the realistic ceiling? Given the domain's authority, the competition's authority, and the niche size — what's the honest 12-month traffic potential?

For Northspark, the audit surfaced 140 fixable technical issues, 12 pages that had real ranking potential, and a realistic 12-month ceiling of 4,000-6,000 monthly organic visits. We ended at 4,800 — right in the middle of the forecast.

Months 1–2 — Architecture rebuild

Northspark's old site architecture was a flat list of service pages — every page roughly equal weight, no clear hub-and-spoke structure. To Google, this signals "no topical authority on anything in particular." We rebuilt around topical authority.

The new structure

• Pillar pages — one authoritative ~3,000-word page for each major service area. These pages target the broadest keyword in the topic.
• Cluster pages — 4-8 focused pages per pillar, each targeting a specific sub-question (e.g. "how to choose," "comparison vs. alternatives," "cost," "implementation timeline").
• Internal linking — cluster pages link up to the pillar; the pillar links down to every cluster. Every page on the site signals what it's the canonical answer for.

This is well-known SEO theory. It works because it gives Google a clean signal about topical authority. The execution challenge is that it requires real planning — you can't bolt cluster structure onto an existing site without rewriting most of the content.

Months 2–4 — Technical SEO triage

We worked through the 140-issue audit list in priority order. Three buckets:

Bucket 1: Indexing blockers (week 1)

Northspark's robots.txt was quietly blocking their blog. A 2022 migration had left 12 canonical tags pointing at the staging domain. Both fixed in week one. These are zero-cost, high-impact fixes that no one notices because they don't add anything — they just stop subtracting.

Bucket 2: Core Web Vitals (weeks 2-6)

Mobile LCP was 4.1 seconds — failing. We image-optimised, lazy-loaded below-fold media, removed three render-blocking scripts, and moved fonts to font-display: swap. New mobile LCP: 1.9 seconds. Pass.

Bucket 3: Schema and metadata (weeks 6-8)

Every page got proper title tag and meta description hygiene. Service pages got Service schema. The leadership page got Person schema. The FAQ page got FAQPage schema. This isn't a ranking factor directly — but it dramatically improves CTR from search results, which compounds into rankings.

Technical SEO doesn't add traffic. It removes the things suppressing your traffic. That's why it has to happen first.

Months 3–9 — The 48-piece content engine

This is the part everyone wants to skip to. Don't. Content production without the architecture and technical work above produces 48 pieces that rank for nothing.

The content brief

Every piece started with a structured brief:

• Target keyword (volume, difficulty, commercial intent)
• Search intent (informational, navigational, transactional)
• Top 5 SERP competitors and their gaps
• Internal links it should make (up to the pillar, sideways to relevant clusters)
• Schema type
• Target word count (set by SERP analysis, not by feel)

Production cadence

Four pieces per week, every week, for 27 weeks. Same writer team. Same editor. The consistency mattered more than the volume. A team writing 4/week for 27 weeks beats a team writing 8/week for 3 weeks and 0/week for the rest.

What didn't work

We tried AI-generated drafts in months 4-5. The output was technically fine — but every Google update through 2025 has gotten better at down-weighting unedited AI content, and ours fell with it. We pivoted to AI-assisted (model writes a skeleton, human writes the meat) which kept the velocity without the quality penalty.

Months 6–12 — Editorial backlinks

We started link-building in month 6, not month 1. Reason: building links to a site that isn't ready to retain authority is wasted spend. Once the architecture and content engine were in place, link-building had something to amplify.

The 32 backlinks we earned (not bought)

No directory listings, no link farms, no guest posts on irrelevant sites. The pitch was always: "Here's an angle for a story you might want to write, drawn from our actual client work." Hit rate: 16% pitch-to-placement.

The published links came from B2B trade publications, two podcast appearances (which produced show-notes backlinks), and three pieces of data we shared exclusively with industry journalists. The data pieces produced 11 of the 32 links by themselves.

What we'd do differently next time

1. Start link-building outreach in parallel with content, not after. The lead time on a placed backlink is 4-8 weeks. We could have shaved 2 months off if we'd started outreach in month 3 instead of month 6.
2. Do programmatic SEO on long-tail comparison queries. We left these on the table. Northspark is adding them in year 2.
3. Schedule a quarterly content refresh from day one. The pieces published in month 3 needed updating by month 9 — we did this manually and reactively. Building it into the cadence would have caught the wins faster.

Full case study with the methodology and results: Northspark — 400 to 4,800 visits in 12 months →