Salesix AI: building topical authority with a 43-asset content structure

Overview

Salesix AI sells AI voice agent software. I was asked to build a strategy to increase its topical authority by creating and ranking content around e-commerce use cases. The site had a small organic presence: 142 clicks and 7.43K impressions over the previous three months. Over 12 weeks, I planned, structured and created 43 SEO content assets. In Search Console, the last three months against the previous three show clicks up from 142 to 216 (+52.1%) and impressions up from 7.43K to 11.2K (+50.7%).

The challenge

• A competitive space, with multiple businesses targeting the same keywords and use cases
• A few use-case pages already existed, but they weren't focused on e-commerce, weren't well written or structured, and weren't optimised
• The content needed to cover many different search intents
• A real risk of publishing scattered content that wouldn't build authority

The goal

• Establish topical authority in the AI voice agent for e-commerce space
• Improve organic visibility for e-commerce-related searches
• Rank content across multiple types of search intent
• Make the content easier for users and search engines to discover

My strategy

1. Content structure. I planned 1 pillar page, 6 use-case pages, 6 playbooks and 30 supporting blogs, starting broad with the pillar and narrowing to specific blogs, so the structure covers the whole topic without gaps.
2. On-page optimisation. Every page needed optimised titles, meta descriptions and keyword targeting so it could rank, not just exist.
3. Internal linking. Every page links to related pages: blogs to their use case, use cases to the pillar and the relevant playbook, so the structure works both ways, not just pillar-down.
4. Link building. Backlinks were concentrated on the pillar page and other key pages, so authority would lift the whole cluster instead of being spread thin across 43 pages.

Content architecture

The pillar covers the whole topic. Use-case pages cover the high-impact areas. Playbooks give action-oriented guidance for each. Blogs capture long-tail and comparison searches.

What I did

• Researched keywords for AI voice agents in e-commerce
• Built a pillar-cluster SEO framework
• Planned, structured and created 43 SEO content assets
• Optimised on-page SEO elements
• Wrote FAQ sections and definition-style content blocks across pillar, use-case and blog pages to support AI search and answer-engine visibility
• Improved the internal linking structure
• Built high-authority backlinks
• Used Google Search Console to identify underperforming pages, then updated their titles, meta tags and content to recover lost visibility

How I produced it at scale

To produce 43 assets consistently, I built a five-step AI-assisted workflow:

1. Research: researched use cases, mapped real problems, ran keyword research, matched search intent and built blog outlines
2. Writing: turned approved outlines into full blog drafts, optimised for readability, search and AI answer engines
3. Media planning: reviewed finished blogs and generated featured-image prompts and supporting media ideas
4. Design: created the visuals from those prompts
5. Link strategy: analysed each page's topic and cluster role and built targeted backlink campaign angles

Results

Source: Google Search Console, last 3 months compared with the previous 3 months.

• Delivered all 43 planned assets (1 pillar page, 6 use-case pages, 6 playbooks, 30 blog posts) on schedule within 12 weeks.
• The average position moved from 19 to 33.6. I read that as a side effect of adding many new pages, which typically rank lower before climbing over time.
• The starting base was small, so the percentages rest on a modest number of clicks (+74 in total). I treat them as an early signal, not a benchmark.

What I learned

• Structure before volume. Planning the whole topic first is what let 43 pieces work as one system instead of scattered posts.
• One strategy, not separate tasks. Keyword research, content planning, production and internal linking only worked because they were connected.
• Read position with context. Adding many new pages pulls the average position down at first, so I read it alongside clicks and impressions.

How I'd apply this approach to a Shopify store

This is method, not a result from this project.

• Give each topic one clear page, such as a collection page, with supporting products and guides linking to it and back.
• Match each page to one search intent before optimising titles, meta descriptions and headings.
• Link supporting content to the collection and product pages it should help rank.
• Concentrate authority-building effort on the pages that matter most commercially.

This was an internship project on a company's site, not an online store. Organic performance depends on many factors, so these figures show what changed during the project, not a guaranteed outcome.

Visuals

Project reports