Why this matters
Publishing more is not a strategy. In Zyppy's 2026 survey of 131 SEO professionals, original research and first-party data rated among the highest-impact content signals, while content published at scale with little added value rated negatively. The dividing line was not whether AI was involved, it was whether the page added anything the results already ranking did not.
That is the standard we write to. Every piece has to carry something only you can supply: your project data, your pricing reality, your pattern recognition from doing the work. If a topic has nothing like that behind it, we would rather kill it than fill it.
What we actually do
> Strategy
- ▶Topic selection based on real search demand and what currently ranks, plus an honest read on whether you can add anything to it
- ▶Clusters and internal linking planned before the first draft, so pages reinforce one topic instead of competing for it
- ▶Every piece assigned a job: rank for a query, answer a sales objection, or support a page that converts
> Production
- ▶Subject-matter interviews with your operators, so pages carry pricing context, project detail, and judgment that cannot be scraped
- ▶Sourced claims with links, no invented statistics, no borrowed numbers presented as ours
- ▶Structure built for direct-answer extraction: the finding early, section-level answers before methodology, FAQ blocks that map to schema
> Distribution and maintenance
- ▶Repurposing into social, email, and sales collateral, so one interview funds more than one asset
- ▶Refresh cadence on pages that already rank, since defending a top position is cheaper than earning a new one
- ▶Reporting on which pieces are compounding, which have plateaued, and which should be consolidated or retired
What we won't ship
No AI-generated volume with a keyword swapped between drafts, no near-duplicate location or comparison pages, no statistics we cannot source. Google's 2026 enforcement and the practitioner survey landed in the same place: scaled content without added value is a liability on the site that hosts it, not a neutral bet. Fewer, heavier pages hold up better and cost less to defend.
> Read the breakdown: what 131 SEO pros say actually works_Linta Roofing
Content and SEO work took organic search revenue from $395K to $1.39M while ranked keywords grew to 3,842.
> READ THE CASE STUDY_Our approach
Plan
Demand research, competitor page review, and a shortlist of topics where you have something real to add. Topics that fail that test get cut here, not in editing.
Interview
We pull the specifics out of your team: numbers, edge cases, the questions customers actually ask. That material is what makes the page hard to copy.
Publish
Written, edited, sourced, structured for answer extraction, schema applied, internal links placed on both ends.
Compound
Track what earns positions, refresh winners on a cycle, consolidate the pages splitting a topic between them.
Common questions
How many pieces per month?
Typically 2 to 8. The number matters less than the standard: we would rather ship four pages carrying original material than twelve that summarize what already ranks.
Do you use AI to write?
We use it for research support and drafting scaffolds, never as the writer of record. The survey data and Google's own enforcement both point at the same thing: process and added value decide the outcome, not whether a tool was involved.
Will it sound like us?
Yes. Voice comes out of the interviews and a style guide we build with you, then every draft is edited against it.
Can you work with our in-house writers?
Often the best setup. We build the strategy and sourced briefs, your team writes, we edit and handle the technical and schema side.
How long before content produces results?
New pages typically take a few months to settle into position, and clusters compound after that. Refreshing pages that already rank moves faster, which is why we usually start there.