A lot of content agencies will tell you they do "research-backed content." What they mean is they ran a keyword through a tool and looked at the top 10 results before writing. That's not research. That's summarising existing content with fresh formatting. It doesn't build topical authority, it doesn't map to your buyer, and it definitely doesn't get cited in AI answers. Every CTW engagement starts differently - with actual research into your ICP, your competitive landscape, and the questions your buyers are asking before they've even found you.
Why everything starts with research
The most common failure mode in B2B content is writing for a hypothetical buyer instead of an actual one. You end up with articles that cover broadly relevant topics but don't answer the specific questions your ICP is asking, don't address the objections they carry, and don't position your solution the way your best sales conversations do.
The fix isn't better writing - it's better research before writing. Every CTW engagement begins with three research layers:
ICP mapping. Who is the buyer? Not "businesses" or "marketing teams" - specific job titles, company types, revenue bands, and the stage of awareness they're likely at when they find you. What are they searching for at each stage? What do they already believe about your category? What makes them sceptical? This determines which topics to write, how to frame them, and what the CTA should be at the end of each piece.
Competitive gap analysis. What are your direct competitors ranking for? Where are they getting cited in AI-generated answers? What topics have they covered well, and where have they left gaps? The goal isn't to mimic their content - it's to identify the angles they've missed and own them before they notice. This also surfaces the positioning moves you can make that they can't, based on what's distinct about your offering.
AI citation audit. For new clients, I run their domain and their competitors through the key AI tools - ChatGPT, Perplexity, Claude, Google AI Overviews - to see who's getting cited and who isn't when buyers search for your category. This tells me which content types and schema patterns are working in your space, and where the structural fixes need to happen before producing more content would even be worthwhile.
If the research says your buyers are primarily technical leads evaluating tools, I'm writing differently than if they're marketing managers trying to justify a budget. Same topic, different angle, different structure, different outcome. Research is the only way to know which version to write.
The technical SEO audit every retainer starts with
Before I write a single article for a new client, I audit their current site. Not because it's a nice deliverable to hand over - because there's no point producing great content if the platform it lives on is working against it.
The kickoff audit covers:
- Page structure - H1/H2/H3 hierarchy, meta title and description quality, heading keyword targeting
- Schema coverage - what's implemented, what's broken, what's missing entirely
- Internal linking - is authority flowing through the site correctly, or are key pages isolated?
- Crawlability - robots.txt, sitemap, canonical tags, redirect chains, duplicate content
- Core Web Vitals and page speed - particularly LCP and CLS, which affect both rankings and user experience
- llms.txt presence - does the site have an AI context file? Is it well-structured?
- Competitor gap summary - top 3 competitors, what they rank for that you don't
You get a prioritised fix list. Not a 60-page report with a thousand findings - a clear stack-ranked list of what to fix and in what order. High-impact technical fixes are often things your developer can implement in a day. They don't need a consultant to do them; they need to know about them.
Implementing the fixes is your call. I can quote separately if you want help with it, but the audit deliverable stands regardless.
What "an article" actually includes
When I say long-form article, I mean a piece that's built to rank, built to be cited, and built to move a specific buyer along their journey. Here's what goes into each one:
Keyword and intent research. Primary keyword, secondary keywords, related questions, and a clear read on where the buyer is in their journey when they'd be searching for this. Awareness, consideration, or decision - the answer changes the structure, length, and angle of the article.
Competitor content review. What's already ranking for this keyword? What are those articles doing well? Where are the gaps? Every article is written to be the most useful, most complete, and most AI-citable version of that topic - not just another take on the same content.
The article itself. Written in your brand voice, for your ICP, at the right depth for the topic. Long-form typically means 1,200–2,500 words, but length is always determined by what the topic actually needs - not a word count target. A thorough 900-word piece beats a padded 2,000-word one.
Schema markup, included as standard. Every article gets Article schema (headline, author, date published, date modified), FAQPage schema if there's a FAQ section, and BreadcrumbList schema. This is not optional and it's not an add-on. Content without schema is less visible to AI systems - so shipping content without it would undermine the whole point of the engagement.
Meta title, meta description, and suggested slug. Written to match search intent, within character limits, with the primary keyword placed correctly. Ready to copy-paste into your CMS.
Internal link suggestions. Two or three existing pages on your site that should link to or from this article, with suggested anchor text. Internal linking is one of the most underused levers in B2B SEO - it costs nothing and has a real effect on authority distribution.
The three retainer tiers - what's behind each one
01 / Entry - $1,400/month
The right tier for teams who want to start building compounding search and AI presence without a large initial commitment. Two long-form articles a month is enough to build topical authority steadily - it just takes longer than four or six.
What's included: the kickoff technical SEO audit, two long-form articles per month (each with full schema and meta), a monthly keyword and citation check-in, and articles delivered as files for you to publish.
What this tier is best for: B2B businesses who have a clear ICP and a solid product, know content is a long game, and want to start building search presence without overcommitting. Also a good starting point for teams who want to test the quality and process before scaling.
02 / Content & Enablement - $2,500/month
The most popular tier, because it covers both sides of the revenue equation: content bringing qualified traffic in, and sales assets helping close it.
What's included: everything in Entry, plus two additional articles (four total), and one custom sales one-pager per month built from the same research that drives the content calendar.
The one-pager is built around what prospects are actually asking at that point in time - product comparison questions, objections that are coming up in sales calls, pricing justification, competitive differentiation. It's produced from a monthly debrief so it reflects live sales conversations, not hypothetical buyer concerns.
What this tier is best for: B2B teams with an active sales motion who want marketing and sales to be pulling from the same research. The one-pager is often the piece that gets forwarded in a deal, shared in Slack, or used in a follow-up email. When it's built from real buyer questions rather than guesswork, it performs significantly better.
03 / Growth - $4,000/month
For teams treating content as a genuine growth channel, not a checkbox. More output, more distribution, and a proper strategy loop each month.
What's included: six long-form articles per month, one custom sales one-pager, four to six LinkedIn posts repurposed from the month's articles, a monthly strategy call and performance review, and priority turnaround.
The LinkedIn posts deserve a note. They're not just copy-pasted article intros - they're rewritten for LinkedIn's format and feed behaviour: shorter, more opinionated, optimised for reach. Each one links back to the full article and is written to drive traffic rather than summarise the article for people who won't click.
The monthly strategy call is where we review what's working, what's moving in search and AI citations, and whether the content focus should shift. Markets change, buyer priorities shift, competitors make moves. A static content calendar produces static results.
The CMS publishing add-on
For any tier, you can add CMS publishing and Search Console management for $500/month. This covers:
- Uploading the article to your CMS with correct formatting
- Implementing the schema directly in the page head (not just handing over the code)
- Setting the meta title, meta description, and canonical URL
- Handling internal links from existing pages to the new article
- Submitting each new URL to Google Search Console for indexing
- Monitoring Search Console for crawl errors and indexing issues on published content
Some clients prefer to publish themselves - full control, faster turnaround, no extra cost. Others want to hand it off entirely so nothing falls through the cracks between delivery and going live. Both work. The add-on exists for the latter.
Project work - for teams who need one thing done well
Not every gap fits a retainer. Some teams need a single technical SEO audit. Some need an AI setup - a custom Claude Project or GPT built around their brand, with a prompt library and guardrails so their team's AI output actually sounds like them. Some need a sales deck or interactive ROI calculator.
Project work is quoted per scope. Fixed from-prices so you know the floor before committing: AI setup and brand context from $1,500, interactive tools from $2,000, sales deck or one-pager from $750, standalone technical audit from $900.
Project work is also how some clients test the waters before moving to a retainer. If you want to start with an audit and see what it surfaces before committing to monthly content, that's a completely legitimate path.
The packages are priced the way they are because the research takes time. Cutting it out would let me produce more content for less money - but it would be the kind of content that doesn't compound, doesn't get cited, and doesn't move the needle. If the goal is just words on a page, there are cheaper options. If the goal is search and AI visibility that builds over time, the research is what makes the difference.