The work, and the
numbers behind it.

Capital project assurance. It means checking that a large capital project is ready to start, and staying on it once the work begins. They sold software for it. I joined as the only marketer. The product worked. People were buying it. There was no marketing function at all, and the product team had written the website. Every word was accurate. None of it sold anything.

Global consultancies owned the category term. Some were a thousand times the size. They had written the language of the discipline, and everyone in it spoke their words. All of them sold advice. None of them sold software. The buyers most likely to purchase never came across us.

Research. I mapped the rankings in SEMrush and Ubersuggest. Moz for the questions people actually type. Then I read the competing pages myself, all of them. The gap was structural. The consultancies had documented the discipline thoroughly. Not one of them told a buyer how to choose software for it.

• A dedicated category page, written to answer the buyer's question directly
• Four buyer personas defined first, covering end users and the champions who bring the budget conversation internally
• Homepage messaging restructured so the category term occupied the primary position
• The product page rewritten around the category and the personas accountable for the outcome
• An FAQ set and topic clusters, each answer self-contained with its evidence, so a single passage stays accurate when quoted out of context
• Off-site presence: listings on G2 and Software Advice with full feature documentation and video, and supporting content using the same category language throughout

Why the directories matter now. Large language models read them. Ask an assistant what to buy and part of the answer comes from G2. So the listings are visibility work. Treat them that way.

Verified while authenticated and in a private browsing session, and monitored on a recurring basis afterwards.

The company belonged to the industry institute its buyers belong to. The category runs on webinars. A vendor hosts one. The vendor emails its own list. Its own customers turn up. Everyone calls that reach.

Our list held 1,200 to 1,500 records. Too small to justify the production. And to anyone outside it, a vendor session reads as an advert.

So I co-branded it with the institute. Membership made that possible. The session carried their standing alongside ours, and they distributed it as an industry media partner. The subject served their members first. Why does capital expenditure predictability keep failing at organisations that already have governance frameworks, stage gates and project controls in place?

• I produced every promotional asset, for the organisation's channels and for the institute's
• The institute distributed it to its own list, in its upcoming-event emails and newsletters, and in its member forums
• Between 60 and 70% of execution stayed in house, across organic social, email, the landing page and the follow-up
• Registration ran through a form feeding a workflow I built, assessing every registrant automatically against region, company size, role and company revenue
• Qualifying registrants went to sales as marketing qualified leads. Everyone else entered a nurture sequence

From a single session. It was subsequently established as a quarterly programme.

The company exhibited at its industry's annual conference. What we brought was new. An AI layer applied to capital project assurance. The organisers describe their exhibition programme, in their own words, as a place to separate hype from reality. So an AI capability walks in already carrying the burden of proof. All of it.

I joined about two months before it opened. Two months. Learn a technically demanding product. Learn a capability with no precedent to borrow from. Then produce everything the company would stand behind on the floor.

• A dedicated landing page for attendees, reachable by QR code from the stand
• Two brochures, one for the platform and one for the new capability
• Booth backdrops, standees and branding elements, with the design concepts worked out to make an abstract capability visually legible
• Video production for the stand
• Email, website and social promotion before and during, including placement in the institute's member forums
• Logistics for the physical build
• A drip sequence I wrote, so a scan at the stand entered a nurture programme, off the spreadsheet

This was an educational event. No pipeline figures were recorded for it, so none are claimed here.

The platform connects to more than 1,400 systems natively, which makes it applicable to almost any operational team. That breadth is the marketing problem.

Four buyers with almost nothing in common. IT service management. Finance. Cybersecurity. Early-stage startups. One message for all four convinces none of them.

I ran four lead generation campaigns concurrently, one per segment.

• Ten thousand US dollars a month across all four, allocated by conviction
• Creative written separately for each segment
• Exclusion lists applied so the four audiences did not compete for the same impressions
• Its own lead generation form and landing page per segment, each built for retrieval by AI assistants as well as search
• Continuous monitoring and optimisation across all four in parallel

The structure admits a fifth segment without rebuilding the first four.

A HIPAA-compliant telemedicine platform, given away free. Appointment management, patient records, prescriptions, secure video consultation, practice payments, a virtual waiting room, and AI-assisted transcription of consultations. Buyers were physicians, clinicians and hospital administrators. It launched into Switzerland, the United States and India at once. No brand. No site. No presence anywhere.

The product was identical in all three markets. The reason to trust it was not. One message would have been credible in one country. The other two would have ignored it.

One voice held across all three. Clinical credibility without coldness. A physician who cannot trust the product medically stops at the first sentence. Underneath that voice, three separate arguments.

• United States. Compliance first. American buyers screen on HIPAA before they evaluate anything else, so the compliance position led every page and every asset
• Switzerland. Data residency. The Swiss question was where patient data physically sits and who can reach it, so privacy and data handling led
• India. Access. The pull was reach and affordability for practices that could not fund a platform, so the free model and the breadth of coverage led

Underneath those three:

• SEO architecture built per market, keyword research run separately for each geography, topic clusters split by buyer, physicians on one side and hospital administrators on the other
• All content written and published in house. Clinical accuracy needed somebody who could read the product documentation correctly, so no agency touched it
• Multilingual production for the Swiss market
• Listings secured on G2, TrustRadius, Software Advice, Trustpilot and Gartner-owned review sites, so a physician checking third-party opinion found the product where they already looked
• A CRO programme on the product pages: A/B tests on headline copy, CTA placement, form length and page layout

A United States managed services organisation and Microsoft Dynamics Gold Partner. The buyers were North American enterprise IT. I owned demand generation content across cloud, cybersecurity and managed IT.

Enterprise IT buyers in this category are marketed to constantly. They read very little of it. Another piece about digital transformation goes straight past the technical audience. And those are the people who approve the purchase.

• One deep asset, chosen over a content calendar. A guide on how DevOps resolves the five recurring failures in software product development, written for practitioners
• Written at the technical register the audience uses, so the reader recognises their own problem in it
• Two segmented newsletters, one on DevOps and one on Azure, each with its own subscriber base and its own editorial line
• Persona research and account-specific email underneath, the base form of account-based marketing

Newsletters: 1,500 subscribers at 33% open with a 15% organic traffic lift, and 1,200 at 25% open with a 9% lift.

A technology services group with several products and a limited budget for paid distribution. Owned content had to carry the traffic.

Generalists write technical subjects badly. The engineers who search for them read the result as filler. Depth is the whole job here. You cannot hand it to a writer who has not understood the material.

• Subjects chosen where search demand and genuine technical depth overlapped: Kubernetes security, chatbot architecture, blockchain architecture
• Written to be useful to a practitioner, at a length the subject actually required
• Guest placements pursued separately on established industry publications with domain authority of 55 and 57, for authority, with volume never the point
• Two long-form case studies built from the customer's problem and outcome, so sales could use them in a live conversation

Case studies drew 4,400+ and 3,700+ pageviews, converting at 9% and 3.5%.