How a Series A B2B SaaS scaled content production with AI 

The company

• Industry: B2B SaaS, Ecommerce (with multiple verticals)
• Stage: Series A, ~£10M ARR
• Marketing team: 1 (the marketing leader, with the full remit)
• Sales team: 4 outbound SDRs

---

The challenge

This company is so typical of who I work with. One marketer running it all — content, events, partners, campaigns, email. She was doing the work of three, and most of it was content.

She knew she needed to adopt AI. She’d dabbled with ChatGPT — got it to help her draft a few emails, sketch out a campaign plan. She could see the potential. But most of what she’d seen so far, the output was poor.

Here’s the thing: she’s a content marketer by trade. It’s her best skill and what she enjoys most. So when she’s using AI, she’s not just trying to make it work — she’s half trying to catch it out. Spotting every place the writing falls down. Where it gets the brand voice wrong. Where it says something her audience would never say.

In her words:

“I have to ship a lot of content — not just blog content and newsletters, but nurture flows that I know need to be more targeted. At the moment we just have a generic one. The SDR team has been growing too, and there’s pressure to supply them with content for outreach and building their profiles on LinkedIn for prospecting. It’s just not something I’ve had time to get to.”

“Every campaign started with me opening a blank doc. Some weeks I’d close my laptop on Friday feeling like I’d been busy all week and shipped nothing.”

“I knew AI could help. I just didn’t have time to figure out where to start.”

That’s when her CEO put her in touch with me.

---

The first conversation

We had a call. I got to understand her challenge, the team, the lay of the land, what she was doing and where she was struggling for time.

We mapped it on a Miro board together. Seeing everything she was juggling in one place — every channel, every vertical, every piece of content waiting on her desk — made the real bottlenecks easier to spot. It’s hard to fix what you can’t see all at once.

It’s the only way I know how to build something that actually helps you, rather than something I think helps you. You know your function. I know what’s possible with AI. The work is in the middle.

By the end, we had a priority list. The brand layer was first.

---

So what did we do

Here’s the thing most marketing leaders miss: AI is only as good as what you feed it. If you don’t have a clear brand layer, your AI outputs are a coin flip. If you do, every agent you build sits on the same source of truth — and the quality compounds.

So that’s what we built first.

Together we wrote down how the company speaks, who it speaks to, what words it uses and avoids, and the patterns she’d been correcting in every draft. This became the brand layer — every agent we built next refers to it before it drafts anything.

Then we built three agents on top of it.

1. An agent for vertical-specific content generation

We built an agent that understands the product, the different verticals, and the pain points of buyers in each. It drafts nurture flows tailored to each vertical — pulling the right case study, statistic, and proof point for that audience.

The framework she’d been using mentally — but had never written down — became the agent’s context.

Eight persona nurture flows shipped in a single day. Work that previously took weeks.

“I’d been planning out the multi-vertical nurture programme for six weeks. We built it in a morning.”

2. A self-serve content library for sales

We built an AI agent that pulled the most useful chunks — quotes, stats, soundbites — out of every new content piece as it published. Tagged by vertical. The SDR team could grab what they needed for cold outbound and for building their LinkedIn profiles around prospecting.

The shift was immediate. The SDR team started posting on LinkedIn. They had real material to send in outbound. The standing weekly “have you got anything I can send?” conversation stopped happening.

“I didn’t realise how much of my Friday afternoons were going into reformatting blog posts into Slack messages for the sales team. Now they just go and get what they need.”

3. An agent that drafts content in your brand voice

We built an agent that drafts content in the brand voice — landing pages, listicles, nurture emails, social posts. All drafted by AI first, then edited by her.

She moved from writing-from-scratch to editing AI-first drafts.

“Now I edit instead of writing. Honestly, I should have done this six months ago.”

---

What changed

The point of AI Build isn’t just to save time. It’s to clear space for the strategic work that gets put off when you’re stuck drafting.

Before: one marketer doing the work of three. Drafting from scratch every week. Sales team waiting on content. Verticals underserved. Strategy work permanently in the queue.

After: a brand layer in place. Three agents drafting in voice. The SDR team self-serving for outbound and LinkedIn. Marketing leader spending her time on strategy and the highest-value work.

She didn’t hire anyone. She didn’t add new tools. She built one layer and three agents — and the function started behaving like one that had three more people on it.

The proof points:

• Meeting connects up 1.5% — direct attribution to the SDR team having useful content to send
• 8 vertical-specific nurture flows shipped in a single day (vs multi-week sprints previously)
• They started ranking for competitor keywords — from a competitor displacement campaign which in just 3 months drove 2 closed won deals

---

What’s next

We’re mapping the AI roadmap. Top of the list: an events AI agent, built before September. The team has six events that month — including their biggest tradeshow of the year, the single largest events-driven revenue moment for the business.

Getting AI in place before they hit the floor isn’t a nice-to-have. It’s the difference between leaving with leads the team can follow up — and leaving with a list of names that goes cold.

---

Want this for your marketing function?

This is exactly the kind of work AI Build is designed for. A 90-day 1:1 programme where I sit with you, understand your specific function, and build the AI layer + the agents that take work off your desk.