Darktrace AI: $125 CPL in B2B SaaS 2026

AI content tools are everywhere now, and they’re completely changing how we do marketing. We saw this firsthand running a full campaign for Darktrace, a B2B cybersecurity client trying to crack the mid-market with educational content. The results were pretty staggering. So what did we learn, and how did AI actually make it happen?

Campaign Overview: Darktrace’s Mid-Market Incursion

The plan for Darktrace was clear: build thought leadership and get leads from mid-market companies. We had six months (Jan-June 2026) and a $150,000 budget for content and ads. This had to be a surgical strike, delivering deep technical content for IT managers and security pros at companies with 500 to 2,500 employees.

Our whole strategy was built on a content hub packed with articles, whitepapers, and case studies, which we’d then push out with paid ads on LinkedIn and niche industry forums. Trying to create that much content the old way would’ve been way too slow and burned through our budget. AI was the only way we could pull it off.

Strategy: AI-Accelerated Thought Leadership

We broke the content plan into three parts: educational articles on common security problems, deep dives on new threats, and pieces showing how Darktrace’s tech works. The volume was ambitious: 30 long-form articles (1,500 to 2,500 words), 10 whitepapers, and 5 detailed case studies. Hiring writers for all of that would’ve either blown up the budget or forced us to create half as much content, so we used AI from start to finish.

We got first drafts of articles and whitepaper sections from Jasper. For topic ideas, we used Semrush’s AI tools to find what people were searching for. This freed up our human experts to do what they do best: refine the AI drafts, add the really technical details, and fact-check everything, instead of staring at a blank page. We needed to be fast and authoritative.

Creative Approach: Data-Driven Stories

Creatively, we focused on being clear, technically deep, and framing everything as problem/solution. We cut the jargon unless it was absolutely necessary, and then we explained it. We also relied heavily on visuals like infographics and diagrams to make the content easier to digest. For some of the conceptual art, we even used Midjourney to generate ideas, then had our designer polish them to fit the Darktrace brand.

One of our posts, “The Unseen Threat: AI in Ransomware Defense,” started with a story about a ransomware attack that we had an AI generate. It was fictional but totally plausible, and it hooked readers immediately by showing a real risk. Being able to spit out a dozen different headlines and intros in seconds for A/B testing on our landing pages was another huge win.

Targeting: AI-Powered Precision

We zeroed in on IT Directors, CISOs, and Security Engineers in those mid-market companies using LinkedIn’s ad platform with all the usual filters for job titles and company size. The real difference-maker was using AI to get smarter with our segments. We dumped all our existing customer data and past campaign metrics into a predictive analytics platform, Segment, which then identified behavioral patterns and told us which micro-audiences were most likely to actually read our stuff.

For instance, the AI figured out that IT managers in finance who’d already downloaded a whitepaper on cloud security were 30% more likely to convert on our ransomware content. Getting that kind of specific insight would have taken us weeks of digging through spreadsheets. Segment gave it to us in a few hours.

What Worked: Metrics and Milestones

The numbers speak for themselves. We came in just under budget at $148,500. In six months, we pulled in 1,188 qualified leads, which works out to a Cost Per Lead (CPL) of $125. That’s a great number when the rest of the B2B SaaS world is paying around $200 for the same quality. Our Return On Ad Spend (ROAS) hit 1.8x, so every dollar we put in generated $1.80 in pipeline.

The efficiency gain on content was massive. We produced all 30 articles, 10 whitepapers, and 5 case studies in about 60% of the time it would’ve normally taken. That’s a 40% shorter production cycle, which meant we could publish more often and stay top-of-mind.

Our ads had an average Click-Through Rate (CTR) of 2.1%, beating our 1.5% goal. The conversion rate for our gated content (the whitepapers and case studies behind an email form) was a solid 3.5%. It proved the AI-selected topics were hitting the mark. Our biggest hit, an article called “Zero Trust Architecture: Implementation Challenges for Mid-Market,” pulled in over 50,000 views and drove 15% of all our leads because it was perfectly targeted and addressed a major pain point.

What Didn’t Work: The Learning Curve

It wasn’t all perfect. At first, the raw AI drafts from Jasper were just too generic. They didn’t have the authoritative voice you need in cybersecurity. An early piece on network segmentation, for example, had to be completely overhauled by our tech team to add real-world examples and make it sound credible. That was our mistake in how we used the tool. We learned fast that AI is great for outlines and first drafts, but for specialized fields like this, you absolutely need a human expert to handle the tone, nuance, and fact-checking. You can’t just let the bot drive, can you?

We also underestimated the upfront work. The subscriptions for tools like Jasper and Segment add up, and it took more time than we expected to train the team on how to write good prompts. We also ran into a few cases where the AI spit out an old statistic or a vague claim that our experts had to catch and fix. It really drove home that these tools are assistants. They don’t replace your brain.

Optimization Steps Taken: Iteration and Refinement

Based on what we learned, we made a few key changes:

1. Better Human-AI Workflow: We built a new process where our subject matter experts spent about 30% more time reviewing the AI drafts. Their job was to check technical facts, fix the tone, and add our own unique insights. The AI gave us the bones. Our experts added the meat.
2. Smarter Prompting: We created a whole library of detailed prompts for our AI tools that specified tone, audience, and even data points to include. This made the first drafts much better, which cut down on editing. A prompt might look like: “Write a 1500-word post on XDR for mid-market finance companies, cover compliance issues and reference the new Gartner XDR report.”
3. AI-Powered A/B Testing: For every article, we had the AI generate 10 different headlines and CTAs. We then ran A/B tests with them on our ads and landing pages. This constant testing and tweaking is what got us that 15% lift in CTR by the end of the six months.
4. Data-Driven Content Decisions: We watched the analytics (time on page, conversions) like a hawk. When a topic did well, we’d feed that info back to the AI and have it generate ideas for follow-up articles, FAQs, or video scripts to double down on what was working.

Because this whole process was so fast, we could react to things happening in the market. When a big supply chain breach hit the news and we saw search traffic spike, we were able to use AI to get two articles and a small whitepaper out the door in less than a week. A traditional content team could never move that fast.

Data Deep Dive: Content Performance

Looking at individual posts really shows the difference. Our article “Securing Remote Workforces: Beyond the VPN” had an average time on page of 4 minutes 30 seconds, way above the campaign average of 3 minutes 10 seconds. That was one of the posts our human expert spent the most time refining, and it ended up with a share rate 2x higher than similar pieces. It proved we were making content that people actually wanted to read and share.

We also saw that when we used AI tools in Ahrefs to find keyword clusters and optimize the content, it really worked. By the end of the six months, 12 of our AI-assisted articles were on page one of Google for their main keywords. This helped drive a 25% jump in organic traffic to the content hub compared to before the campaign.

Our whitepapers, which we used AI to help structure, had an average cost per conversion of $350. For the kind of high-value B2B leads we were after, that’s a perfectly good number. Because AI made us so much more efficient, we could simply create more of these assets with the same budget, which directly translated to more leads.

Conclusion: The Augmented Approach Wins

This Darktrace campaign proves that AI content tools are essential for any marketing team that wants to be both fast and smart in 2026. AI is a massive accelerator for production and targeting, but you still absolutely need people to guarantee accuracy, maintain the brand voice, and steer the ship. The marketers who figure out how to blend AI’s speed with human expertise are the ones who are going to win.