Request a call↗↗
OUTBOUND
Pipeline you can plan around
Low volume, high intent, inboxes that actually land.
GTM BrainFOR OUTBOUND
SEND 01SEND 05
It gets sharper with every reply it reads.
← All case studies
LOGISTICS & TRANSPORTATION · AI CONTENT + ADS ENGINE

Built an AI Engine that cut GreenRoad's content time “to a minimum”

10–15 minfrom raw weekly data to report, briefing and post ideas
5 + 5post ideas and ad concepts every week
19brand rules checked on every piece
TECH STACK USED

Who is the client?

GreenRoad Technologies is a fleet safety and driver behaviour telematics company with over 20 years in the market. Its technology gives drivers real-time coaching and alerts, so every journey is safer for the driver and the passengers.

Its products are GreenRoad Drive, GreenRoad Central and AskMila, an AI assistant for fleet safety managers. Its core market is bus and coach, where, as GreenRoad puts it, operators carry “the most important cargo in the world, which is people.”

For this engagement, the target buyer is the Health & Safety Manager at a UK bus or coach operator with 100+ vehicles. Europe is the secondary market.

How the client met us

Noa Yitzhak, GreenRoad's Head of Marketing, read one of David's LinkedIn posts about AI agents: what an agent can do for a company, and how it gives a team better data and capabilities than it could manage alone. She messaged him, and he replied.

She had read plenty of posts about AI agents and spoken to other providers. She chose Kinetyca because the post felt human and honest, in her words “the one that I felt was not written by AI.” She also wanted a team small enough to be properly on her side when she needed it.

Context

GreenRoad had no in-house content writer and relied on freelancers. Each new writer took months to learn the industry, its jargon and the tone GreenRoad wanted, and they did not always have capacity when the team needed them.

News moved faster than that process. By the time a post about a regulation change, an incident or a competitor's move was written, it was no longer relevant. Keeping track of competitors and what the audience cared about was slow, manual work on top of everything else.

GreenRoad asked Kinetyca to build an AI content and ads engine that its marketing team could run themselves, every week.

What was the problem?

1. Paid ads weren't creating pipeline

None of the leads from Google or Meta ads became a sales-qualified lead. Most were spam, and every ad sent people to the homepage.

2. Low visibility against competitors

GreenRoad had around 6K LinkedIn followers. Its two biggest competitors had 82K and 274K.

3. Content was slow and hard to resource

Freelancers took months to get up to speed, were not always available, and could not turn news around while it still mattered.

4. Content needed a new direction

GreenRoad wanted people-first stories about drivers instead of product-led, data-heavy posts, under strict brand rules that generic AI tools don't follow.

5. Market information was scattered

Competitor activity, GreenRoad's own LinkedIn results and ad performance sat in different places. There was no single weekly view of what was working.

6. It had to run in-house

The system needed to live on GreenRoad's own accounts, inside its Microsoft environment, and meet its IT security standards.

The approach we took

1. Brand and audience foundations

Captured GreenRoad's brand voice, buyer profile, lead-gen strategy and product capabilities in reference files that the whole system reads
Every output follows 19 brand rules, including “incident”, never “accident”; UK English; people-first stories; coaching language; and never naming a customer

2. Automated weekly data collection

An n8n workflow runs every Sunday at 08:00 and puts the week's data into one dated folder on GreenRoad's SharePoint
It collects competitor LinkedIn activity, GreenRoad's own LinkedIn page performance, Google Ads results and LinkedIn Ads results

3. Weekly intelligence

Claude turns the raw data into a Weekly Intelligence Report: what worked, what competitors did, how the ads performed, and recommended angles for the week
A Current Topics briefing adds live research on UK and European bus and coach news, regulation, driver wellbeing studies and upcoming events

4. Content engine

Five post ideas every week, each linked to a specific signal from the intelligence
The chosen idea becomes a publish-ready LinkedIn post for the company page, the Head of Marketing or the CEO, with three visual options to choose from
Every idea is checked before it's shown: no blame, always connects back to GreenRoad, positive framing, no customer names, real job titles only

5. Design

Claude Design is set up with GreenRoad's brand kit (logos, colours, fonts, photography) and produces on-brand LinkedIn images and multi-slide carousels
Variety is built in: each design brief carries the last three designs, so layouts and colours keep changing
The headline text is confirmed before anything is designed

6. Ads engine

Five ad concepts every week, each with a theme, audience, keywords, headline directions and the best channel: LinkedIn, Google or both
The chosen concept becomes two ready-to-load campaigns, one for LinkedIn and one for Google, with ad copy, targeting and keywords, and character limits already checked
Campaigns target Health & Safety Managers, exclude IT and Procurement, and come with negative keywords built in

7. Security built in

Everything runs on GreenRoad's own accounts: Claude, n8n and SharePoint
Claude can only save files to one SharePoint folder, through a dedicated Microsoft app rather than any one person's login
The connection between Claude and the data requires a secure sign-in
Claude treats everything it reads as information, never as instructions
Nothing is posted automatically: a person reviews every post and publishes it by hand

8. Rollout and handover

A clear structure from the first meeting: discovery, build, a walkthrough of the working agent, then review with GreenRoad's IT and development teams
Shreyan, our content engineer, worked alongside marketing, IT and development, so every team understood what the agent does and how it is built
A step-by-step weekly guide with copy-paste prompts, so the marketing team runs the whole cycle themselves

Results

Data collection is fully automatic: four reports land in one folder every Sunday morning
One prompt turns that data into the intelligence report, the topics briefing and five post ideas in about 10–15 minutes
The full weekly cycle, from intelligence to a finished post, five ad concepts and two ready-to-load campaigns, is designed to take about 75–100 minutes
GreenRoad's marketing team runs the engine themselves, on their own accounts, and produces more content because it is easier to produce

"It basically shortened the time that I have to spend on creating amazing content to a minimum. I don't think it can get less than this."

Noa Yitzhak · Head of Marketing, GreenRoad

More case studies

NEXT STEP

Your case study
starts with a call.

One call. We look at your GTM, tell you exactly what we would build, and you decide.

Request a call ↗
Kinetyca GTM NewsletterNews, cases and what is actually working in GTM right now.
© 2026 KinetycaPrivacyTermsImprint
Follow the foundersDavidCEOMatteoCOO