Colleagues, today we’re putting a most curious job opening under our microscope. **Mercanis** is looking for a Marketing Director in Berlin. And this is one of those cases where the irony isn’t just the cherry on top, it’s the whole cake itself.
**A brief about the company:** Mercanis is “The AI Procurement Platform.” Their mission is to make procurement digital, transparent, and intelligent. They use AI to automate the entire procurement process, aiming to “free procurement departments from operational tasks to focus on strategic objectives.” Remember that phrase; it will come in handy later. The company has raised over $30 million in a Series A round, so their intentions are serious.
**Their current situation and pain points:** They have a product, investor funding, and a team of 6 marketers. Now, they need magic – scaling to €30 million ARR. This requires a one-person band who will:
* Develop and execute a GTM strategy for entering the UK and French markets.
* Transform the company into a “Thought Leader” and build relationships with analysts from Gartner/G2.
* Shift the company from outbound marketing to powerful inbound marketing.
* Act as a “player-coach” for the team, combining strategy with hands-on work.
* And, of course, work closely with sales and development, leveraging data and automation.
Essentially, Mercanis is looking for a brain. A strategic, creative, experienced, multitasking, and empathetic brain to run their marketing machine. And they’re willing to pay €100-120k + bonuses for it. A decent salary for a good human brain. But wait… they themselves are building a business on replacing routine and even strategic human tasks with AI. Don’t you find that amusing?
**How could this task be solved with AI?**
Let’s imagine a dialogue in a Mercanis meeting room. On one side, the hypothetical CEO; on the other, me, your humble servant, in the role of an IT manager with AI evangelist tendencies.
**CEO:** We need a Marketing Director. A strategist. A leader. Someone who will get us to 30 million. Someone who understands the market.
**Me:** Absolutely. But what if I told you that the core of this “director” could be assembled in a couple of months, would cost less, work 24/7, and make decisions purely based on data, not “gut feeling”? We’re an AI company, after all. Let’s eat our own dog food.
**CEO:** Are you talking about a chatbot that writes LinkedIn posts? We need someone who can talk to Gartner!
**Me:** Not exactly. I’m talking about creating an “AI Marketing Director” system. It’s not a single tool, but a combination of neural networks and services that will cover 80% of the tasks in your job description. For the remaining 20%, we’ll hire not a director, but an operator-validator for this system.
### Approaches, Tools, and Implementation Steps
So, what would our “AI Director” look like in practice?
**1. Strategic Core (The Brain):**
* **Task:** Developing GTM strategy, analyzing UK and French markets, positioning.
* **Tools:** We take GPT-4, Claude 3, or another powerful LLM. We “feed” it all internal documentation: data on current clients, win/loss analysis from sales, product descriptions, all marketing materials. We connect it via API to market analytics services (like SEMrush, Ahrefs, Similarweb) and news aggregators.
* **Process:**
1. We set the task: “Analyze our top 10 competitors in the UK B2B SaaS Procurement market. Identify their USPs, promotion channels, and communication tone. Propose 3 positioning options for Mercanis, considering our target audience – CFOs and Heads of Procurement.”
2. Next: “Based on the chosen strategy, develop a 3-month content plan to establish us as a Thought Leader. Topics should be oriented towards the pain points of the target audience. Suggest formats: white papers, webinars, articles, posts.”
3. The AI delivers not just ideas, but a structured plan with KPIs, budgets, and timelines.
**2. Content and Inbound Engine:**
* **Task:** Content creation, campaign management, lead generation.
* **Tools:** A combination of Jasper/Copy.ai (for text generation), Midjourney (for visuals), Descript (for video and podcasts), HubSpot/Pardot (for automation). All this managed by our “strategic core.”
* **Process:** The AI core assigns tasks to content generators: “Write an article on topic X in style Y for audience Z.” The finished text is automatically sent for localization via DeepL API for the British and French markets. Then, the system automatically creates social media posts, sets up email newsletters, and assigns tasks for visual creation. At this stage, a human (operator) merely reviews and clicks “Approve.”
**3. Analytics and Analyst Relations Module:**
* **Task:** Building relationships with Gartner/G2, data analysis.
* **Tools:** Integration with CRM, Google Analytics. Parsers for monitoring Gartner reports and G2 reviews.
* **Process:**
* The AI monitors mentions of competitors and key topics in analyst reports 24/7. When a relevant report appears, it automatically prepares a summary for the team and suggests talking points for an analyst pitch.
* It can even draft emails for analysts: “Dear [Analyst Name], I noticed your latest report ‘Magic Quadrant for Procurement Suites.’ Our Mercanis platform solves problem X, which you discussed, using Y. We would be happy to provide you with a demo.”
* The entire sales and marketing funnel is analyzed in real-time. Instead of weekly reports, the AI itself writes in Slack: “Attention, conversion on the landing page in France dropped by 15% after the last update. I recommend reverting the headline text to version B.”
**Mitigating Distrust:** No one is suggesting firing the entire team tomorrow and trusting Skynet. Implementation should be phased:
* **Pilot Project:** Launch the AI system in one market or for one product line. For example, entrust it with fully managing a campaign in the UK.
* **Human vs. Machine A/B Test:** Give the same task (e.g., launching a campaign to promote a new webinar) to a human team and to the AI system managed by a single operator. Compare speed, cost, and results (CPL, number of registrations).
* **Human-in-the-loop:** Initially, the human is not just a validator but a mentor. They correct the AI’s results, thereby training it. The team’s role shifts from “doers” to “coaches” and “high-level strategists.”
### How to Validate AI Results?
Validation here is even simpler than evaluating a human director’s work. AI doesn’t have “bad days,” “burnout,” or “personal motives.” There is only data.
1. **Strict KPIs:** We compare metrics directly. Cost Per Lead (CPL), Return on Marketing Investment (ROMI), Net New ARR growth rate. The numbers either add up or they don’t.
2. **Speed and Volume:** We measure how many hypotheses and campaigns the system can launch and test in a month compared to a human team. The results will surprise you.
3. **Qualitative Feedback:** We use AI to analyze sales call recordings. Do they hear the same key messages from clients that our “AI Director” generated? If so, the positioning is working.
Instead of searching for one person who must be a strategist, a creative, an analyst, and a manager, Mercanis could invest part of that salary into creating a system that performs these functions. And with the remaining money, hire a brilliant operator to conduct this AI orchestra.
After all, if you’re building the future of procurement with AI, perhaps it’s time to start with the future of marketing? Just food for thought.
Источник: https://www.linkedin.com/jobs/view/4393502803/
