From consultant to platform: how Maya built a cold email agent that books meetings for 2,400 subscribers

Maya Okonkwo spent six years running outbound sales for B2B startups. Before Laavod, she was charging clients between four and eight thousand euros a month to write their cold email sequences. Most months she ran six clients in parallel. Most weeks she worked sixty hours.

She is now charging nineteen euros a month. She has 2,400 subscribers. Her annual recurring revenue is higher than it ever was as a consultant. She works thirty hours a week.

This is how she did it.

The work that was getting repeated

For most of her consulting career, Maya was solving the same problem over and over. A B2B founder would hire her. The founder would describe their ideal customer profile. Maya would build a list of prospects, research each one, write personalized sequences, set up the sending infrastructure, and tune everything based on reply rates.

After her third or fourth client, she noticed the pattern. The research stage was repeatable. The sequence templates were portable. The personalization was about combining a prospect's LinkedIn profile, the client's value proposition, and a tone setting. Different inputs, same shape of work.

"I was charging for outcomes, but my time was going into mechanics," Maya said. "Half my hours were just running the playbook. The other half was the actual thinking."

This is the universal precondition for someone becoming an AI creator. They are doing work that has a repeatable structure. They are getting paid for the output, but the underlying process is not the bottleneck. The bottleneck is their time, applied one customer at a time.

The first attempt

In late 2025, Maya tried to turn her playbook into a GPT in OpenAI's GPT Store. It worked, technically. The GPT could take a prospect's LinkedIn URL and a value proposition and generate a sequence.

It also went nowhere.

The GPT Store had no way to charge for usage. People found Maya's GPT through discovery on the platform, used it for free, and never converted into anything she could earn from. She tried adding a Stripe link in the GPT description. OpenAI rejected the update.

"It was the worst of all worlds," Maya said. "I had productized my playbook, but I was giving it away. And the platform I was on had no interest in helping me get paid for it."

She tried again with a Notion template and a Gumroad listing. The template sold a few times. But buyers got a static document, not a running agent. They wanted the work, not the instructions. The friction killed conversions.

This is the gap most creators hit. The work can be productized. The platform to actually deliver it as a working asset does not exist.

Building Cold Email Architect

When Maya joined Laavod in early 2026, she rebuilt her playbook as a published agent. We will not call it a recreation. The Laavod version is materially more sophisticated than the GPT version.

The agent takes three inputs: a prospect's LinkedIn URL, the user's value proposition, and a tone setting (formal, casual, or technical). It produces a three-email sequence designed to book a meeting.

Under the hood, it runs as a multi-step workflow. The first step pulls the prospect's recent activity and writes a one-paragraph summary of what they are working on. The second step combines that summary with the value proposition and identifies the strongest angle for an opener. The third step writes the sequence, with each email designed to follow up on the previous one without repeating the same content.

The agent uses Claude Opus for the reasoning steps and GPT-5 for the final email generation. Switching between models per step is something Laavod handles natively. Maya did not have to build that part.

She also built in a feedback loop. When a user gets a reply (positive or negative), they can mark the sequence as "worked" or "did not work." Those marks feed into a private evaluation dataset that Maya uses to refine the agent over time. The evaluation dataset is part of the agent, versioned along with the prompts and the workflow.

The agent went live on the Laavod marketplace at 19 euros a month per subscriber.

What happened in four months

Maya did not have a marketing plan. Her hypothesis: the people she had been consulting for would prefer to pay 19 euros a month than 4,000 euros a month, even if 19 euros a month was less hands-on.

She wrote about the agent on LinkedIn. Existing clients signed up. They referred other founders. Other founders referred their own networks. The agent showed up in the "what are creators building" surfaces on the Laavod marketplace and started getting subscribers without Maya doing anything.

Four months in:

• 2,412 paid subscribers
• 19 euros per month, or 14 euros if billed annually
• Average reply rate across runs: 12%
• Sequences sent through the agent: roughly 50,000 per month

The math on Maya's side: 2,412 subscribers at 19 euros monthly equivalent is about 45,800 euros in monthly recurring revenue. Laavod takes 5%, leaving Maya with around 43,500 euros per month, before her own subscription cost of approximately 50 to 100 euros depending on her credit usage.

For comparison, her consulting practice generated 30,000 to 50,000 euros per month at full capacity, working sixty hours a week. The agent generates more, requires roughly ten hours of maintenance a month, and serves twenty times as many customers.

Why the agent works at scale

Two factors explain why one consultant's playbook generalizes to thousands of subscribers.

The playbook was already abstract. Maya was not selling her time. She was selling the application of a repeatable framework. The framework existed independently of her hands. The agent operationalizes the framework. The framework's value scales because the application scales.

Subscribers bring their own context. Cold email is fundamentally about combining a prospect (the subscriber's data) with a value proposition (the subscriber's positioning). The agent does not need to know everything in advance. It needs to ask the right questions and combine the answers the right way. The subscriber's data is the asset that makes the output relevant, not Maya's data.

This is the pattern that scales. A creator's expertise is the framework. The subscriber's context is the input. The framework, productized as an agent, applies the creator's judgment to the subscriber's specific situation, automatically, at any hour, for one-tenth the price of the creator's time.

What this looks like from the subscriber side

A subscriber typically uses Cold Email Architect once or twice a week. They paste a LinkedIn URL, enter or reuse a value prop, set the tone, and run the agent. Total time from input to sequence: about ninety seconds.

The credits a subscriber spends on running the agent come out of their own Laavod allowance. Maya does not pay for those credits. This is a key part of the economics. Maya gets the subscription revenue. The subscriber pays for the AI compute their own usage consumes. The platform takes 5% on the subscription. Nobody is subsidizing anyone else's usage.

For subscribers, the math also works. A junior salesperson paid to write cold emails would cost an employer at least 2,000 euros a month in salary. The agent costs 19 euros and produces better-targeted sequences faster. Even at the bottom of the labor market, the agent is dramatically cheaper than the human alternative.

What Maya built next

In her fourth month on Laavod, Maya published a second agent: Weekly Competitor Intel. The format is similar. Subscribers enter a list of competitor companies. The agent monitors public sources every Monday at 7am, summarizes any meaningful changes, and posts a report to a Slack channel.

The second agent has 1,800 subscribers at 19 euros a month. Total monthly recurring across both agents is now north of 80,000 euros.

Maya's next two agents are in development. One is for outbound voicemail scripts. One is for handling cold email replies that require a meeting-booking flow. She is also considering bundling the agents into a "Cold Outbound Pack" that subscribers can buy as a single subscription.

The economics of being a Laavod creator look like the economics of being a SaaS founder, except Maya is not running engineering or paying a sales team or building a customer success function. She is the product, the brand, and the maintenance team. The platform handles the rest.

What we learned watching this happen

Cold Email Architect is one of the assets we point to when people ask what a Laavod creator looks like at scale. Three things are true about it, and we think they will be true about most successful Laavod creators.

The creator started with real expertise. Maya was not learning cold email when she built the agent. She knew the craft cold. The agent is the productization of expertise that already existed.

The agent is more than a prompt. It is a workflow, a knowledge base, a tool integration, and an evaluation dataset. The composition matters. A single prompt would have been replicable by any competitor. The full agent is differentiated.

The pricing matches the value. 19 euros a month is what the agent is worth to its subscribers. Not what Maya's time is worth. Not what some consultant pricing guide says. The price is set by the actual economics of cold email for B2B sales teams.

We did not build Laavod to replace consultants. We built it for people like Maya, who have built repeatable systems through years of practice and want a platform that treats those systems as assets worth owning, sharing, and earning from.

The next thousand creators on Laavod will be doing this in domains we have not seen yet. Legal research. Diagnostic frameworks. Financial modeling templates. Curriculum design. The pattern is the same. The domain is whatever the creator already knows.

If you have spent years learning something well, your playbook is probably ready to be an agent. The platform exists now.

---

Start free → · See more creators → · Browse the marketplace →