How We’re Building Flamingo

Straight pricing shouldn't need a sales call. Here's how we built OpenFrame's.
Vlad Marchenko, Head of Marketing at Flamingo, covers the $1-per-device model, why we went per device over per technician, and how we priced AI tokens so the bill actually makes sense.

We publish content across 4 platforms, run an active Slack community of 600+ MSPs, manage 193 vendor analysis pages with thousands of community votes, and host online events. All of it is run by two people with AI assistance. In this episode, Michael shows how we manage the content calendar, community moderation, online event logistics, and why doing all of it from a single system matters.

Generating documentation. We don't have a product manager. Instead, our system watches what the engineering team ships and automatically generates reference docs, tutorials, and feature specs from the source code. We talk about the time it hallucinated setup instructions that didn't exist, how we fixed the trust hierarchy between generated docs and manual notes, and why auto-documentation is the single biggest force multiplier for a small team.

Interviews and case study management. Every beta customer who did a case study agreed to go on video - 100% participation rate across 11 customers managing 5,000+ endpoints. We built an internal conviction dashboard that tracks every customer conversation, maps intent tiers, and connects interview data to revenue projections. Before we turn on billing, we're running structured validation interviews with our top 50 customers to pressure-test willingness to pay.

Roadmap and delivery management. Our product roadmap is public. Users can see what we're building, vote on features, and track what shipped. Behind the scenes it's synced to our project management system so votes from the community feed directly into sprint priorities. We also track every bug fix and enhancement with live status updates, and coordinate across a global team through the same tool.

Lead validation and access code management. We built an internal system that scores every MSP signup before they ever get access to the product. It detects whether a company is actually an MSP, researches them across the web, assigns confidence scores, and syncs everything to our CRM automatically. We control who gets in through cohort-based access codes - 2,090 signups, 585 granted access, 196 actively using the product. Every step is gated and measured.

How does a 15-person startup compete with companies that have thousands of employees? Not by hiring more people – by making every internal process as automated as possible.
In this intro episode, Flamingo CEO Michael Assraf walks through the operating philosophy behind the company: one unified codebase, one shared database, and five internal hubs that run marketing, product, revenue, people, and investor reporting with minimal human touch.
Michael covers why transactional efficiency is non-negotiable for PLG businesses with sub-$100K ACVs, how Flamingo's unified architecture works across its website, community platform, and product, and what the five hubs (Marketing, Product, Revenue, People, and Company) do to disconnect scalability from headcount.
This is the first episode of How We Are Building Flamingo – a weekly show for founders and operators who want to see inside another startup's operating system. Subscribe to follow along.
Connect with Michael on LinkedIn: linkedin.com/in/michaelassraf Learn more: flamingo.run Our MSP podcast: openmsp.ai/podcasts