Datagate
An AI analytics MVP that organised noisy social-media conversation into research for digital-asset investors.
At a glance
- DEC 2023-JUN 2024
- CO-FOUNDER / STRATEGY & PRODUCT
- AI ANALYTICS / DIGITAL ASSETS / PRODUCT STRATEGY
- The MVP reached investor conversations, but Datagate never became a lasting company. Archived here with an honest account of why.
The question
Digital-asset markets produce a constant stream of public conversation. The volume is enormous, the quality is uneven, and the relationship between attention and useful investment information is difficult to establish.
Our starting question was:
Could AI help investors move from an unstructured social-media feed to a more systematic view of sentiment, narratives and market-relevant signals?
Summarising posts was never the goal. The product had to organise that noise into something a research process could actually use.
The approach
We framed Datagate around three layers:
- Signals: public social-media activity and changing narratives.
- Analysis: AI-supported structuring, classification and interpretation.
- Decision context: an interface intended to make the information usable for an investor.
The MVP came together in ten weeks, with a nine-person team covering definition, build and presentation. Alongside the product we developed financial models, a go-to-market direction and an investor narrative.
Ten weeks
Turn a broad claim about AI and social data into a proposition narrow enough to build.
Bring together the people, responsibilities and working rhythm needed to move quickly.
Create the MVP and translate the analytical idea into a usable product surface.
Present the company and its economics across more than 20 investor conversations.
My role
My focus was Strategy and Product: the product definition, coordination across the team, the financial models and go-to-market thinking, and much of the investor story.
The role sat between four different realities: what the product could technically demonstrate, what users might value, what the business would need economically, and what an investor would require before committing.
Decisions
Make the output decision-oriented
Raw social activity was not the product. The interface had to produce structured information that could fit into a research or investment process.
Build the model and the business story together
Product choices affected costs, positioning and market size, so the financial model and go-to-market work developed alongside the MVP rather than after it.
Use the MVP to expose the next uncertainty
The build mattered because it converted abstract confidence into concrete questions about data quality, customer behaviour, differentiation and distribution.
Count objections, not conversations
More than 20 pitches produced reactions and objections. Meeting count proved interest; it never proved demand.
What existed
Datagate reached:
- An AI-powered social-media analytics MVP
- A nine-person team
- Product and operating coordination
- Financial modelling
- Go-to-market development
- More than 20 investor pitches
It did not reach a lasting company, and that distinction is central to this case study.
What it taught me
The experience changed how I read speed. Moving quickly pays when it buys better evidence sooner; it becomes theatre when activity substitutes for proof that the business works.
Datagate showed me that an MVP can answer “can we make this?” while leaving “who needs it enough?”, “why us?” and “can this become durable?” almost untouched.
It also taught me that a founder role is not one job. Product, team, economics and narrative move together, and a weakness in any one of them eventually constrains the others.
Ten weeks was enough to build the product.
It was not enough to find the company.