Two phases, one stack, no faff. Brochure site to support the recruitment drive, then a proper portal once the network's earning its keep.
The proposition: recruit creators, coach them to perform better on Live, give them weekly analytics, plug them into a wider network for collaborative growth, and run them through structured TV-style formats nobody else is doing.
For creators who break out, the full talent agency stack from the Gleam days comes online — publishing, licensing, brand deals.
This document covers what I'd build for you, in two phases, with rough effort and cost. Phase one supports the recruitment drive. Phase two is the operational backbone, once there are enough creators on the books to justify it.
Clean, fast, creator-facing. Sets out the proposition, builds credibility, converts interest into signups. The site exists to support the recruitment drive — not to do everything at once.
Next.js, Tailwind, deployed on Vercel. Sanity or flat content for the blog, depending on publishing cadence. Form submissions via Resend or piped to your CRM of choice.
Two to three weeks from kickoff to live, assuming copy and key visuals are signed off promptly.
| Discovery, scoping, copy structure | £500 |
| Design and frontend build | £2,200 |
| CMS setup, signup form, integrations | £900 |
| QA, launch, handover | £400 |
| Total — fixed | £4,000 |
Hosting (Vercel) and domain costs sit outside the build fee, paid by Glowmotion directly. Roughly £20 a month for the basics.
Logged-in. Stats, messaging, briefs, opportunities. For creators it's where they live. For Glowmotion staff it's the operational backbone that makes the network actually scale.
Phase two kicks in once there are enough creators on the network to justify the build. My suggestion: wait until you've got 50–100 active creators on the books, or there's a clear operational reason — talent managers drowning in WhatsApp groups, manual analytics taking hours per week, etc.
TikTok's API access for Live data is more restricted than its main content API. Real-time viewer counts, gift earnings, and Live-specific metrics aren't always exposed at the granularity creators care about.
My recommendation: hybrid approach. Official API where it works, creator self-report through a quick weekly form for the gaps. I'll bake that research into phase one delivery so phase two starts with eyes open.
Next.js 14, Prisma, Supabase (auth, Postgres, file storage). The same modern stack I'm running across two other live SaaS products right now, which means proven patterns and fast moves. Tailwind and shadcn/ui for the UI. Stripe ready if you ever take payments through the portal.
Hosted on Vercel with Supabase doing the backend heavy lifting. Realtime messaging via Supabase Realtime. Background jobs (analytics pulls, weekly reports) on Trigger.dev or Inngest.
MVP in eight to twelve weeks from greenlight, solo with AI agents, assuming clear product direction.
Roughly: weeks 1–2 on auth, schema, and admin scaffolding. Weeks 3–6 on creator dashboard, analytics, and messaging. Weeks 7–9 on talent manager tooling and format management. Weeks 10–12 on polish, QA, and launch.
Directional figures for budgeting, not a quote. Final number depends on what we cut from MVP and what gets pushed. Worth a fresh scoping session once phase one is live and we know more about what creators actually need.
Creator dashboard, messaging, manual analytics input.
£18–22kAll the lean features plus talent manager tools and TikTok API integration.
£28–35kEverything above plus format management and advanced reporting.
£40–50kOperating costs (Supabase, Vercel, third-party services) sit outside the build fee. Expect £50–£100 a month at MVP stage, scaling with usage and creator count.
Built and shipped multiple SaaS products in the last 18 months — retreat management software for influencers, a white-label insurance platform, content automation pipelines — so the modern web stack is hot in my hands.
Working solo with AI agents (Antigravity, Claude Code, Cursor) means I move at speeds that would normally need a team of three. You get one point of contact, fast iteration, and no agency overhead.
And honestly — having known you for 15 years means I'll tell you when something's a bad idea instead of just charging for it.