I recently joined a new greenfield project as a Frontend Lead. We’re building two apps: Main app Admin/CMS app

Both share the same auth flow (minus registration), same design system, same utilities, and a lot of reusable CMS components.

Team size: around 8–10 developers.

My proposed architecture: I suggested we go with a modular monorepo using Nx because: Easy sharing of code/modules. Single place for bug fixes (no versioning hell for the design system). Strong module boundaries via tags. If we ever need MFEs later, the structure already supports that progression.

During development, I already needed to fix multiple things in the design system. With Nx, I patched them directly without having to open PRs across repos and publish new versions.

For early-stage products, I believe MFE should be driven by business needs, not technical curiosity. And right now the business doesn’t require separate deployments, nor do we have the scale that justifies microfrontends.

The issue: Even though our company is building the project, the client also has their own IT department, and every architecture decision must be approved by two architects on their side.

They’re not explicitly saying “We want MFE,” but they keep circling back to the same question: “Why aren’t you using microfrontends?” The only justification they give is separate deployments, which we could easily achieve by: Nx affected commands Completely independent pipelines per app Or even separate build targets triggered only by changes None of this requires MFEs.

My concern Implementing MFEs at this stage will: Slow us down significantly Increase complexity and overhead Require us to maintain multiple environments, shells, adapters Impact delivery time and feature velocity Add long-term cost without short-term value

I even asked for the client architect to confirm in writing that microfrontends are an explicit requirement — and that he acknowledges the delays and complexity this brings. He didn’t give a direct answer.

My question to the community Would you: Stand your ground, stick with a modular monorepo + Nx, and push back until the business provides a real reason for MFEs?

Or

Give in and architect the whole thing as MFEs even though the business doesn’t require it, and the project risks missing deadlines? Curious how others in similar leadership roles would handle this.

TL;DR Greenfield project, two apps, 8–10 devs. I proposed an Nx modular monorepo because business needs don’t justify MFEs. Client architects keep asking “why not MFE” but give no real reason besides “deployment flexibility,” which can be achieved without MFEs. Should I push back and stick to monorepo simplicity, or give in and build MFEs even though it adds unnecessary complexity?

Sorry for the long post.