We asked Webflow developers if they plan the whole site before opening the canvas. Here is what they said.
A Reddit thread started with a simple confession: a developer used to open Webflow first and figure out the site as they went, and it kept costing them redesigns halfway through. Once they moved to a strategy first workflow, research, then information architecture, then wireframes, then Figma, then the Webflow build, the redesigns stopped. The thread that followed had Webflow developers and agency founders compare notes on client sign off, AI assisted planning, and why the build itself is often the easy part. Here is what they told us.
Gareth, Founder at BlockyHQ
Strategy first, every time. My list matches this almost step for step, but there is one addition that saved us after we already had the ordering right: put the client sign off between the IA and wireframe stage and Figma, not after design.
The redesigning halfway through problem has a quieter cousin. The client sees wireframes, nods politely, and only really reads the site once it is designed. Then the feedback that shows up is a structure and copy disagreement wearing a design costume. "This section feels off" usually means "I never actually agreed this section should exist."
The gate that fixed it for us is simple. The client explicitly approves the sitemap, every page's sections, and the actual copy before anything opens in Figma. It is not exciting, but it turns endless revisions into polish.
One caveat I learned the hard way: sign off on every single step can backfire. Clients get approval fatigue and start rubber stamping everything, and the one gate that mattered gets the same nod as a font choice. Fewer, heavier gates work better. One on IA and content before design, one on the design itself, and each one explicit: once you approve this, changing it is a change request, not a revision. The moment sign off has consequences, clients actually read what they are approving.
Pavel Vainshtein, Co-Founder at WebflowForge
One of the most important factors is getting the client to sign off on each step of the design. I usually give three revisions per flow or design, and I keep that limited to components rather than the whole design. If a client already agreed on a design, I will not change the whole thing. That request mostly comes from smaller clients, and it goes away fast once you explain a full redesign costs extra.
I ask big clients for a brief with everything I need up front, since building from scratch without one usually ends up costing them more. Before anything else I analyze competitors to see what works for them, then build a strategy. The website itself is the easy part. Lead generation and acquisition is the hard part.
My UX and UI planning now runs through a few AI platforms, thinking through SEO, AEO, and GEO for impressions, then building an acquisition flow from forms to onboarding to email sequence, plus a scoring system that routes leads to sales, CS, or SDR. For the build itself, I will design the UI and UX with a prompt, move it through Figma via MCP, then let AI carry over the full client and component structure from Figma to Webflow, variables included. That gets a full responsive skeleton built in around two days, compared to roughly three wireframes a day by hand, all after customer approval of course.
Isaac Farrow, Co-Founder and Webflow Developer at Whatifweb.co.nz
Opening Webflow too early is the mistake. Nearly every project that has blown out on us traces back to starting the build before we actually understood the business.
The part I would add is that information architecture is where the time actually gets saved. Map the pages and CMS collections properly first, and both the Figma and Webflow stages go faster, because you are not redoing decisions halfway through. Client First keeps the build maintainable too, which matters the moment someone else has to pick up the project.
Shreyansh Kotak, Lead UI/UX Designer at Perpetual
My team has completed a couple dozen Webflow projects over the last several years, and we follow a very similar strategy first workflow. What has consistently worked best for us:
- Discovery: talk to stakeholders across different business units, and users when possible, then synthesize the findings so everyone is aligned on the core goals first.
- Wireframing: build low fidelity wireframes based on those insights and get client feedback on structure and general direction.
- Design and CMS planning: once wireframes are set, move to high fidelity designs in Figma for final visual design and approval, while planning the CMS structure in parallel.
- Webflow build: only after designs are in a good place do we start building in Webflow. We have tried skipping Figma to go straight into Webflow, and iterating on layout or structure directly in Webflow gets painful if the design direction pivots.
For post launch A/B testing and minor refinements, we will occasionally draft in Figma first, but we often jump straight into Webflow for those quick updates.
Cedrik Dudeke, Klarkode, Webflow Developer and Expert
Strategy always comes first. At Klarkode we spend time on research, user journeys, and information architecture before touching Webflow. Treating Webflow purely as an execution tool keeps projects fast, organized, and scalable. Starting directly on the canvas only leads to wasted time and constant revisions.
Vikas Pasi, Developer at BlitzStudio.xyz
I learned this the hard way too. Early on I would jump straight into Webflow because I was excited to start building, then halfway through I would realize the page structure or CMS was not right and end up rebuilding parts of the site.
Now I spend more time planning before I even open Webflow, and the actual build is much smoother. It saves more time than it feels like in the beginning.
The pattern across every answer is the same. The build is the fast part once the thinking is done. Research, strategy, information architecture, and sign off on structure and copy are what keep a Webflow project from turning into a redesign loop.
Written by
Neal
Content Writer