Research & Positioning · July 14, 2026
I delay design tools on purpose. Strategy, customer language, and decision mapping come first-so pixels serve a real argument instead of inventing one under deadline pressure.
Opening Figma too early creates false progress. Frames look like decisions. They are not. I start every project in conversation, research, and narrative-because design should translate strategy, not invent it under deadline pressure while everyone stares at a blank artboard hoping inspiration will arrive with coffee and a new plugin.
This is not an anti-design rant. I care about craft. Craft without an argument becomes a very expensive mood. The process below is how I keep pixels honest at nau-whether the deliverable is a homepage, an onboarding flow, or a broader product experience. For the formal boxes I fill first, see the strategy framework before designing a screen. The short version: always decide what must be believed first, then design.
Design tools reward visible output. Stakeholders relax when they see frames. Teams debate corner radius while the offer sentence still wobbles. You leave the workshop with a mid-fidelity homepage and the same strategic fog you walked in with-only now it has a type scale. That fog ships. Bounce rates do not care about your auto-layout hygiene.
Nielsen Norman Group has spent decades showing that usability problems often start in unclear information and tasks, not in missing gradients-NN/g. Baymard’s research culture similarly privileges evidence over aesthetic opinion-Baymard. Early Figma tends to invert that order: opinion first, evidence later, if at all.
I collect the words buyers use for the problem, the workaround, and the fear. Those phrases become headline candidates and onboarding copy. Invented brand language can wait. If a founder says “synergistic workflow OS” and a customer says “I just need the roster to stop breaking every Friday,” Friday wins. Y Combinator’s user-facing advice still applies-Y Combinator Library.
What must someone believe before they click, trial, or pay? Typical gaps: “this is for companies like mine,” “it will work with our stack,” “the switching cost is worth it,” “these people are credible.” Each page and section gets a job against a gap. Decoration without a gap job is optional. This is how I avoid the failure mode in why beautiful websites don’t always convert.
Plain text. No hero image. No gradient. If the argument fails in a document, Figma will not rescue it. I pressure-test with someone outside the company. Related reading: clarity vs marketing problems and repositioning in 60 minutes.
What does “better” mean in fourteen days? Qualified CTA rate? Time-to-first-value? Demo show rate? Without a metric, design debates become taste contests. With a metric, taste serves an outcome. First Round’s operator writing is full of this discipline-First Round Review.
| Pre-Figma artefact | Question it answers | Owner |
|---|---|---|
| Verbatim phrase bank | What language is already trusted? | Research / founder |
| Belief-gap map | What must change before action? | Strategist / PM |
| Offer sentence | Who is it for and what changes? | Founder + strategist |
| Success metrics | How will we know the design worked? | Founder + product |
Every screen gets a job: reduce a belief gap or advance a job-to-be-done. Hierarchy follows the argument. Imagery supports meaning. Motion directs attention to the next decision. Components become a system for clarity, not a playground. Apple’s product pages often demonstrate meaning-first hierarchy; Stripe’s marketing historically did the same with restraint. You are not copying their visual skin. You are copying their refusal to let style invent the pitch.
Performance and accessibility enter the design conversation early-not as a QA surprise. Designing six hero variants that each weigh a megabyte is a strategic own-goal. Keep web.dev, web.dev vitals, and WCAG in the room with the artboard.
Canvasly wants a “premium redesign” of their project tool for creative studios. Day one request: open Figma. I refuse politely. We interview four studio producers. Phrases: “client feedback chaos,” “version soup,” “I need one link that isn’t embarrassing.” Belief gaps: will clients adopt it, will it replace email, is migration painful? Offer sentence: “One client-ready link for creative studios tired of version soup-feedback in context, without another bloated PM suite.” Metrics: demo rate from studio ICP, time-to-first shared link. Then we design. The hero shows one link and a short feedback thread-not an abstract purple blob labelled synergy. Conversion improves because the argument arrived before the pixels.
Figma the company, ironically, is a fine reminder that collaborative clarity beats ornamental complexity. Use the tool. Do not let the tool set the sequence.
I keep briefs short enough that founders actually read them. If a brief needs a table of contents, it has become a novel. The sections below are usually enough to unlock design without inviting decorative wandering.
Out of scope is the kindness that prevents mid-project scope creep dressed as “quick explorations.” Explorations are cheap only when the argument is stable. When it is not, every exploration rewrites strategy in pixels-and nobody notices until the bounce rate files a complaint. For client intake discipline that protects this sequence, see questions I ask before taking any client.
I call the pre-Figma phase done when a sceptical outsider can repeat the offer, the team agrees on exclusions, belief gaps have owners, and success metrics are written where everyone can see them. Not when the mood board feels exciting. Excitement is allowed later-preferably after the first conversion review shows the argument working on a phone held by someone who does not love you.
If stakeholders demand visual options immediately, I offer two structural wire directions after the sentence locks-not five brand worlds. Structure debates are cheaper than aesthetic debates. Aesthetic debates without structure are how projects burn weeks choosing between two equally unclear futures. For end-to-end sequencing from strategy into launch, see from strategy to launch.
You do not need a six-week research programme to earn the right to open Figma. Five calls, a support-tag review, a cold paraphrase test, and a mobile walkthrough will expose most belief gaps. The Interaction Design Foundation’s research overviews help teams avoid cargo-cult methods-Interaction Design Foundation-but the standard is still evidence, not ceremony. If research becomes a way to delay decisions forever, you have swapped one form of false progress for another.
Document decisions in the brief so the design file does not become the only source of truth. Files change; premises should be harder to accidentally overwrite with a prettier frame. That is also why I keep success metrics outside the design tool-so “looks done” cannot silently replace “worked.”
When the brief is locked, Figma becomes a translation layer with constraints-and constraints are liberating. Designers make better hierarchy choices when they know which sentence must win. Engineers estimate more honestly when the first outcome is named. Founders argue less about colour when the belief gaps are on the page. That is the real speed gain: fewer circular debates disguised as craft. If you want the same discipline applied to whether the product should exist at all, pair this process with how I decide whether a product is worth building.
Engineers can join pre-Figma work by pressure-testing feasibility of the first outcome and instrumentation. Marketers can pressure-test whether the offer sentence can live in an ad without a paragraph of footnotes. Founders bring veto power on ICP honesty. Nobody needs a style tile to contribute. When polish arrives early, quieter experts stop challenging the premise-because arguing with a beautiful frame feels rude. Rudeness in service of clarity is underrated. Invite the challenge while the brief is still words; it is cheaper than inviting it after three hero variants.
Starting late in Figma is how you avoid building the wrong features beautifully-hidden feature costs-and how you avoid marketing theatre-dig deeper than marketing. It is also why my first days in a company look like diagnosis, not design sprints: if I joined tomorrow.
Harvard Business Review’s pieces on design as strategic advantage assume design is integrated with choice-making-HBR. The Interaction Design Foundation’s UX strategy materials reinforce the same sequence: understand, then shape-Interaction Design Foundation. Google’s content and quality guidance rewards pages that help real people-Google Search Central-another reason argument precedes art direction.
If the morning feels “unproductive,” notice the itch. That itch is addiction to visible output. Learning to tolerate invisible clarity work is a competitive advantage. So is shipping systems instead of spikes-systems over campaigns.
When Figma finally opens, every screen has a job. That is how beautiful work also converts-and how I avoid designing my way into a prettier version of the wrong problem.
It slows false starts. Projects that open Figma first often redo the same screens after strategy catches up. Front-loading clarity is usually faster end-to-end.
Sprints work when the problem and customer are defined. If those are fuzzy, a sprint becomes a workshop for decorating ambiguity. Define first, sprint second.
Yes-on foundations, instrumentation, and spikes tied to the success metric. Avoid polishing UI for undecided narratives.
Low-fidelity structure sketches are fine after the offer sentence. High-fidelity exploration before argument is what I avoid.
Pixels are powerful. They persuade, reassure, and guide. They should not be asked to invent the strategy while the clock runs. Close the tool until the words are true. Then open it-and make those words impossible to ignore. The blank artboard will still be there tomorrow; a fuzzy offer will still be fuzzy after a week of decorative frames.
Want a project process that starts with clarity-not decoration?
See how I workContinue reading
A complementary essay from the same thread.
Swipe left for next, right for previous.
Up next · Research & Positioning
A practical build filter for greenlighting product work: urgency, willingness to pay, differentiation, time-to-first-outcome, and kill criteria-so curiosity does not become a roadmap.
Read essay →Let the right audience find you, with 10 videos only. That's the progress we build together.
Stefani Dimitrova
Organic GTM & Product Storyteller
