An early-stage B2B SaaS founder wanted to drop the medical niche and broaden the brand. I'm pushing in the opposite direction. This is the work in flight.
When I came on, Trilith had a real product behind it but a confused surface. I traced the trouble to three overlapping tensions:
1. The brand was trying to cover too much. Branding services. Hiring services. AI features. For an early-stage company, that's three pitches in one breath. Visitors couldn't tell what Trilith was for.
2. The tagline didn't help. "Listen like a clinician, build like an engineer." Poetic. Also opaque. It didn't say what Trilith does, who it does it for, or why someone should care.
3. The founder wanted to broaden, not narrow. The instinct was to drop the medical specialism, soften the positioning, and chase any traction. The instinct is common. It's almost always wrong for early-stage brands.
"Three pitches in one breath isn't ambitious. It's invisible."
This is an active engagement. The arguments below aren't a tidy after-the-fact narrative — they're the live conversation between me and the founder, still in motion.
"Let's drop the medical line. We'll get more leads."
Reasoning: Wider audience = more demand. Niche feels limiting. Generic feels safer for an early-stage company looking for any traction.
"Hold the medical niche for hiring. Branding can be generic — they're different jobs."
Reasoning: Hiring is where the specialism matters most — clinic owners need someone who understands their world. Branding services scale across industries. Treat them as two products, not one fuzzy brand.
"AI should be in the brand vision — it's our differentiator."
Reasoning: AI is the current category narrative. Investors and prospects expect to see it.
"Cut AI from the vision for now. The brand has to be legible before it can be ambitious."
Reasoning: Three pitches at once — branding, hiring, AI — produces a vision statement no one can repeat. Ship the first two clearly. Add AI when the brand can carry the weight.
"Listen like a clinician, build like an engineer."
Reasoning: It captures the team's character — empathy plus precision. It feels distinctive.
"It describes the team, not the offer. Visitors don't buy a team's character — they buy a clear promise."
Reasoning: A tagline has to survive being summarized to a colleague. This one doesn't. Replace it with what Trilith actually does, in plain words.
"Going generic doesn't widen the funnel. It empties it of conviction."
The argument I'm holding, across all three:
1. A generic brand has nothing to defend. If Trilith is "for any business," then established platforms already do it cheaper, with more polish, and with five years of brand equity. There is no story to tell.
2. A narrow brand creates language. "Hiring for medical practices" is a phrase a clinic owner can repeat to their colleague. "AI-powered branding and hiring for any business" is forgettable.
3. Cutting is the cheapest possible advantage. Pulling AI from the v1 vision isn't giving up on AI — it's refusing to let an early-stage brand drown in scope.
Once we'd separated branding from hiring as distinct propositions, the headline question changed. It wasn't "what's our tagline?" It was "what does each side of this brand actually do, for whom, and why?"
Each surface — landing page, onboarding, email — has to answer those three questions before it earns a CTA.
Note what changed. Same product surface. Same team. The new line doesn't promise more — it promises less, more clearly, to the right people. That's the trade.
The other thing the site needed: a way to push every visitor toward an action without sounding like every other SaaS landing page. ("Get started today!" is content design's white noise.)
I proposed adding a small contextual FAQ block at the bottom of each page — three or four questions tuned to the page content, with answers that ended in a soft CTA matched to where the user actually was in their thinking.
No. Start month-to-month. We won't ask you to commit until you've seen results. Most practices know within 30 days whether the platform fits how they hire.
Probably not. Half our customers run solo or two-clinician practices. The hiring problem is the same; the budget is the constraint. Talk to us about the smaller plan.
Book a 15-min call →Trilith works alongside recruiters, not instead of them. We handle branding and the front of the funnel; your recruiter still closes. Most clinics see fewer rounds, not fewer placements.
The pattern works because the FAQ is doing the job a CTA usually fails at: it meets the visitor where their hesitation actually is. The "Book a call" button only appears under the question where booking a call is the right next step.
Some of these are live in production. Some are still in revision. The list below reflects the engagement at this point in time.
Voice, tone, terminology, and visual brand standards in a single reference document. Delivered, in use.
Site map, page-by-page content structure, and the navigation labels that point people where they need to go. Proposed, in revision.
Hero, value prop, social proof framing, pricing language. In active rewrite — back-and-forth with founder ongoing.
Button labels, error messages, tooltips, empty states, confirmation dialogs. Live in production on trilith.in.
Contextual FAQs at the bottom of each page that double as soft CTAs — meeting the visitor where their hesitation actually is. Pattern proposed; design and implementation pending.
This is an active engagement and the brand is live but young. The fairer measure right now is internal, not metric — the founder can describe Trilith's two products in two sentences without a prompt. Six weeks ago, that wasn't true.
When the engagement closes — or when there's enough live traffic to draw signal from — I'll add the harder numbers here. I'd rather show an honest case study than a flattering one.
I'd push for the audience-narrowing conversation earlier. The founder and I spent the first two weeks aligning on style guide details before we'd settled the bigger fight, and a lot of that early work had to be revisited once we sharpened the positioning.
I'd also build the FAQ-as-CTA system into the IA up front, rather than as an addition. The pattern is strong enough to deserve a dedicated content type in the design system — not a section bolted onto each page.
The lesson, restated: positioning before voice, voice before microcopy. Get the order wrong and the work compounds.
I'm freelancing selectively and open to remote roles. If you're early-stage and trying to figure out what you actually sound like, I can help.
Get in touch →