DEV Community: Taehak Kim

From Google Form + Linktree to a Real Lead-Capture Page in 30 Seconds

Taehak Kim — Tue, 05 May 2026 10:05:06 +0000

There's a stack a lot of solo operators end up running by accident, and it goes like this:

Linktree in the Instagram bio (because the platform restricts to one link)
A Google Form somewhere in there for "inquiries" or "bookings"
A Gmail label where the form responses pile up and slowly stop being read

It's free, it works, and it's leaving real money on the table. Here's why, and what to replace it with — without paying the Squarespace / Wix / Webflow tax to do it.

What's actually wrong with Linktree + Google Form

Both tools are fine. They're just being asked to do a job they weren't built for.

Linktree is a directory of links. That's its whole job. It does not do conversion. A visitor lands, sees seven links titled things like "Latest Reel" / "Book a session" / "Newsletter", picks the path of least resistance, and most of them pick "Latest Reel." Your inbound flow is being absorbed by your own content. The link that converts is buried.

Google Forms is a spreadsheet with a UI. It captures data well. It does not sell. There's no headline, no proof, no pricing signal, no "here's why I'm the right person" before the form. By the time someone is ready to fill out a Google Form, they're already 90% sold. You're losing the other 10% on the way to the form.

The combined stack is functional. It's also a lead-capture page chopped into two pieces, with a third piece missing entirely: the page that does the selling between the click and the form.

What a real lead-capture page does instead

The job description for the missing page is short:

1. Headline that names the visitor and the outcome. "I help indie SaaS founders write the first 200 newsletter subscribers' worth of emails" outperforms "Welcome to my page" by an order of magnitude. The visitor decides whether they're in the right place in three seconds. Make it easy on them.

2. Three to five sentences of proof. Not testimonials (yet — write the page first, gather them later). Just the what and why-me. Past results. Specific clients. A statistic. Something that anchors you as not-a-random-internet-stranger.

3. A pricing or scope signal. Even "Engagements start at $1,200" or "Free 20-minute scoping call, paid work starts at $X" qualifies the visitor. The 60% who'd waste your time self-eject. The 40% who fit show up pre-warmed.

4. The form, embedded inline, with the right fields. Not a Google Form on a different domain. Same page. Above-the-fold-on-mobile if possible. Fields that filter (project type, budget tier, deadline, "where did you find me"). Auto-confirmation email that doesn't go to Promotions tab.

5. One single call-to-action button. Not "Book a call" and "Read my newsletter" and "DM me on Twitter." Pick one. Visitors don't decide between buttons — they bounce.

That's a lead-capture page. Linktree gets you to step 0. Google Form gets you step 4. The 1-2-3-5 in between is what's missing.

The slow way to add the missing middle

Three classic paths and what each one costs:

Squarespace / Wix: A weekend of your time, $16-$23/month. The form lives on your site (good), the form goes to your email (still bad — no follow-up automation, no CRM, no way to track conversion source).
Webflow + Memberstack + Mailchimp: A real lead-capture stack. $40-$60/month combined, plus a freelancer to wire it up unless you're already a no-code person. Powerful, overkill for most solo operators.
Carrd: $19/year, single page, lovely. Form goes to email or Mailchimp. No CRM, no AI-generated copy, no follow-up. Fine for the first 6 months, painful around month 12.

Each works. Each has a tax. The tax is either time (a weekend), money (Webflow stack), or future-you (Carrd's ceiling).

The 30-second way

Full disclosure: I've been building LabelWebs for solo operators stuck on this exact stack. The workflow is built around the Linktree-to-real-page upgrade, so it's worth describing even if you end up doing it elsewhere.

You describe the business in plain English. "Newsletter ghostwriter for indie SaaS founders, packages start at $1,200, I want a one-page site with a clear hook, three sentences of proof, an inline lead form asking project type / budget / deadline, and a single 'request a free scoping call' CTA." The AI then does four things in one shot:

Auto-injects copy and generates the homepage — headline, sub-headline, three proof sentences, scope-signal sentence, FAQ. Niche-specific, not "Welcome to my freelance services."
Builds the page section by section as code blocks — every section is generated for your business, so it doesn't read like a templated freelancer site.
Drops in a customizable lead form wired into a built-in CRM — fields are yours (project type, budget, deadline, source), inquiries land in a dashboard with status (New / Replied / Qualified / Closed), and you can actually follow up without spelunking through Gmail.
Generates a CaroSpark Instagram carousel from the same prompt — useful for the "now I have a real page, time to drive traffic to it" part, which is the step everyone forgets.

Total time from "I'll fix this" to a live page on yourbrand.labelwebs.com with a working form and CRM: about thirty seconds, plus a couple of minutes to tighten the headline.

For the upgrade-from-Linktree path specifically, the workflow is:

Build the page on the free tier first. Test the form on your phone.
Replace the one most-clicked link in your Linktree with the new page URL. Don't delete Linktree yet — just redirect the converting traffic.
After a week of seeing real form submissions land in the CRM, swap Linktree out entirely and put yourname.com (or the labelwebs.com subdomain) directly in the bio.
The $19/month plan adds your own domain (.com / .net / .org / .me / .info) and bumps the AI to Gemini 3.1 Pro for sharper niche-specific copy on regenerates. The platform also handles the subdomain → custom-domain mapping for you, so you don't end up in DNS-record purgatory the day you switch.

The fields that actually convert

Independent of the platform you use, these are the form fields that move the most for solo-operator lead pages:

Project type (dropdown, 3-5 options — not free text)
Budget tier (3 options, named the way you talk to clients — e.g. "Quick fix ($500-$1.5k)" / "Full project ($1.5k-$5k)" / "Ongoing ($5k+/mo)")
Deadline (3 options: "Yesterday" / "This month" / "Flexible")
Where did you find me? (free text — the data you'll learn the most from over six months)
Email + optional name

Skip "company name." Skip "phone number." Skip "tell us about your project" as the first field — it's an essay-test that scares off 40% of qualified leads.

Try it

If you want to test the 30-second flow on your specific business, try LabelWebs free. The free tier covers everything described above on a *.labelwebs.com subdomain. The $19/month plan adds your custom domain, removes the watermark, and runs your prompt through the better AI tier.

But the bigger point is the takeaway, regardless of platform: Linktree is a router, not a destination. Google Forms is a spreadsheet, not a sales page. The thing you're missing is the page in between, and it's the cheapest leverage you have left in 2026.

Building a Contractor & Home Services Website Without a Developer in 2026

Taehak Kim — Tue, 05 May 2026 07:41:29 +0000

Most contractor websites I look at are doing one of two things badly:

The first kind is the template that screams template — same hero photo of someone in a hardhat shaking hands with a homeowner, the same "Quality Workmanship Since [year]" tagline, the same six "services" boxes that say Roofing / Siding / Gutters / Decks / Remodels / Free Estimates. A homeowner googling for a roof leak at 9pm doesn't book off this. They scroll past it.

The second kind is the dead Facebook page treated as a website. Phone number in the bio, last post six months ago, no way to actually request a quote without a phone call you don't want to make at 9pm. The contractor wonders why their leads dried up; the homeowner already called the next guy on the list.

A working contractor site doesn't need to be pretty. It needs to do three jobs in under fifteen seconds: answer "do you do my exact problem?", answer "do you serve my zip code?", and capture the lead before the visitor closes the tab. Everything else is decoration.

What a contractor lead page actually needs

Strip the bloat. Build these and only these:

1. A specific hero hook, not a category. "Emergency roof leak repair in Fairfax County — same-day response" outperforms "Quality Roofing Solutions" by a wide margin. Specificity makes the visitor feel found, not sold to. Every word that isn't doing work is doing damage.

2. The service list as actual sentences, not icon boxes. "We replace 25-30 year asphalt shingle roofs and handle most insurance claim paperwork directly" tells a homeowner more than six service boxes ever will. Three to five service paragraphs beats fifteen icon-boxes every time.

3. A clear service area. Either a list of cities/zip codes or a one-line "We serve [region]". Half of all bounces on contractor sites are people in the wrong service area realizing it too late and bouncing back to Google. Tell them up front.

4. A lead form with the fields you actually need to triage. Address, problem type, urgency (emergency / this week / no rush), photos. Email is fine but a phone field with tel: link on mobile is better. The form should ping you immediately — homeowners who fill out a form are usually filling out three at once. First responder wins the job.

That's the page. Five sections, no marketing fluff, no "About Us" timeline.

The slow way to build it

The classic options each have a tax:

WordPress + a "contractor theme": $200 setup if you DIY, $1500-$3000 if you hire someone, plus ongoing plugin updates and a security surface area you don't want.
Wix / Squarespace: Look fine for a weekend of work. The lead form goes to your email, where it sits next to spam and Home Depot coupons. No CRM, no follow-up automation, no way to know which job site brought the call.
Local SEO agency: $400-$1200/month. They build the site, but the moment you stop paying, the site goes dark or gets held hostage. Read the contract before you sign.
Facebook page only: Free. Also worth what you pay for it as a lead source in 2026.

The pattern: most contractors end up with something, hate it inside six months, and go back to running on word-of-mouth. The site never gets fixed. The leads never get tracked.

The 30-second way

I've been building LabelWebs specifically because the contractor / home-service / local-service segment is the one that bleeds the most from "site exists but doesn't generate calls." Here's the workflow that works on it.

You describe the business in plain English: "Roofing contractor in Fairfax County VA, residential asphalt shingle roof replacement and emergency leak repair, 24-hour response, work directly with insurance adjusters." The AI then does four things:

Auto-generates the copy and injects it into a homepage — niche-specific hook, services described as sentences not icons, service-area sentence, FAQ block. The output reads like a contractor wrote it, not like a stock theme.
Builds the page section by section as code blocks — generated for your business, so the next roofer in Fairfax doesn't get an identical page. This matters for local SEO; identical pages on the same generator get filtered.
Drops in a customizable lead form wired into a built-in CRM — fields you choose (address, problem type, urgency, photo upload), inquiries land in a dashboard you can triage from your phone between job sites.
Generates an Instagram carousel from the same prompt via CaroSpark — five-slide carousel for posting "before / process / after" content the same day you publish the site. Free shortcut for contractors who can't be on social media all day.

Total time from "I should do this" to a live page at yourbrand.labelwebs.com with a working lead form: about thirty seconds, plus however long you spend tightening the hook line.

If you want it on yourbrand.com, the $19/month plan removes the watermark and connects any .com / .net / .org / .me / .info domain you already own — the platform maps the subdomain to your custom domain automatically without making you touch DNS records. The $19 plan also runs your prompt through Gemini 3.1 Pro instead of Gemini 3 Flash, which matters most for the regenerate when you're tightening service-area copy.

Contractor-specific prompts that produce useful output

Don't write "make me a roofing website." That gives you the same template-screaming-template output as every other AI builder. Try this format instead:

"[Trade] contractor in [county / metro], [specific service 1], [specific service 2], [specific service 3]. Service area: [list of cities or zip codes]. Emergency response: [yes/no]. Insurance work: [yes/no]. Years in business: [N]. License #: [optional]. Lead form should capture address, problem type, urgency (emergency / this week / no rush), and allow a photo upload."

The license number, the insurance-work mention, and the years in business are the three trust signals that move the most homeowners. Include them.

For the lead form fields specifically, the ones that pay off most:

Property address (shows you commitment up front, also lets you check service area)
Problem type (dropdown, not free text — you're filtering, not interviewing)
Urgency (3 options, not 5)
Photo upload (lets you scope the job before you drive out)
Phone number, marked optional but in practice 90% will fill it

Try it

If you want to skip the WordPress / Wix decision entirely, try LabelWebs free. The free tier gets you a live page on a labelwebs.com subdomain, customizable lead form, CRM. The $19/month plan adds your custom domain and the better AI tier.

If your current site hasn't generated a real lead in 90 days, the bar for "improvement" is very, very low. A page that does the four things above, even badly, will outperform a beautiful site that does none of them.

How to Build a Photographer Website in 30 Seconds (With AI in 2026)

Taehak Kim — Tue, 05 May 2026 07:38:52 +0000

Most photographers I know have the same site situation: a half-finished Squarespace from 2022, three "real" portfolios on Instagram / Behance / a Google Drive folder, and a DM inbox doing the actual work of converting strangers into paid sessions.

The honest truth is that nobody books a wedding or a brand shoot off a perfect grid layout. They book off a page that says what you shoot, where you shoot, what it costs to start a conversation, and a form that actually puts their name in front of you. Everything else is portfolio, and portfolio belongs on the platforms people are already scrolling.

This is a guide for that page. Not a portfolio site — a lead-generating page that lives at yourname.com and does one job: turn cold traffic into a filled-out inquiry form.

The four things a photographer's lead page actually needs

Strip away the trends and almost every booking site that converts has the same skeleton. If you're rebuilding from scratch, build these in order:

1. A one-line hook above the fold. Not "Photography by Lena." That's a name, not a hook. Try the format [Type of work] for [type of client] in [city]. Example: "Brand portraits for product founders in Brooklyn." The visitor knows in two seconds whether they're in the right place. If they don't, no amount of Lightroom will save you.

2. Three to five hero images, max. People decide style in seven seconds and then start scrolling for proof you're not a flake. Five images is plenty. The full portfolio link goes to Instagram or a separate gallery. Don't waste the lead page on showing every wedding you've ever shot.

3. A pricing or starting-rate signal. Even just "Sessions start at $450" will save you ten back-and-forth DMs per week. People who can't afford you self-select out, and people who can show up pre-qualified. This single line moves more bookings than any headshot upgrade.

4. A lead form connected to something that pings you. Not a contact email. A form. With the fields you care about: shoot type, date, city, budget tier. Plus an automatic follow-up so you don't miss a hot inquiry while you're on a Saturday shoot.

That's the whole structure. Hero, hook, proof, price signal, form. Nothing else needs to exist on the lead page itself.

The slow way to build it

Pick your poison: Squarespace (a weekend, $16/mo, fights you on form fields), Wix (an entire wasted Saturday, $17/mo, AI builder generates copy that screams "AI builder"), Webflow (you'll need a freelancer, $30+/mo), or some Notion/Carrd hybrid (free, but the form goes nowhere useful and you'll outgrow it in three months).

I've been through all of them. The pattern is always: spend the weekend, hate the result, decide to "redo it properly later," never redo it, watch the Instagram-only era continue for another year.

The 30-second way (if you're willing to try it)

I've been using LabelWebs for the last few weeks. It's a tool I'm building, but the workflow is genuinely the fastest path I've found for the photographer use case specifically — so it's worth describing the steps even if you end up doing it on another tool.

You describe your business in plain English ("brand portrait photographer in Brooklyn, sessions start at $450, I want a lead page that captures shoot type, date, and budget"). The AI then does four things in one shot:

Generates the copy and injects it into a homepage — hook, sub-hook, the three sentences under each section. Niche-specific. No "Welcome to my photography business" filler.
Builds the page section by section as code blocks — not a drag-and-drop template. Each section is generated for your business, so two photographers in the same city won't end up with visibly identical pages.
Spins up a customizable lead form wired into a built-in CRM — the fields are yours to edit (shoot type, date, city, budget), and inquiries land in a dashboard you can actually triage from.
Optionally, uses the same prompt to generate a CaroSpark Instagram carousel — you enter the prompt once, the AI fills text into image templates, and you download a carousel-ready image to post the same day you publish.

Total time from "I'll do this" to a live yourbrand.labelwebs.com page with a working form: roughly thirty seconds plus however long you spend tweaking the hero copy. The free tier covers all of this on a *.labelwebs.com subdomain with a small watermark.

If you want it on yourname.com, the $19/month plan removes the watermark and lets you connect any .com / .net / .org / .me / .info domain — and the platform handles the subdomain → custom-domain mapping automatically (no DNS config rabbit hole). The $19 plan also upgrades the AI from Gemini 3 Flash to Gemini 3.1 Pro, which produces noticeably tighter niche-specific copy on the regenerate.

What to actually fill in (photographer-specific)

If you go this route on LabelWebs or anywhere else, the prompts that produce the best output for photographers tend to look like this:

"I'm a [wedding / brand / product / family / boudoir] photographer in [city]. My ideal client is [one sentence — e.g. 'product founders launching their first DTC line']. Sessions start at [$X]. I want a lead page with a clear hook, three hero images, a one-line pricing signal, a lead form asking shoot type / date / city / budget, and a short FAQ block about turnaround and edits."

Don't write "make me a photography website." Generic prompt → generic output, every time, on every AI builder.

For the lead form, the fields that filter the most aggressively (in my experience):

Shoot type (dropdown — limit to your actual offerings)
Tentative date (so you know if you're even free)
Budget tier (the single most underused filter — three buckets is enough)
"Where did you find me?" (free text — you'll learn what's actually working)

Try it

If you want to test the 30-second flow specifically, try LabelWebs free — no credit card on the free tier. If you hate the output, you've lost thirty seconds. If you like it, you can connect your own domain and skip Squarespace entirely.

Either way, the lesson holds: photographers don't need bigger portfolio sites. They need smaller, sharper lead pages. The portfolio is already on Instagram. The page that books the work is the one missing.

I built an AI website builder that turns a sentence into a live lead-capturing page in 30 seconds

Taehak Kim — Tue, 05 May 2026 07:34:29 +0000

TL;DR

I built LabelWebs, an AI website builder that takes a one-sentence business description and outputs a complete lead-capturing business page in 30 seconds. It also auto-regenerates individual sections as code blocks, ships a customizable lead form that doubles as your CRM, and includes CaroSpark — an AI Instagram carousel generator for the marketing posts that drive traffic to the page. Plus plan: $19/mo, watermark off, stress-free custom domain.

The problem that wouldn't go away

Every freelance photographer, interior designer, and local service contractor I know runs into the same wall: their "online presence" is an Instagram bio with a Linktree, or a Google Form they paste into their bio. Clients land, see no real website, and click away. 75% of consumers judge business credibility by the website. No website = no calls.

The actual options solo operators face today are bad:

Web agency: same outcome you want — a real business page that captures leads — but $2k–$5k upfront, 2–4 weeks of back-and-forth, plus maintenance fees. Most freelancers and local service owners can't justify it.
Google Form + Instagram bio link: what most people actually do. It captures leads, but it's a Google Form. No real page, no SEO, no credibility. The form lives in a vacuum — clients see "Google Forms" and trust drops.
Generic bio-link / page tools (Linktree, Carrd, etc.): fill the visual gap, but none of them are built around lead capture. No CRM, no niche-specific copy, no auto-generated form fields. You end up bolting Google Forms to a Linktree to an email inbox to a spreadsheet.

What if you could just describe your business and get a real lead-capturing page in 30 seconds — no agency, no Google Form duct tape?

How it's built

AI engine: Google Gemini. Free plan generations run on Gemini 3 Flash; the Plus plan extends to Gemini 3.1 Pro for tighter niche-specific copy on regenerates.
Templates: a curated set of conversion-tested layouts. The AI fills content into proven structures rather than inventing layouts from scratch — that's most of why the output looks like a real business page instead of an AI hallucination.
Hosting: pages go live on *.labelwebs.com subdomains with edge caching. The Plus plan adds stress-free custom domain mapping — paste your .com / .net / .org / .me / .info and the system handles the mapping automatically, no manual DNS edits.
Payments: Paddle (subscription billing, MoR).

What it actually does

1. AI text injection → homepage in 30 seconds

You type one sentence about your business. Gemini does the rest: hero copy, features, CTAs, lead form fields, SEO meta — all tailored to your niche. Photography studio, cleaning service, marketing consultant — the copy reads like it was written for that specific business, because it was.

2. Per-section code block regeneration

Each section of the generated page is its own AI-managed code block. Don't like the hero? Regenerate just the hero. The lead form copy feels off? Regenerate just the form section. No "rebuild the whole site to tweak one paragraph" loops.

3. Customizable lead form → built-in CRM

The form isn't a generic "Name / Email / Message". It auto-generates fields that fit the business — service date for cleaners, event date for photographers, square footage for renovators. Inquiries land directly in a built-in CRM with status tracking. No Zapier, no Google Form, no leads-die-in-inbox problem.

4. CaroSpark — AI Instagram carousel generator

This is the part most people don't expect. Once you have a page, you need traffic to it. Most of that traffic comes from Instagram, and the posts that work are the multi-image carousels with text overlays — the ones that look like mini-articles. Writing those is a chore.

CaroSpark fixes it: pick a carousel template, type a prompt about what you want to say ("3 reasons local plumbers lose leads"), AI injects the text into the slide-by-slide image templates, and you download the whole carousel ready to post. The page lives at labelwebs.com, the carousels drive people to it, and the lead form catches them when they land.

The $19/mo Plus plan: stress-free everything

Watermark removed — looks like your site, not ours
Custom domain connection — bring your own .com, .net, .org, .me, .info. The flow is the part most people dread, so we automated it: paste your domain, the system handles the mapping, and your existing subdomain.labelwebs.com page now lives at customdomain.com without you touching DNS records you don't understand. No stress, no DIY DNS.
Unlimited subdomains for separate niches/campaigns
Full CRM features
Monthly AI credits

What's next

More niche-specific templates (currently strong on services + creators)
Naver SEO (a meaningful slice of the audience is Korean small business)
Auto-generated location+niche landing subpages

Try it

labelwebs.com — free plan available, no credit card to start. Honest feedback would mean a lot. What would you change?

I built an AI website builder that turns a sentence into a live lead-capturing page in 30 seconds

Taehak Kim — Sat, 02 May 2026 09:18:59 +0000

TL;DR

The problem that wouldn't go away

The actual options solo operators face today are bad:

Web agency: same outcome you want — a real business page that captures leads — but $2k–$5k upfront, 2–4 weeks of back-and-forth, plus maintenance fees. Most freelancers and local service owners can't justify it.
Google Form + Instagram bio link: what most people actually do. It captures leads, but it's a Google Form. No real page, no SEO, no credibility. The form lives in a vacuum — clients see "Google Forms" and trust drops.
Generic bio-link / page tools (Linktree, Carrd, etc.): fill the visual gap, but none of them are built around lead capture. No CRM, no niche-specific copy, no auto-generated form fields. You end up bolting Google Forms to a Linktree to an email inbox to a spreadsheet.

What if you could just describe your business and get a real lead-capturing page in 30 seconds — no agency, no Google Form duct tape?

The stack

Frontend: Next.js + Tailwind for the marketing site, dashboard, and the rendered customer pages
AI text injection: Gemini 3 Flash for fast section drafts, Gemini 3.1 Pro for the deeper niche-specific copy (sales angle, lead form fields, SEO meta)
Templates: a curated set of conversion-tested layouts. AI fills the content into proven structures rather than inventing layouts from scratch.
DB / CRM: Postgres with a lightweight pipeline schema (lead → status → follow-up)
Hosting: wildcard subdomain (*.labelwebs.com) with edge caching; custom domain auto-connection via CNAME
Payments: Paddle

What it actually does

1. AI text injection → homepage in 30 seconds

2. Per-section code block regeneration

3. Customizable lead form → built-in CRM

4. CaroSpark — AI Instagram carousel generator

The $19/mo Plus plan: stress-free everything

Watermark removed — looks like your site, not ours
Custom domain connection — bring your own .com, .net, .org, .me, .info. The flow is the part most people dread, so we automated it: paste your domain, the system handles the CNAME logic, and your existing subdomain.labelwebs.com page now lives at customdomain.com without you touching DNS records you don't understand. No stress, no DIY DNS.
Unlimited subdomains for separate niches/campaigns
Full CRM features
Monthly AI credits

What's next

More niche-specific templates (currently strong on services + creators)
Naver SEO (a meaningful slice of the audience is Korean small business)
Auto-generated location+niche landing subpages

Try it

labelwebs.com — free plan available, no credit card to start. Honest feedback would mean a lot. What would you change?