The Demo Link Is the New Landing Page — And Cross-Border Sellers Should Be Paying Attention
If you sell across borders, your biggest conversion killer isn’t ad spend, shipping, or even tariffs. It’s friction. A prospect lands on your Shopify store, hits a login wall on your SaaS upsell, or tries to trial your Amazon-adjacent app and bounces because they don’t trust a random install prompt from a brand they’ve never heard of. Every extra click between “interested” and “inside” is a tax you pay in CAC. So when a tool shows up that collapses that distance — that turns a link into a live, stateful, pre-configured environment instead of a static marketing page — I sit up. That’s exactly what Portal, built by Zachary Goldman, is betting on.
What Portal Actually Does (And Why It’s Not Just Another Demo Tool)
The pitch is deceptively simple: a link that drops someone into the actual product, running in a cloud computer, configured in whatever state you want. Not a screen recording. Not a clickthrough replica. Not a sandbox clone that drifts from production. The real thing, with guardrails.
Goldman’s framing is sharp, and it comes from lived experience. He was a PM at Microsoft AI working on Copilot in the browser, and the trigger moment was running user research where people refused to install an extension because they assumed it was malware. That’s a very specific pain, but it generalizes fast. Think about any cross-border SaaS, tool, or platform you’ve tried to get a prospect into. The install step, the signup wall, the “verify your email,” the cold-start empty dashboard — each one bleeds intent.
Portal’s answer is to pre-build the state. You configure what’s blocked (buttons, API calls) with a no-code builder, generate a link, and share it like a Google Doc. The recipient is inside, in the state you set. You can join them, or they can explore alone with an optional AI that watches the screen and answers questions. That AI tutor angle is the part most people gloss over, but it’s arguably the most interesting piece for anyone selling complex software internationally, where language and onboarding gaps compound.
The use cases Goldman lists are worth reading as a menu of what’s possible: autonomous user research sessions, first-round interviews where candidates perform a real task in a real account, intake conversations, whiteboarding during calls, support onboarding for enterprise trials that can’t afford 1:1 calls, and forward-deployed engineers sending a POC instead of forcing a cold signup. That last one is the money use case for B2B cross-border sellers.
Why Amazon Sellers Should Care More Than Shopify Ones
Here’s a contrarian take. If you’re a pure Shopify DTC operator, Portal is a nice-to-have. Your storefront is your demo. Nobody needs a sandbox to browse a product page.
But if you sell software, tools, or services to Amazon sellers — or you run a brand with a companion app, a subscription portal, or a B2B wholesale arm — Portal solves a problem you’ve probably been throwing money at. Amazon sellers are some of the most skeptical buyers on earth. They’ve been burned by every “guaranteed ranking” tool and every dashboard that looked great in a Loom video and fell apart in production. A live, pre-configured environment they can poke at without signing up is a trust accelerant that a landing page can never be.
The same logic applies if you’re running an agency, a repricer, an analytics tool, or anything that plugs into Amazon Seller Central. Your demo is your sales pitch. Portal lets you make it interactive without giving away the keys.
How It Stacks Up Against What You’re Probably Using Today
Let me be direct about the competitive set, because “interactive demo” is a crowded category and the differences matter.
Navattic and Storylane are the incumbents most people in SaaS-adjacent e-commerce have encountered. Both do HTML-based clickthrough demos — you screenshot the product, then build a guided tour on top. They’re fast to build, they look polished, and they’re great for marketing sites. But they’re approximations. The prospect clicks a fake button and sees a pre-recorded response. For a simple UI walkthrough, that’s fine. For anything involving real data, real state, or real edge cases, it falls apart.
Walnut sits in the same neighborhood and has pushed harder into personalization and sales-led use cases. Still fundamentally a replica layer.
Reprise and similar tools let you build interactive replicas of your app. Closer, but again — replica, not the real thing.
Loom and screen recordings are the low-tech version of all of this. Cheap, fast, and completely non-interactive.
Portal’s differentiator is that it’s not simulating the product. It’s running the product. That’s a meaningfully different technical bet, and it changes what’s possible. You can’t have a prospect hit a real API, see real database state, or experience a real edge case in a Navattic demo. You can in Portal — because it’s the actual environment.
The tradeoff is complexity. Replica tools are lightweight and safe by construction. Portal is running a live cloud computer, which means it has to solve for isolation, reset, cost, and abuse. Goldman mentions a “reset hook” for wiping the database after a prospect makes a mess — that’s the kind of detail that tells you they’ve thought about the operational reality. But it also tells you this isn’t a five-minute setup.
Where the Math Breaks
Let’s talk unit economics, because that’s where I get skeptical.
Running a cloud computer per demo session is not free. If you’re a PLG company sending thousands of demo links a month, the cost per session matters. Portal’s pricing isn’t disclosed on the launch page, and the self-serve builder just launched, so I can’t tell you whether this pencils out at scale. But the mental model should be: this is a high-touch tool, not a high-volume one. It’s for the demo that closes a $5K ACV deal, not the one that converts a $20/month self-serve signup.
For cross-border sellers, that reframes the use case. Portal isn’t a replacement for your marketing site. It’s a replacement for the sales call, the onboarding session, the POC setup. Those are expensive, human-heavy touchpoints — and if you’re selling across time zones, they’re even more expensive because someone has to be awake at 3am to run them.
The “Send a POC Instead of a Signup” Play
The single most underrated use case in Goldman’s list is the forward-deployed engineer angle: send a POC rather than making a client sign up and hit cold-start friction. If you’ve ever watched a warm enterprise lead die in the signup funnel — because their IT team blocked the domain, because the onboarding email went to spam, because the empty dashboard was too intimidating — you know how much that costs.
A Portal link sidesteps all of it. The prospect is inside, in a working state, in one click. No install, no login, no IT ticket. For cross-border B2B, where you’re often selling into orgs with stricter security postures and longer procurement cycles, that’s a real unlock.
What Cross-Border Sellers Can Borrow From This
Even if you never touch Portal, the thinking behind it is worth stealing.
First, audit your friction stack. Map every step between “prospect is interested” and “prospect is inside your product.” Count the clicks. Count the logins. Count the installs. Then ask which of those you could eliminate by pre-building state instead of asking the user to build it. This applies to your Shopify checkout, your app onboarding, your B2B portal, your agency intake form. Everywhere.
Second, treat your demo as a product, not a marketing asset. Most sellers treat demos as something you do to a prospect. Portal’s framing flips it: the demo is something you ship to a prospect, and it should be as well-designed as your product. That means versioning, that means resetting, that means thinking about what state you want them to land in.
Third, the AI tutor angle is a wedge for localization. If you’re selling into markets where your product’s UI isn’t in the local language, or where your support team doesn’t cover the time zone, an AI that watches the screen and answers questions in-context is a force multiplier. It’s not a replacement for real localization, but it’s a bridge. And for cross-border sellers, bridges are how you test markets before committing to full localization spend.
Fourth, reconsider the “business card replacement” use case. Goldman mentions Portal being used as a business card replacement — instead of handing someone a URL, you hand them a live environment. For B2B sellers at trade shows, conferences, or in cold outreach, that’s a genuinely differentiated move. It’s the difference between “here’s my site” and “here’s my product, try it right now.”
The Analytics Layer Nobody’s Talking About
One of the commenters, Rahul Arulkaran, flags the analytics as a standout feature — and he’s right to. If you’re sending Portal links as your demo, you get to see what prospects actually do inside. Which buttons they click. Where they get stuck. How long they spend. That’s the kind of signal that used to require a full product analytics stack plus a sales team taking notes.
For cross-border sellers, this is a feedback loop that’s usually missing. You ship a demo, you send it out, you hear “looks great” and nothing else. With Portal-style analytics, you’d know that 80% of prospects bail at the same screen, and you’d know it in a week instead of a quarter.
Where I Think Portal Falls Short
I’ll be honest about the gaps, because that’s what you’re here for.
The pricing opacity is a problem. The launch page doesn’t disclose pricing, and “try it for free” without a clear cost ceiling is a red flag for anyone planning to scale usage. For a cross-border seller running demos into multiple markets, you need to know whether this is a $50/month tool or a $5,000/month tool. Not disclosed.
The technical setup burden is real. A “no-code builder” sounds easy, but configuring a cloud computer to run your product in a controlled state — blocking the right API calls, setting the right seed data, handling resets — is not a marketing intern’s job. This is an engineering-adjacent task. For lean cross-border teams, that’s a meaningful cost.
The security surface is non-trivial. You’re letting strangers into a live environment. Portal has guardrails, but every guardrail is a potential bypass. If you’re running this against a product with real customer data, you need to be very careful about what state you’re exposing. The launch page doesn’t detail the isolation model, which is the question I’d want answered before putting this in front of enterprise prospects.
The category is crowded and moving fast. Navattic, Storylane, Walnut, Reprise — these are well-funded, well-entrenched players. Portal’s “it’s the real product” angle is a genuine differentiator, but it’s also a harder technical problem, which means it may be slower to ship features and integrations. The launch page mentions backing from South Park Commons, which is a good signal, but not a guarantee of longevity in a category that’s seen consolidation.
The AI tutor is underspecified. “An optional AI that sees the screen and answers questions” is a great idea. But how good is it? Does it know your product’s docs? Can it be trained on your support macros? Is it multilingual? For cross-border sellers, the multilingual question is the whole ballgame, and the launch page doesn’t say.
What I’d Watch / Test Next
If you’re a cross-border operator considering this, here’s what I’d do this week.
First, go build one Portal for your highest-friction demo — the one where prospects currently have to sign up, install, or sit through a 30-minute call. Use the Linear demo as a reference for what “good” looks like. Time yourself. If it takes more than a day, note where the friction is.
Second, send it to five warm prospects who’ve gone cold in your funnel. Not new leads — people who already showed interest and then stalled. Watch the analytics. See if the live environment re-engages them where a landing page didn’t.
Third, pressure-test the security and reset story before you put anything customer-facing in front of it. Ask Portal directly what the isolation model is, what happens if a prospect does something destructive, and how the reset hook actually works. If those answers aren’t crisp, don’t ship it to enterprise.
Fourth, model the unit economics. Estimate your demo volume, multiply by whatever Portal charges once you get a quote, and compare it to the fully-loaded cost of the sales calls or onboarding sessions it replaces. If it’s cheaper and faster, it’s a buy. If it’s not, it’s a toy.
The bigger takeaway: the landing page era is ending. The link that does something is the next default. Whether Portal is the tool that wins that category or just the one that names it, the direction is clear — and cross-border sellers who internalize that early will have a real edge.
