The six-tap problem is your six-tap problem
Every cross-border operator I know is running some version of this. Not solar panels — dashboards. You have Amazon Seller Central open in one tab, Shopify admin in another, TikTok Shop Seller Center in a third, a Temu supplier portal in a fourth, and somewhere in a spreadsheet is the number you actually wanted: today’s units. The number that tells you whether the ad spend is working, whether the inventory plan is holding, whether the day is going to close above or below yesterday. And getting to that number takes six taps, three logins, and a mental model of which platform owns which metric this week.
That’s the lens I want to use for Helio, a small self-hosted solar dashboard that launched on Product Hunt this week. On its face it has nothing to do with e-commerce. It’s a rooftop-solar companion app for people with Eybond / ShineMonitor inverters. But the design decisions its maker Rishi Shah made — one number on open, self-hosted, no database, no tracking, deploy in five minutes — are a masterclass in the kind of internal tooling that most cross-border sellers still refuse to build for themselves. And the fact that a solo maker shipped it because his own vendor app was too slow tells you something about where the real operational leverage is in 2025.
What Helio actually solves (and what it refuses to solve)
The pitch is almost aggressively narrow. Rishi’s own words: his rooftop solar came with the SmartClient app, and every day he opened it for one number — how many units the roof made today — and it took six taps to find it. So he built Helio to open straight on today’s units, read from the inverter’s own counter, with live power, history, and savings in the user’s own currency around it.
That’s it. That’s the product. And that’s precisely why I find it interesting.
The default SaaS instinct for a problem like this would be to build a “solar intelligence platform” — forecasting, panel-level analytics, financing integrations, a mobile app with five tabs, a subscription tier. Rishi went the other way. He built a single-purpose view that surfaces the one metric the user actually opens the app for, and then layered in a small number of secondary features that are genuinely adjacent: panel health (it distinguishes a cloudy day from a dirty panel, and tells you when cleaning will actually help), a grid check that shows when high voltage made the inverter switch off, a daily Telegram summary every evening, and CSV export of every day.
Notice what’s missing. No AI assistant. No “insights.” No push notification strategy. No account system in the SaaS sense — the login stays on your own server, there’s no database, no tracking. You deploy your own copy to Vercel in five minutes, free. The only supported hardware is Eybond / ShineMonitor, and Rishi is explicit that inverter brand determines what he supports next.
For a cross-border seller, the takeaway is not “go buy Helio.” The takeaway is the shape of the thing: a one-metric tool that opens on the number you actually check, self-hosted so nobody else has your data, and deliberately limited in scope so it doesn’t become another dashboard you have to maintain.
Why this matters more for Amazon sellers than Shopify ones
Shopify merchants have it comparatively easy. The admin is clean, the analytics are decent, and if you want a custom view you can pull the Shopify Admin API and build a Metafield-backed dashboard in an afternoon. Amazon sellers live in a different world. Amazon Seller Central is a sprawl of reports, most of which are lagged, some of which disagree with each other, and none of which are designed around the question “did today’s units clear yesterday’s?” If you sell on both FBA and FBM, add Amazon Vendor Central for wholesale, and layer in TikTok Shop Seller Center for the social channel, you are now maintaining four different mental models of “today’s units” and reconciling them by hand.
That’s the six-tap problem, scaled. And Helio’s answer — read the number from the source system’s own counter, don’t recompute it — is the right answer. Most seller dashboards fail because they try to recompute revenue from order events and end up disagreeing with the marketplace’s own settlement report. Reading the counter is boring, but it’s the only number that reconciles at month-end.
How it stacks up against the incumbents you’d actually compare it to
If you’re a seller thinking “okay, so what would I build,” the honest comparison set isn’t other solar apps. It’s the tooling you’re already paying for.
On the marketplace side, the default answer is Helium 10 for research and Jungle Scout for sales estimates, plus SellerApp or DataHawk if you want a consolidated Seller Central view. These are good at what they do — competitive research, keyword tracking, PPC analytics — but they are not built to be the one screen you open at 9am to see whether yesterday closed above plan. They’re built to be the screen you open when you’re doing analysis, which is a different job.
On the DTC side, the default answer is Shopify plus Klaviyo plus Triple Whale or Northbeam for attribution. Triple Whale in particular has done a good job of becoming the “one number” dashboard for DTC operators — blended ROAS, MER, new-customer CAC — and if you’re a Shopify-first brand doing $2M+ a year, that’s probably the right buy rather than build. But the moment you add Amazon or TikTok Shop as meaningful channels, the attribution math gets ugly and the “one number” starts to lie to you.
Where Helio differs from all of these is the deployment model and the scope discipline. Triple Whale and Helium 10 are SaaS — your data lives on their servers, the pricing scales with your GMV, and the feature surface grows whether you want it to or not. Helio is a self-hosted single-purpose tool you deploy yourself, with no database and no tracking, and the scope is set by one person’s opinion about what belongs on the screen. That’s a fundamentally different product philosophy, and for internal operational tooling — the kind of thing that shows you today’s units, not the kind of thing that advises you on strategy — it’s often the better one.
The “self-hosted, no database” decision is the interesting one
Most sellers I talk to recoil at “self-hosted” because it sounds like ops work. But read Rishi’s framing again: deploy your own copy to Vercel in five minutes, free, login stays on your server, no database, no tracking. That’s not a burden — that’s a security posture. For a seller whose Seller Central credentials, ad account tokens, and supplier POs are the crown jewels, a tool that never phones home is a feature, not a limitation. The no-database part matters too: it means there’s no state to corrupt, no migrations to run, no backup policy to write. The tool reads from the source system every time you open it. That’s the correct architecture for a dashboard whose only job is to reflect reality.
What cross-border sellers can borrow from this
Four things, in order of how quickly you can act on them.
First, the one-number rule. Before you build or buy any dashboard, write down the single number you actually open your admin for every morning. Not the number you think you should care about — the one you actually check. For most Amazon FBA sellers it’s units shipped yesterday. For most DTC brands it’s revenue and blended ROAS. For TikTok Shop operators it’s GMV and refund rate. Whatever it is, that number goes on the first screen, above the fold, with no login friction beyond what’s strictly necessary. Everything else is secondary.
Second, read the counter, don’t recompute it. Helio reads today’s units from the inverter’s own counter rather than deriving them from power events. The e-commerce equivalent is reading units from the marketplace’s own report — Amazon’s Business Reports or the settlement report, Shopify’s orders API — rather than recomputing from webhook events. Recomputing is where 90% of “my dashboard is wrong” tickets come from.
Third, pick a notification channel your team already lives in. Rishi’s daily Telegram summary is a small feature that does a lot of work: one tap to connect, one message every evening, no app to install, no push permission to negotiate. For cross-border teams that are already running ops in Slack or Telegram or WhatsApp Business, the same pattern applies. A daily 6pm summary of units, revenue, refunds, and stock-outs in the channel your ops lead already has open beats a beautiful dashboard nobody opens.
Fourth, ship the CSV export on day one. Helio exports every day’s data to CSV. That sounds trivial. It isn’t. The CSV export is what lets the tool coexist with your existing spreadsheet-based finance and inventory workflows instead of trying to replace them. Every internal tool you build should have a CSV export in v1, because the moment finance asks “can I get this in a spreadsheet” and the answer is no, the tool dies.
Why Amazon sellers should care more than Shopify ones
I said this above and I’ll say it again because it’s the most actionable point in this essay. If you’re Shopify-only, the platform gives you enough out of the box that the marginal value of a custom one-number dashboard is small — buy Triple Whale, move on. If you’re Amazon-first, or Amazon-plus-TikTok-Shop, or Amazon-plus-Temu, the platform gives you almost nothing that opens on the right number, and the analytics vendors are all built around research and PPC rather than daily ops. That gap is where a small self-hosted tool earns its keep. A single Vercel deploy that pulls yesterday’s units from Seller Central, yesterday’s GMV from TikTok Shop, and yesterday’s refunds from Shopify, and posts a summary to Slack at 6pm, is maybe a weekend of work for a competent engineer and saves your ops lead 20 minutes a day forever.
Where my judgment says it falls short
I like the philosophy. I’m less convinced the product as shipped is the right shape for anyone outside a very narrow slice of solar owners, and there are three specific things I’d push back on.
The hardware lock-in is a real ceiling. Eybond / ShineMonitor only, for now. Rishi is explicit that inverter brand determines what he supports next, which is honest, but it also means the addressable market is a fraction of the rooftop-solar population. For a cross-border seller reading this, the lesson is the opposite of what Helio does: if you’re building internal tooling, build it against the API surface you control, not against a vendor’s proprietary counter that could change without notice. Marketplace APIs are more stable than inverter firmware, but not by as much as you’d like.
“Self-hosted” is doing a lot of work in that pitch. Deploying to Vercel in five minutes is true for someone who already has a Vercel account and knows what a deploy is. It is not true for the median solar owner, who is not a developer. The people who will actually get value from Helio are a narrow intersection: solar owners who are also developers who are also willing to run their own infrastructure. That’s a fine market for a side project. It’s not a market that supports a company.
The “honest” positioning cuts both ways. No database, no tracking, login stays on your server — great. But no database also means no historical analytics beyond what the inverter itself retains, and no cross-user benchmarking, and no way to build the kind of cohort analysis that would actually tell you whether your panels are degrading faster than expected. The simplicity that makes Helio trustworthy also caps how useful it can become. That’s a legitimate trade-off, but it’s a trade-off, not a free lunch.
Where the math breaks
The savings estimate — today’s units plus a tariff the user sets themselves — is the feature most likely to mislead. A user-set tariff is only as good as the user’s understanding of their own utility billing, which for most people is a flat rate plus a time-of-use schedule plus a monthly connection fee plus a solar feed-in credit that varies by state and retailer. A single number can’t capture that. Rishi’s framing is honest about this — “a tariff you can set yourself” — but the output is presented as savings, and savings implies accuracy the input doesn’t support. If you’re building the e-commerce equivalent (say, a “profit per order” number), be careful that the inputs you’re letting users set aren’t the inputs that determine whether the output is true.
What I’d watch / test next
Three concrete things an operator can do this week.
One: write down your six-tap number. Open whatever admin you open first every morning, and count the clicks to the number you actually want. If it’s more than two, that’s your build spec. Hand it to whoever on your team can ship a Vercel deploy, and give them a week.
Two: audit your counter sources. For each channel you sell on, identify the one report that the platform itself treats as authoritative — the settlement report on Amazon, the orders API on Shopify, the GMV report on TikTok Shop — and check whether your current dashboard reads from that source or recomputes. If it recomputes, you have a reconciliation problem waiting to happen at month-end.
Three: stand up a daily summary in the channel your ops team already uses. Telegram, Slack, WhatsApp — doesn’t matter. Units, revenue, refunds, stock-outs, one message, 6pm local time. The Helio pattern — one tap to connect, one message a day, no app to install — is the right pattern for cross-border ops teams that are already drowning in notifications. Start there before you build anything more ambitious.
Helio itself is probably not the tool you’ll use. But the design brief behind it — one number, read from the source, self-hosted, no database, CSV out, daily summary in the channel you already use — is a better brief than most of the seller dashboards I’ve seen pitched this year. Steal the brief, not the app.
