A size quiz couldn't tell these systems apart.
They sell and install three home battery-backup systems, all in the same size class: wall-mounted and whole-home-capable. The obvious answer, a quiz that sorts by size, was useless here. Sorting by capacity lands almost everyone in the same bucket, without explaining what makes one system fit their home.
We sorted by the homeowner's real situation.
The strategy we landed on with their team was fit for your situation, never best product. No system would universally win, and their people helped define what should actually drive a recommendation.
- Their technical lead had written up the real differences between the three systems.
- Their team checked the build against the plan they'd approved.
- Their operations manager tested the live tool and found a buying situation we'd missed.
Understand: find the differences that matter.
With any tool built to help people choose between a client's own products, the first thing I do is pull the real technical specs on every system myself instead of relying on a generic quiz template. We landed on climate and altitude, the solar brand they already own, their EV plans, and room to grow their storage. An earlier call with their team locked that frame in place. Those answers tell us more about fit than asking how much of the house someone wants to run.
Design: check against the plan they approved.
The first build quietly drifted into a generic four-question quiz with flat scoring and no climate logic. It was marked done and reviewed, but the client caught the mismatch by checking against the approved plan. I rebuilt the questions and recommendation engine as a routing table, with rules read from data. That meant a later change could be a table edit instead of a rewrite.
The rules put climate before brand loyalty.
My build-and-integrate process combines AI-assisted drafts, checks against the real specs, and live browser testing. The battery selector uses an ordered set of rules, then explains its recommendation from the homeowner's answers.
Build: stress-test the recommendation.
Above roughly 9,000 feet, altitude takes priority: the tool recommends SigenStor even when the homeowner already owns Enphase gear. An AI-assisted draft got SigenStor's cooling method wrong, calling it fan-cooled when it's passive. It also put brand ahead of altitude; I caught and fixed both mistakes before the client saw that version. A separate review had already caught an earlier build that described a different manufacturer's product entirely.
- Check altitude first, then exclude systems that don't suit outdoor cold.
- Consider the existing solar brand after climate requirements are met.
- Use EV and storage-growth plans to break remaining ties; ordinary EV charging stays neutral.
Ship: let their team find what I missed.
I shipped to the same public link their team was already reviewing and verified every path in a real browser. Their operations manager found a gap: a non-Tesla EV owner who didn't want to power the home from it had no honest answer. I added a fourth EV answer, deliberately neutral so it wouldn't tilt the recommendation. The fix was verified live about a week after their team flagged the gap.
Extend: work inside the platform they use.
The same method runs beyond this one tool. On Shopify, I write changes through the platform's API, its connection for other software, and read them back from the live store. On WordPress and WooCommerce, I make those writes at the product level and check the live page every time. The work belongs inside whatever system already runs the client's business.
A recommendation their team could put to the test.
3 systems, 6 questions
Sorted by the homeowner's situation, with each recommendation path checked in a real browser.
1 climate-safety rule
High altitude overrides brand preference, routing an existing Enphase owner to SigenStor.
About a weekfrom their team's catch to the fix verified live on the same URL.
These are capability proof points, not measured sales results. If your team keeps answering "which one do I need," a fit-based tool can turn that conversation into something buyers can work through themselves.
What could this look like for your brand?
Let's talk about what your team wants to build and how I can help.
Let's Connect →