CASE STUDY · CONSUMERS ENERGY

Solar

Gardens

Solar

Gardens

Building the missing front door: a mobile-first conversion redesign of a community solar program.

ROLE — Researcher & Designer

TIMELINE — 14-week co-op

SCOPE — Research, IA, content strategy, prototype

Landing page — redesign

Experience page — redesign

TL;DR

The short version

Solar Gardens lets Consumers Energy customers subscribe to a share of a community solar installation and receive a credit on their bill — built for people who want renewable energy but can't put panels on their own roof. The existing site spoke only to visitors who already knew what community solar was, so I rebuilt it as a mobile-first conversion path: a demand-generation landing page, a storytelling page about the gardens, and a separated enrollment flow. Solo researcher-designer on an eight-week co-op sprint, working from a live-site audit, stakeholder interviews, and FullStory funnel data.

THE BRIEF

The Brief

The live Solar Gardens landing page — current site

The brief said redesign. Research said something else.

RESEARCH & METRICS

Research & Metrics

33.2K

Total visitors

93%

Arrived on mobile

165

Mobile visitors reached Choose Subscription

0

Recorded progressing past it

½ kW increments and program jargon, unexplained — current site

Mobile isn’t a segment. It’s the product.

Current site on mobile

Pricing led with $8, and ‘No Contracts or Cancellation Fees’ ran against the program’s one-year commitment — current site

Customer journey map — the awareness stage exists in the journey but had no counterpart on the site

THE STRATEGIC SHIFT

The Strategic Shift

The site was written for people who already knew what community solar was. Almost nobody arriving at it did.

The program's target customer is someone who supports renewable energy but can't install panels — a renter, a condo owner, someone with the wrong roof. That person isn't searching for "community solar subscription." They don't know the category exists.

The existing page assumed awareness and jumped straight to configuration: pick a block, here's the price, here's the form. Every problem I documented was a symptom of that one missing layer. **The site had no demand-generation stage** — it was a signup page wearing an information page's clothes.

That reframe promoted the work from a visual refresh to a conversion redesign, and produced the structural decision everything else hangs on: the site was doing three jobs on pages that couldn't separate them.

THE SOLUTION

Three pages, one job each.

Landing Page — demand generation

Establishes that a category exists, that this person qualifies, and that it costs less than they assume, before any configuration decision appears. Block sizes are framed in appliances rather than ½ kW increments — the number stays accurate, the reference point becomes something the customer already owns. RECs are explained as a receipt: documented proof the clean energy was generated and that it's yours. The one-year commitment and the absence of a cancellation fee appear in one sentence near the price, replacing a "no contracts" claim that collapsed under scrutiny. Pricing leads with the microblock, so the cheapest way to say yes is the first thing you see.

Pricing — current site

Pricing leads with the MicroBlock — redesign

Experience the Solar Gardens — belonging

One primary CTA into the Experience — redesign

Deliberately not educational; that job belongs upstream. Four garden locations, each with its own narrative hook rather than interchangeable cards with the megawatt figure swapped out.

Experience page hero — redesign

GVSU garden — its own story — redesign

Cadillac garden — brownfield comeback — redesign

Enrollment — separate workstream

Scoped apart from the start so an unmovable technical constraint couldn't hold the two pages I could ship.

Built as a fully responsive Figma prototype using auto layout with Fill/Hug behavior and a single Fit Width flow — rebuilt properly rather than patched, so the mobile-first argument was demonstrated rather than asserted.

Validation was structured as two-phase unmoderated think-aloud testing via UserTesting: a pilot of 2, then a full round of 6–8. Task scenarios were written directly against documented pain points, so findings map back to specific decisions instead of general impressions.

DESIGN VALUES

Design Values

Mobile is the product.

Not a breakpoint. At 93% phone traffic, the desktop layout is the adaptation.

One page, one job.

A page that educates, persuades, and converts does all three badly.

One primary CTA per decision zone.

Zones are defined by where a visitor has finished absorbing information, not by layout sections. Wherever someone has just learned enough to act, there's exactly one thing to do next.

Explain it in the customer's units.

Appliances, receipts, dollars on a bill — not kilowatts, rate classes, and certificates. The translation isn't dumbing down; it's doing the work on our side of the screen instead of leaving it on their

Say the hard part out loud

Every instinct in marketing copy is to soften the one-year commitment. But vagueness was itself a documented pain point — it was costing more trust than the honest version would.

BUSINESS CASE

Business Case

The funnel gap sits at Click and Engage — visitors arriving and leaving before enrollment. That's a conversion problem on existing traffic, cheaper to solve than an acquisition problem, and at 93% mobile the improvements compound on the majority segment rather than a minority one. The new garden has fixed subscription capacity, so every point of conversion lift has a defined ceiling of business value — which makes the projection bounded and credible.

Retention is the second lever. The cancellation contradiction and the unclear bill impact are plausible contributors to churn among customers who enrolled with mismatched expectations; fixing expectation-setting at the top of the funnel is also a retention intervention.

Measurement: 30/60/90-day checkpoints against the mobile funnel baseline established pre-launch, with new instrumentation for the Click and Engage stages current funnels don't cover.

[Placeholder — disclaimer line, verbatim.]

WHAT I’D DO NEXT

What I’d Do Next

Run the study

Pilot, then the full round, synthesizing findings in three separate streams: funnel metrics, baseline observations on the live site, and design validation on the new screens. Keeping them separate is what keeps the argument honest; n=2 is directional evidence, not a conclusion.

Build the bill impact estimator

The single largest unanswered customer question. Even a rough range beats silence.

Instrument Click and Engage

Before launch, so the checkpoints measure the stage the redesign actually targets.

Design the exit and win-back path.

I studied why people don't join; why they leave after a year deserves its own research.

REFLECTION

Reflection

The hardest part wasn't design. It was the stretch between finding the problem and getting agreement that it *was* the problem. I was handed a redesign brief; what the research supported was a claim that the site was missing an entire stage of the funnel — a harder thing to bring to a stakeholder than a set of mockups. What made it land wasn't the argument itself, but that every decision I presented traced back to a documented pain point, a persona need, or a number in the funnel. Nothing was there because I liked it, and that traceability turned a subjective conversation into a reviewable one.

I also learned how much design at a regulated utility is translation. The tariff, the rate classes, the one-year commitment — none of it was mine to change. The available move was to explain it accurately in language a person actually uses, which turns out to be most of the job.

What I’d do differently

Instrument first. I built the FullStory dashboards after the content audit, and the funnel data reframed decisions I'd already made. Starting with the numbers would have gotten me to the demand-generation insight faster and with more evidence behind it on day one.