A campaign that takes cold Meta traffic, puts it in a room for forty-five minutes, and hands your team booked site assessments from homeowners who already understand what they are buying.
On 10 August we pulled the live Malaysian ad set from the Meta Ad Library. Around 210 active ads mention the SuRIA rebate. They are, with very few exceptions, the same ad.
When the message is identical, the only thing left to compete on is the number. That has already started:

Exhibit A. Meta Ad Library, Malaysia, active ads, 10 August 2026. Eight advertisers, one message. Most of them end at the same place: a WhatsApp button.
Why this matters more this year than last. The rebate has a hard end date of 31 December 2026, and a fixed RM150 million pool behind it. Every advertiser knows it. Between now and December the bidding gets more crowded and more expensive, not less. A price war on a deadline is the worst possible fight for whoever has the higher build quality.
You are running roughly nineteen ads across seven different messages: the instalment plan, the Merdeka package, the ATAP explainer, the 95% bill cut, the rebate tiers, a brand video, and after-sales protection. Every one of them ends in "contact us now" or a WhatsApp button.

Exhibit B. Your live set, same day, same source.
Your newest ad asks: what costs are covered if the system needs repairs, is labour covered, will the provider still be there in ten years. Behind it sits a four-hour response commitment, onsite inside two business days, Tier 1 BloombergNEF batteries and a nine-step delivery process.
No competitor in the set we pulled is saying any of that.
That argument needs about four minutes to land. A Meta feed gives it about three seconds, next to a competitor offering RM161 a month.
Quality arguments lose in a scroll and win in a room. Everything that follows is built to get your buyer into a room.
Every ad in this category asks a stranger for their phone number so a salesperson can call them. Buyers know exactly what that costs them, which is why the price has to keep dropping to buy the click.
"Come and learn what actually moves your bill" is a far cheaper yes — and it is the only format with enough room to make your case.
Ryan's instinct was right. The research backs the webinar. What we would change is where it aims: teach the bill first and let solar be the conclusion, rather than teaching solar to people who have already decided.
Forty-five minutes, live, free. Built so the first four segments are useful to someone who never installs anything, and the fifth is the reason they choose you.
The price war works for you here. By segment five, every cheap competitor ad the attendee has already scrolled past becomes evidence for your case rather than against it.
The fast lane matters. A homeowner who is already decided should never be made to wait three weeks for a webinar. The lander and the confirmation page both carry a direct booking link.
Only one of them mentions the rebate, and it reframes it from claim this to understand this. None of them asks for a phone number so a salesperson can call. Three languages, because a large part of the live competitor set advertises in Malay and Chinese.






Mockups. Brand colours are sampled from plusxnergy.com and the wordmark is set in type — the live build would use your own brand assets. Every claim on these is one you already make publicly, except the tariff mechanics, which are from the July 2025 restructure.
A corporate site has to serve everyone: homeowners, commercial buyers, investors, job applicants. A campaign page only has to do one thing. That is the whole difference, and it is why this sits alongside your site rather than inside it.

Built mobile-first. Most of this traffic will never see a desktop.
Name, WhatsApp number, email, property type, monthly bill. Property type and bill size are there so your team knows who is worth an assessment before they pick up the phone.
It does not ask for an organisation name, a billing address, a subject line or a message, and there is no CAPTCHA in front of it. Everything asked for either delivers the join link or qualifies the lead.
The page also disqualifies people, out loud: under RM200 a month, renters, high-rise without roof access. Turning away the wrong registration is cheaper than sending your team to the wrong roof.
Open it on your phone. The form validates, rejects a bad number, and confirms. Live, it posts to your CRM and fires the Meta registration event so the ad account can optimise against real signups.
Everything up to here has been about solar. This one section is not. It is the last campaign we built, in a different industry, and it is here so you can judge whether we can actually run what we just described. Solar resumes in section 09.
In July we ran this exact sequence for a management training programme: cold Meta traffic, a registration page, a live room, a booked appointment. Below is that ad account and that CRM, unedited. Read it for the mechanism, not the industry — the buyer and the price are different, the machine is identical.

Exhibit C. Meta Ads Manager, 21–28 July 2026. Twenty versions ran. Most were switched off inside 48 hours; one produced a single lead at RM49.66 and was stopped in two days. Finding the winner cheaply is the work, and this is what that looks like from the inside.
Two angles ran side by side. Same week, same audience, same registration page. One named a problem only the buyer feels. The other described what the product does. The difference was not the cost per lead. It was who answered.
The winner was cheaper, produced three and a half times the volume, and nearly doubled the share of people who could actually sign. That is the whole job. It is also why cost per lead on its own is a poor score, and why we report booked appointments instead.
This is the transferable part. Meta offers no income or job-title targeting in Malaysia, so audience settings alone cannot find a homeowner who can afford a system. The creative has to do the filtering. For you that means stating the conditions in the ad itself — landed property, a bill above a named figure — which costs reach on purpose.
79 of 130 registrations attended live — 61%, against a 20 to 40 percent norm for a free webinar. Job role was a required field at registration, so every attendee was graded one by one.

Exhibit D. 47 of the 79 people in the room could authorise a purchase, including 28 owners. Seniority did not reduce attendance: owners turned up at 62% and directors at 63%, against 61% overall.

Exhibit E. Minute-by-minute attendance. The room peaked at 75 people at 8:31pm and still held 96% of them half an hour later, 79% an hour after peak. The session ran two and a half hours. People stayed — which is the only reason a forty-five minute argument about build quality is worth making at all.
That was a management programme at a management programme's price. What it tells you is that the sequence works and that we can run it — not what a solar lead will cost. The next section is our estimate for your campaign, built from scratch and modelled harder than the numbers above.
The middle column is the case we would plan against. The outer two are the honest edges. Every registration figure here is modelled worse than the RM17.91 we actually achieved, because solar is a crowded auction and your first month is not our fifth.
| At RM5,000 ad spend | Cautious | Planning case | Good |
|---|---|---|---|
| Cost per registration | RM40 | RM30 | RM20 |
| Registrations | 125 | 167 | 250 |
| Turn up live (we hit 61%) | 40% · 50 | 48% · 80 | 55% · 138 |
| Book an assessment | 20% · 10 | 25% · 20 | 30% · 41 |
| Fast lane, skipping the class | 6 | 8 | 12 |
| Booked site assessments | 16 | 28 | 54 |
| All-in cost each (fee + spend) | RM729 | RM418 | RM220 |
Total cost for the month: RM11,850 — RM6,850 fee plus RM5,000 of ad spend.
| At RM10,000 ad spend | Cautious | Planning case | Good |
|---|---|---|---|
| Cost per registration | RM40 | RM30 | RM20 |
| Registrations | 250 | 333 | 500 |
| Turn up live (we hit 61%) | 40% · 100 | 48% · 160 | 55% · 275 |
| Book an assessment | 20% · 20 | 25% · 40 | 30% · 83 |
| Fast lane, skipping the class | 12 | 17 | 25 |
| Booked site assessments | 32 | 57 | 108 |
| All-in cost each (fee + spend) | RM518 | RM297 | RM157 |
Total cost for the month: RM16,850 — RM6,850 fee plus RM10,000 of ad spend.
Doubling the budget does not double the cost. Our fee is fixed, so it spreads across more appointments: RM418 each at RM5,000 of spend, RM297 each at RM10,000. That is the argument for starting at the top of the range rather than the bottom, and it is the only reason we would push for it.
Read this as a range, not a promise. The nearest real Malaysian anchor we have puts cost-per-appointment on a comparable funnel between RM228 and RM449, and that figure excluded management. The first thirty days exist to replace this model with your numbers. We report the real ones whether they flatter us or not.
This adds a lane. It does not take anything over.
Your marketing manager keeps the brand. Positioning, tone, product truth and final approval stay in-house. What we take off their desk is the part that eats a week every month: building creative, rebuilding pages, chasing reminders and pulling numbers.
One system covers the month.
On the planning case of 28 booked assessments, that is a close rate of about 3%. At RM10,000 of spend it is 57 assessments and about 2%.
RM15,000 is your own published Merdeka figure and it is revenue, not margin — your gross margin decides what the rest is worth. The point is only the height of the bar. Your team will know whether closing three in a hundred attended assessments is a low one.
Ad spend is yours and goes straight to Meta. We never mark it up and you hold the account, exactly as you do now.
Three-month minimum. A single month cannot produce a reliable cost per appointment in a vertical this competitive. We would rather say that now than explain it in week five.
Everything we build belongs to you — pages, creative, sequences, audiences and the registration list. If we stop working together, none of it leaves with us.
The fee does not move with spend. Scaling the budget improves your cost per appointment rather than ours.
Invoiced by Fastcom PLT (202204000698 · LLP0030970-LGN).