After 6 months of building solo from Japan 🇯🇵, I'm launching SplanAI on Product Hunt in 5 days.
It's an AI sales assistant for the 200K+ small American home builders who can't afford a CRM or sales team.
→ Auto-qualifies inbound leads in 60 seconds
→ Drafts personalized follow-ups in English & Spanish
→ Pulls MLS data so builders see lot pricing & comps in one dashboard
7 builders in beta. Paying customers from Texas and Florida.
Why home builders? They're losing 60-70% of leads — not from bad product or pricing — but because no one follows up fast enough. And no one is building for them at $49/mo.
May 26 on @ProductHunt. Follow along for the launch ride 🚀
🔗 https://t.co/m9OfTFU8Q8
A quiet lead doesn't always mean a buyer hasn't decided to build. Sometimes they already decided -- and they're quietly comparing which builder gets there first, without ever mentioning it.
A custom home cost estimate is quietly conditional -- on today's material costs, an unlocked scope, a lot not yet fully walked. Almost nobody states how long it holds. So when it moves, it doesn't feel like a window closed. It feels like the number was never real.
A home builder proposal with three concepts is real progress. One with no visible price difference between them just moves the guessing from which house to which one I can afford -- and most buyers won't ask that question out loud.
A new CRM will show you, for the first time, exactly how long a lead has been sitting at "send them a concept." That's real visibility. It did not make producing that concept any faster. The stall you can finally see is still the same stall.
A custom consult often ends with "think it over and get back to me." No deadline, no format -- a blank page handed to the buyer. The silence after gets read two ways: the builder thinks the buyer went cold. The buyer thinks the builder moved on. Same silence, both wrong.
A buyer skimming your site before they ever call isn't grading craftsmanship. They're counting photos. A production builder's gallery runs forty deep. A solo custom shop's runs eight. That gap reads as inexperience before either builder says a word.
A lot that fails feasibility and a builder who just didn't want the job produce the exact same sentence to a buyer: no. Nothing distinguishes a careful rejection from a brush-off once it's only spoken, never shown.
The lead who emailed again this morning gets the reply. The lead who went quiet three days ago, closer to deciding, doesn't. Recency isn't readiness -- it's just the easiest signal to notice on a busy day.
A custom home cost estimate almost always goes out before any design exists to go with it -- a number with nothing next to it yet for a buyer to judge it against. The number usually isn't the problem. Having nothing to attach it to is.
Walking five model homes in one afternoon isn't just comparing floor plans -- a buyer is absorbing "this company has range" as a separate impression. A small custom builder never had a row of models to walk. There's one lot, one meeting, one static picture.
A home builder proposal shows the design a lot allowed. It almost never shows the zoning, setbacks, or permit limits that actually did the allowing -- so when one of those shows up later, it reads like a different house, not a technicality.
A builder reads a quiet lead as price. Often it's a closing date instead -- a home that hasn't sold yet, a lease that doesn't line up with the build timeline -- and the buyer never thought that was the builder's business to know.
A spec sale's signature ends the selling. A custom sale's signature just starts a new round of it -- allowance decisions, change orders, substitutions are small sales too, and they rarely get the structure the first one did.
A buyer never learns a builder's headcount. They infer it from whether the PDF, the text, and the number said out loud on a walkthrough actually agree. The mismatch isn't carelessness -- it's what happens when only the buyer ever sees all three side by side.
Two different people type "custom home proposal template" into Google. A builder wants a faster tool. A buyer, already comparing two or three builders, wants a format they can hand to each one. Almost everything written for that phrase only answers the builder.
A lead that goes quiet for a week after a good meeting usually isn't dead. It's often the buyer walking the lot again on their own -- checking things nobody invited them to check -- before they say anything out loud. Silence reads as decline. Most of the time it isn't one.
A wide custom cost estimate is often the honest number -- the lot isn't finalized, the floor plan isn't locked, so the range reflects what's genuinely still open. A buyer doesn't read it that way. They read the width as the builder not knowing the job.
A pricing page without a demo-call gate still has to decide who it's priced for -- and a lot of "best software" roundups answer with a seat minimum: three or five seats before the per-user rate even applies. Read by one person. Billed like five.
A feasibility check answers whether a lot works under this week's numbers -- the buyer's budget, the listing price, the permitting rules. None of those stay fixed. But the "yes" from a few weeks ago usually still gets treated like it is.
A spec sale has one owner from the model-home walkthrough to the signature. A custom lead gets passed between sales, design, and whoever builds the estimate -- and the buyer ends up repeating themselves to someone who should already know what they said.