how we got here

we met in 2012.

coursemates first, then partners on every assignment, then the same first job out of school, then years building in the same companies. somewhere in there it stopped being two careers and became one. we're partners in work and in life. in 2021 we made it official and started pebble.

the through-line, the whole time, was small businesses. we spent years building software for malaysian smbs — custom work, one client at a time. clinic owners, distributors, service shops, sellers. we sat in their actual operations and watched how the day really goes.

and we kept seeing the same thing break.

the pattern, everywhere

small businesses don't fail because the founder isn't capable. they fail because repetition outpaces the team. the same questions get asked every day, and staff answer them differently depending on who's around.

the real knowledge lives in the founder's head and their whatsapp — scattered across chats, a few spreadsheets, and whatever someone happens to remember. nothing's written down in one place because there's never time to write it down.

so the business can't grow without the founder sitting in every loop. they become the bottleneck in their own company. not for lack of effort — the opposite. they're doing too much of the work only they can do, because they're the only place the answers actually live.

the clearest version of it happened to me as a customer.

the night it clicked

i was buying a teacher's day gift for my kid — last minute, the night before. found a shop i liked, couldn't find the promo on their site, so i messaged them. 10:54 pm. auto-reply: away, back during office hours.

the actual reply came the next morning. 10:58 am. twelve hours later. by then i'd already bought somewhere else.

here's what got me: the answer existed. the promo was real, the link was real, the order could've been placed at 10:54 pm. something that simply knew their own offer would've closed it on the spot. instead the sale just evaporated. nobody did anything wrong — there just wasn't anyone awake, and everything they needed to answer me was sitting somewhere nobody could reach at that hour.

i'd watched clients lose the exact same way for years. that night i was on the buying end of it, and it was obvious: this isn't a people problem. it's a gap nothing was filling.

for years we knew what we wanted to build. we just couldn't build it right.

the tech wasn't ready
the old way you'd hand-write rules for every business, every question, every branch — and it still fell apart the moment a customer phrased something slightly differently. it didn't understand anything. it matched keywords and hoped.
the catch to make it work at all, you'd have to drag the business onto your platform, migrate their data, change how they operate. the cure was worse than the problem. we weren't going to ship that.
what changed the models. they can finally read a business's own knowledge and stay inside it — answer like someone who actually works there, without us scripting every path. that's what makes the product tight enough to slip into how a business already works, instead of demanding they rebuild around it. no big migration. it just fits.

so we built nudge. it answers the repetitive questions, within what the business has told it, and hands off to a human the moment judgment is needed. it sits in the chat tools they already use. the owner gets their day back — and the 10:54 pm message gets answered at 10:54 pm.

the product is new. the conviction took nine years. that's the order we wanted it in.

stay lean. stay sharp.

nudge is the first. there'll be more. → read what we believe