How and why we built Upcast.Social
We needed a marketing function and did not want to buy one. So we put the idea through our own validation tool first, built what the evidence supported, and ended up with a product other people pay for.
Why
We are a small company that builds AI products, and marketing them is real work nobody here is employed to do. Hire someone, retain an agency, or buy one of the established tools — each option has a number attached that recurs monthly whether the work gets done well or not.
There was a second problem. We tell clients not to spend on AI until they can say what it is worth, which is hard to keep saying while quietly buying a subscription because it felt easier than thinking about it. So we put the idea through Opptora first.
The verdict was not obvious. Building a social tool to save the cost of a social tool is how teams end up maintaining a worse version of something they could have bought. What justified it was narrower: posting is the small part, and what was actually costing us was producing enough decent, on-brand material without a marketing function.
What
Upcast.Social is social media management for people without a social media team.
- Daily research and drafting — every day it researches what is happening in your market and writes posts about it in your brand's voice, not the generic one every AI defaults to. This is the part we are most pleased with, and it runs without being asked.
- Campaigns, not one-off posts — related material is planned, generated and reviewed together.
- Approval before anything publishes — per post or for a whole campaign.
- Scheduling with retries — because networks refuse things.
- Several brands — a company and its products each need their own voice.
- The cost, shown up front — what the AI work cost, before you approve it.
That last one was not a feature request. A tool that spends money on your behalf without showing you how much is precisely the thing we spend our working lives warning people about.
From internal tool to revenue line
Our problem turned out to be the normal case rather than the exception. Upcast became a product other people pay for, and then grew a reseller programme — agencies run their own clients through it under their own account.
That is a strange arc for internal tooling, so it is worth being honest about where the value came from. Not from clever software: from the fact that the alternative had a knowable cost, so we had something to compare against. The revenue was a bonus we did not underwrite the decision on.
What it's like for the people using it
No media team needed
"I never had the time or budget for a social media team, so most platforms just got ignored. Upcast.social changed that completely. I jot down a few rough ideas and it helps me turn them into professional content, schedules it, and posts it across every channel. What felt impossible is now quick, streamlined, and actually enjoyable."
Consistent social without the effort
"Running a food brand, keeping up with Facebook and Instagram felt like a full-time job on its own. I'd post when I remembered, but it was never consistent. Upcast.social changed that — I give it a few ideas about our products and it helps me craft polished posts, schedules them, and publishes across both platforms. What used to fall to the bottom of my to-do list now just happens."
The interesting number is never the invoice that stops arriving. It is the work that stops needing to be done.
Three things carry into client work, none of them about social media.
- Measure the thing you are replacing before you replace it.
- Put the cost of the AI work in front of whoever authorises it, inside the product.
- Be suspicious of build-versus-buy arguments resting on licence fees.