A visual system built to survive a small team running it alone.

The founder's Instagram grid was the usual pre-agency mix: a launch photo, a Diwali post, a repost from a customer, a screenshot of a review, in no particular order and no particular voice. The product itself was genuinely good (a small-batch snack with real repeat buyers), but nothing on the shelf or the screen said so.

We didn't start with a moodboard. We started by asking who would actually be making the next hundred posts after we left, because the honest answer (a two-person team with no in-house designer) decided almost every choice that came after it.

That ruled out a lot of what usually passes for 'branding': no identity that depended on a specific photographer, no template that needed Figma to edit, no six-page brand guideline nobody would open twice.

What we built instead was closer to a kit than a brand book: a locked type and colour system, a packaging structure that worked across five SKUs without a redesign each time, and a set of Canva-editable content templates built around the actual questions their customers asked in the DMs.

The pushback we gave the founder early was the one that mattered most: drop two of the four content pillars they'd inherited from a previous freelancer, because a two-person team maintaining four pillars well was going to maintain none of them well.

Handover was not a Notion doc and a goodbye call. We sat with their in-house person for two working sessions, watched them build a week of content unassisted, and fixed the friction points before we called it done.

Six months on, unprompted, they were still posting on-system (same fonts, same grid logic, same packaging structure) without us in the loop. That was the actual brief, even though nobody wrote it down that way: not a launch we could screenshot, but a habit a small team could keep without us.