Audience research
Segments by interest, behaviour and lookalikes — built from your existing customers, not guesswork.
Campaigns on Instagram, Meta and VK for products people are not searching for yet. Demand is built by interest and behaviour — creative does the heavy lifting, and it is treated as such.
Segments by interest, behaviour and lookalikes — built from your existing customers, not guesswork.
Static and video variants designed for the feed, not cropped from a website banner.
Cold, warm and returning traffic get different messages and different offers.
Correct event tracking so the platform optimises toward sales rather than clicks.
Creatives and audiences tested in pairs, so it is clear which of the two moved the number.
Separate campaigns for cart abandoners, video viewers and site visitors.
Who buys, why, and what would make them stop scrolling.
A first batch of six to ten variants for the opening test.
Pixel, events, catalogue and campaign structure.
A small budget spread across hypotheses to find the working combinations.
Budget shifted to winners, losers cut, new variants queued weekly.
Yes — design and copy are included. If you have brand assets or a videographer, I work with what you have.
Six to ten for a first round. Fewer than that and there is nothing to compare.
Sometimes. For narrow B2B niches paid search usually pays back faster — I will say so at the brief stage rather than take the budget.
I set up a proper business account structure and keep a backup. It reduces the risk; nobody can eliminate it.

Website, Google Ads and Instagram for a Manhattan nail salon: 1,170 online bookings and $128K in revenue in 10 months.

Ads, a website and an online store for a Russian manicure salon chain in New York: 778+ bookings in a year.

Meta and Google Ads for a chain of buccal facial massage studios in New York: 4,370 conversions at a $0.48 click.

Meta ads for a New York nail salon: 489 messaging leads at $3.31 each in one of the most expensive markets.