TT Ads

How we approach the work

Beauty / DTC · hypothetical playbook

Strong on Meta, new to TikTok

Situation: A DTC brand with working Meta creative and clean Shopify tracking wants a TikTok presence without burning a quarter learning the platform.

How we would run it: We do not port Meta ads. Week one is Pixel plus Events API verification against test orders, because scaling decisions made on unverified signal are guesses. Creative starts with three native formats (creator-style demo, before/after, founder voice) at low spend, and nothing scales until a format has beaten the account average twice. Meta assets get used one way only: as a message map of what already persuades this audience.

First milestone: A creative testing rhythm the brand can sustain in-house, with tracking a finance person would sign off on.

Gaming / app install · hypothetical playbook

CPI creeping after signal loss

Situation: A mobile app whose install costs rose after platform privacy changes, with retargeting audiences that quietly stopped filling.

How we would run it: First we separate measurement problems from performance problems — post-privacy, a chunk of any CPI increase is often attribution undercounting, so we reconcile ad-platform numbers against MMP data before touching campaigns. Then hooks: install campaigns fatigue creative faster than commerce, so the plan is a weekly hook-refresh cadence built from the top-performing angles, not a monthly big-bang creative drop.

First milestone: A CPI number everyone trusts, then a fatigue-resistant testing cadence against it.

Food & beverage / catalog · hypothetical playbook

Catalog campaigns underdelivering

Situation: A regional brand whose shopping placements are suppressed by product feed errors nobody owns.

How we would run it: Feed first, ads second. Missing identifiers, mismatched availability, and disapproved items cap delivery before any optimization matters, so the opening sprint is a feed audit with a named owner and an error budget. Only when the feed passes cleanly do we restructure campaigns — usually into something simpler than the account inherited, because catalog complexity mostly multiplies error surface.

First milestone: A clean feed with monitoring, and campaigns simple enough to diagnose when something breaks.

About this page

Why are there no client names or ROAS numbers on this page?

Because we will not publish results you cannot verify. Agency case studies with unnamed clients and unverifiable multipliers are marketing fiction, and the industry has plenty already. These playbooks show how we think; named case studies with dates and checkable figures will replace them as clients approve publication.

Can you share real examples privately?

Yes — on an intro call we can walk through relevant work in your vertical, including what did not work, under whatever confidentiality the situation needs. Ask hard questions; the difference between an agency and a deck is whether the answers survive them.

What if my situation is not one of these three?

The three playbooks cover the most common inbound situations, but the spine underneath them is constant: verify measurement before optimizing, test creative in structured rounds, and fix upstream data problems before spending against them. The process page describes that spine in full.

Want specifics for your industry?

Bring your account (or your plan for one) to an intro call and we will talk through it concretely — including whether TikTok is the right next dollar for you at all. Get in touch.