Santo Domingo Doesn't Speak “LATAM”

Client:

Treinta

Campaign:

Fintech / SMB SaaS — Paid Subscriptions

Location:

Dominican Republic

Year:

April 2026

Platforms: Instagram Reels + TikTok · 7 creators · 14 pieces · Scope: casting → production → whitelisting

01 — The Premise

The buyer keeps her books in a notebook. She is not waiting for software.

Treinta sells a business-management app to owners of colmados, barbershops, food trucks and clothing stalls. The category problem isn’t awareness — it’s that the buyer already has a system she trusts: a notebook, the notes app, her own memory.

So the brief we accepted wasn’t “make people aware.” It was: make a merchant recognise her own Tuesday afternoon on screen, then give her one thing to do about it. Awareness content in this category converts at roughly nothing, and everyone involved knew it going in.

02 — Why Localisation, Not Translation

Latin America is not an audience. It’s a shortlist of separate ones.

The same product, the same objective and the same creative format produce different results in Santo Domingo, Bogotá and Lima — because the merchant’s vocabulary, humour and reasons for distrust are different. We treat every country as a separate casting and messaging problem.

• Language: Dominican register, not neutral Spanish. “Ten el control de tu plata,” never “optimice sus procesos.”
• Audience geography: creators screened on in-country follower share, not total reach. A 1.5M profile with a diaspora-heavy audience is a worse buy than a 180K local one.
• Objection to beat: “I already know what I sell.” The script has to lose that argument on screen before the CTA lands.

03 — Casting

Cast for performance, not for follower count.

Our casting is run by producers with Hollywood and European film and TV backgrounds, so we audition creators the way you’d audition actors: can she carry a scripted scene, hold comic timing, and still demo a product without the piece turning into an ad read?

The roster mixed two profile types — acted comedy with high engagement, and niche entrepreneur voices with credibility inside the merchant community. Every creator on this campaign was a real person with a real audience; we don’t use AI-generated or synthetic creators.

04 — How It Ran

The unglamorous half is where campaigns are lost.

Casting gets the credit. What actually decides whether a campaign is billable and amplifiable is the operational sequence around publication — codes, disclosure, permanence, tracking. Ours is fixed and written into every creator contract.

• Week 0 — Setup: before a single frame — per-creator tracking links, contracts, brand brief and creator brief. Nothing is shot before the paperwork closes.
• Week 1 — Script, approval, shoot: concept approved in writing before filming; drafts inside 5 days; two revision rounds, 2-business-day response windows on both sides.
• Week 2 — Publication: with paid-partnership tags live. If the tag can’t be switched on, the creator does not publish — she sends screenshots and we fix it first.
• Week 2 — Amplification codes collected: Spark Ads and Instagram partnership authorisations delivered within 5 business days of each post, with 60 days of paid-amplification access.
• Week 3+ — Permanence watch, 6 months: every piece stays live, public and unaltered for six months from publication. Contractually ours to enforce, monitored after the campaign closes.
• Close — Reconciliation report: deliverables, links, content metrics and every documented deviation, filed within 10 business days of campaign end.

One rule that saved the campaign’s ad spend

If a creator publishes before activating the paid-partnership tool, the platform can lock that tool on the post — sometimes for months. The content stays up and looks fine, but it can never be amplified, which quietly deletes most of its commercial value. Our protocol is therefore absolute: no active tag, no publication. On this campaign it meant holding one creator’s post for several days rather than losing the amplification window entirely.

05 — Delivery

What we shipped, and who measured what.

Inside those interactions: 9,308 saves and 5,102 shares — the two signals that matter most for a product a merchant has to think about before paying for it. Brand-name searches appeared among the traffic terms bringing people to the content, which is curiosity the campaign created rather than reach it bought.

Attribution note

Installs, registrations and subscription purchases were tracked entirely on Treinta’s side through their own MMP and per-creator links. We report on what we own — deliverables, content performance and amplification access — and we don’t publish conversion figures that aren’t ours to publish.

06 — The Work

The campaign, as it ran on the merchants’ feeds. Published pieces, embedded live from the creators’ own profiles — the same posts that are contractually required to stay public and unaltered for six months.

07 — The Platform Split

TikTok bought the volume. Instagram bought the market.

Running both platforms per creator turned the campaign into a controlled comparison, and the audience data separated cleanly. On Instagram, 96–99% of the audience on each Reel was inside the Dominican Republic. On TikTok, the same creators reached roughly 62% in-country and 29% United States — the Dominican diaspora.

That isn’t a reason to drop a platform; it’s a reason to brief them differently. TikTok delivered the incremental reach and the highest single-piece jumps. Instagram delivered the audience that can actually download a Dominican app store link and subscribe. For a market-entry pilot, the qualified half is the half that decides whether the next campaign gets funded.

• Instagram Reels: 96–99% in-market. Women 57–80% of the audience, 25–34 dominant — the merchant demographic itself.
• TikTok: ~62% in-market, ~29% US diaspora. Younger skew, and the biggest volume spikes of the campaign.
• Stories: not a reach buy. Used as the direct-click layer — the only placement in the mix with a tappable link and brand sticker.

08 — What We Took From It

Honest notes, including the ones that cost us money.

Platform tooling is a launch dependency, not a formality

Branded-content tags belong in the pre-publication checklist with the same weight as the script approval. We now verify the tool on the creator’s account before the shoot, not on publication day.

Substitutions should be priced before they’re needed

Two slots on this campaign were replaced mid-flight — one of them covered by two creators instead of one. The decisions were sound, but they were made in hours. Pre-approved alternates per profile type keep that from becoming a negotiation under deadline.

Managers need the whitelisting explanation twice

Creators routinely hear “whitelisting” as “you must run ads yourself.” A one-paragraph plain-language explainer at contract stage removes a week of back-and-forth per creator.

If the client owns attribution, agree the reporting boundary on day one

It protects both sides: we’re accountable for content and delivery, they’re accountable for the funnel their MMP measures — and nobody argues about results in the last week.