Case Study: How TripleTen Cut Ad Setup Time 5–7x with Make.com and an AI Creative Intelligence Pipeline
Client: TripleTen
Industry: Education & Research (Online Tech Education)
Scale: 51–200 employees | 1,000+ creative assets produced annually | Meta, YouTube, TikTok, and Google Ads
Tech Stack: Make.com, Google Gemini AI, Google Sheets, Meta, TikTok, Google Ads, Apify, Slack
Executive Summary
TripleTen, an online tech education provider, runs a high-volume performance marketing operation — testing hundreds of creative assets per month across Meta, YouTube, TikTok, and Google. Manual processes for uploading creatives, monitoring competitors, and analyzing what actually drove performance were consuming hours the team needed for strategy.
TripleTen worked with Vatech.io to build three interconnected Make.com scenarios, including a Make AI Agent, that automate the full creative lifecycle from upload to analysis. The result: ad setup time dropped by up to 7x, creative labeling of 1,000+ assets per year went from weeks to seconds, and the user acquisition team freed up hundreds of hours per month to focus on strategy instead of repetitive tasks.
"Given all the results, we will definitely use Make for all our other tasks connected with automation." — Maksim Epifanov, User Acquisition Lead at TripleTen
The Challenge: Three Bottlenecks at Scale
As TripleTen scaled its marketing operation, manual processes stopped keeping up in three specific places.
No systematic way to monitor competitor trends. The user acquisition and social media teams wanted visibility into what competitors were producing and which formats were gaining traction. In practice that meant manually browsing ad libraries, screenshotting examples, and dropping them into Slack ad hoc. There was no consistent, scalable way to track trends across markets or turn scattered observations into direction for creative production.
No visibility into what actually drives creative performance. TripleTen produced over 1,000 creative assets in a single year. With format, target audience, pain point, value proposition, and call-to-action all varying independently, the team had no reliable way to connect creative characteristics to outcomes like Cost Per Lead, CAC, or ROAS. Manually tagging every variable across 1,000+ assets to find the pattern would have taken weeks — and inconsistent human tagging would have made the resulting analysis unreliable anyway.
Hours lost to repetitive ad uploads. Every creative test meant manually building campaigns, naming ad sets, and assigning creatives inside each platform's ad manager, one at a time, across multiple platforms. At TripleTen's testing volume, that ate hours per week that could have gone toward optimization instead of data entry.
The Solution: Three Scenarios Covering the Full Creative Lifecycle
Vatech.io built three Make.com scenarios that together cover publishing, competitive monitoring, and performance analysis — the entire creative intelligence pipeline TripleTen needed.
1. An AI agent for competitive intelligence
To replace ad hoc manual research, we built a competitor monitoring AI agent in Make. The scenario pulls structured data from the Meta Ads Library, TikTok Ads Library, and Google Ads, using Apify to scrape public ad libraries for a defined list of competitors stored in Google Sheets. For each competitor and market, Make extracts raw creative data — video links, headlines, hooks, call-to-action copy, and descriptions.
A second layer of the scenario analyzes that raw data with AI: labeling creative characteristics, categorizing them against TripleTen's internal marketing framework (target audience, value proposition, pain point), and translating content from any source language into English. The aggregated output lands in Google Sheets and is summarized in Slack, surfacing top trends, recurring formats, and standout competitor examples for the creative and social teams to act on directly.
"Previously, my user acquisition team manually searched and explored creatives from competitors, then posted examples in Slack with a message like 'hey, let's discuss this.' Now the agent does that work automatically. It's faster, it's aggregated, and it's not just one example." — Maksim Epifanov, User Acquisition Lead at TripleTen
2. Creative performance analysis powered by Gemini
The second scenario tackles the question every performance marketing team asks: which creative elements actually drive results? Make analyzes TripleTen's full library of creative assets — videos and banners — across 20 defined parameters, including format, target audience, pain point, unique selling proposition, and call-to-action type.
Each asset is sent to Google Gemini, which extracts and assigns values for all 20 parameters. That structured output is stored in Google Sheets and joined with actual campaign performance metrics. The result is a systematic view of which formats deliver the best ROAS, which audience segments outperform others, and which topics resonate most — the kind of analysis that would take a human analyst weeks to compile by hand and would still carry inconsistent tagging.
The output changed what TripleTen chose to produce: the data showed they needed more user-generated and influencer content, that parents outperformed other audience segments in their vertical, and that specific pain points — like concerns about AI replacing jobs — drove stronger engagement than others.
"Tagging creatives shows me exactly which elements drive performance and which fall flat, so I can test more precisely, scale what works, and cut what doesn't." — Polina Zhilkova, Creative Project Manager at TripleTen
3. One-click creative uploads across ad platforms
The third scenario connects TripleTen's centralized creative database directly to multiple ad platform APIs. When a creative asset is ready, a user acquisition manager opens a link, clicks upload, and Make handles the rest — creating the campaign structure, naming ad sets, and uploading the creative with consistent naming conventions applied automatically. What previously took hours now takes under a minute.
"Automating creative uploads for Meta Ads with Make has been a game-changer for us. It cuts the time spent on routine ad setup by 5–7x, reduces manual errors, and frees up the team to focus on strategy instead of repetitive tasks. It's one of those improvements that scales instantly as campaign volume grows." — Ivan Galenko, User Acquisition Manager for Meta at TripleTen
"I'm using a Make scenario to automate creative uploads to Google Ads. Since we test a lot of videos, doing it manually was too time-consuming. Now, we've got it down to one-click launches!" — Muhu Guseinov, User Acquisition Manager for Google Ads at TripleTen
The Results
The three scenarios delivered measurable impact across TripleTen's marketing operation:
- 5–7x reduction in time spent on ad setup for Meta campaigns, holding consistent as campaign volume scales.
- Hundreds of hours saved per month across the user acquisition team on manual creative uploads.
- Creative labeling that would have taken weeks now runs automatically across 1,000+ assets per year.
A Repeatable Architecture for Marketing Teams at Scale
TripleTen's setup is a clear example of Make.com used the way it's designed to be used: not as a single automation, but as connective tissue between AI models (Gemini), data scraping (Apify), ad platform APIs (Meta, TikTok, Google Ads), and the team's existing tools (Google Sheets, Slack). None of the three scenarios required TripleTen to replace their stack — Make sits between the systems they already had and removes the manual labor connecting them.
This pattern — a scheduled scrape-and-classify loop feeding structured data into Slack, paired with an AI tagging layer that turns unstructured creative assets into analyzable data — is the same architecture we cover in more technical depth in Building an AI Competitive Intelligence Agent. If you're running high-volume paid creative testing and want the same visibility into what's actually working, or want your team out of manual ad-platform data entry entirely, that's exactly the kind of automation architecture Vatech.io builds with Make.com.
