Case study
Plurio runs the team’s regular performance marketing routine end-to-end — using an AI agent, skills, and automated rules on full-funnel data across Meta and Google.
Top-rated online tech bootcamp with 2,600+ graduates
Сareer-change programs for working adults: engineering, data, QA, AI/ML, design, and cybersecurity.
7+ paid channels driving US, LATAM web & mobile acquisition
Meta, Google Search, YouTube, TikTok, and other performance channels.
1,000+ ad entities/month · <15% of budget on experiments
Most of the budget goes to proven scale; a defined slice funds new bets.
2-month sales cycle · $6K average customer value
High-ticket conversions on a long timeline — every scaling decision carries real cost.
2 funnels with 5 stages
Lead → Deal → Call → Qualified Call → Purchase — each stage measured and optimized based on different thresholds.
In numbers
Challenge
Running paid at TripleTen's scale means juggling delayed conversions, shifting plans, and thousands of ads — where each region, product, and funnel has its own thresholds. That volume and complexity showed up directly in wasted spend and missed scaling.
Solution
Plurio fully automates the performance marketing routine — based on TripleTen's full-funnel data, its playbook, and Plurio's best practices, applied through an AI agent, skills, and deterministic automated rules. A dedicated AI engineer sets it up and keeps tuning it with TripleTen. The system gets better every cycle — and the team's attention shifts from the manual work to improving strategy.
Client quote
This was a very strong period for the team. We rethought how we approach performance marketing, wrote down our decision logic as a first-class artifact, and rebuilt the rhythm of how we work. Plurio gave us what we needed most — decision speed at the level of every campaign, every ad set, every creative. The analysis that used to eat hours and meetings now takes minutes, and the team spends that time on the hypotheses and experiments that actually move the business.
Max Epifanov
VP of Performance Marketing at TripleTen
Data foundation
Plurio connected to TripleTen’s unified reporting data — the team’s existing signals became the source behind every decision and action.
Data sources
Meta Ads, Google Ads, TikTok Ads, Bing Ads, Influencers, HubSpot CRM, GA4, Retell, QualifyAI
Decision layer
Full-funnel attribution data with built-in predictions
Entity levels
01
Channel
→
02
Campaign
→
03
AdSet / AdGroup
→
04
Ad
→
05
Keyword
Primary KPIs
SQLs
CPL
Cost per SQL
CAC
What Plurio did at the operational layer
Problem
Routine campaign management ate hours every day
Fix
Automated 40+ regular tasks across analysis, planning, and optimization
Every layer of media buying is covered — early underperformers stop, winners scale, budgets rebalance across channels and campaigns, keywords are analyzed, and ad copy is optimized.
Problem
Manual ads management couldn’t react instantly, so losing ad sets kept spending and winning ones stayed underscaled
Fix
Set up deterministic rules with cascading conditions running 24/7
Every rule runs as verified logic and factors in forecasts when conversions are delayed. When conditions align, the agent stops, scales, relaunches, or duplicates directly in ad accounts — from manager approval to full autonomy.
Max Epifanov
VP of Performance Marketing at TripleTen
Problem
Conversions landed weeks after the decision that needed them
Fix
Built self-refining forecasts on the full-funnel data
The agent combines historical conversions, current performance, and every funnel-stage signal, then refreshes forecasts continuously. Rules and workflows always use the latest forward-looking CAC.
Problem
Every region and product ran on its own thresholds — too many variables to apply consistently
Fix
Standardized the team’s decision logic — applied to every campaign, ad set, and creative
Wrote down one shared playbook that covers the team's decision framework: ed entity classification - green (on plan → scale), yellow (borderline → monitor), red (below plan → decrease/stop), 7- and 14-day analysis windows, scale-vs-kill thresholds, and business targets.
What Plurio automated in Meta Ads
For this quiz-funnel campaign, the agent checks every ad set every three days. When CPL is on target, budget increases automatically. Over one month, this scaled the channel 2× while CAC stayed the same.
Stable
Every budget call runs on a live CAC forecast
The forecast sharpens with each new conversion, so the team can increase, decrease, stop, or relaunch from a forward number weeks before lagging CAC proves the case.
Top performers surfaced to scale, past winners to relaunch
The agent scores the whole creative portfolio against TripleTen’s winner formula, with delayed conversions factored in — listing top performers for scaling and past winners for relaunching.
Burned-out creatives pause the moment they break thresholds
Ads pause when CTR drops below target or CPA rises. Each format has its own healthy band, enforced every day so the team acts in time.
What Plurio automated in Google Ads
Every campaign checked against plan CPA, budget shifted up or down
A recurring Google Budget Optimization skill compares every campaign’s CPA against plan, analyzes the full account, recommends an increase or decrease, and executes the change directly.
Low-Quality-Score assets get rewritten
The agent pulls low-QS assets, generates fresh copy from stored brand context, and hands it back to the manager to upload.
Campaigns flagged based on lead-quality forecasts
The agent tracks Call-to-Qualified-Call rates in real time, forecasts poor CAC downstream, and flags campaigns to stop or decrease — with the reasoning attached.
Wasted keyword spend and high-intent traffic spotted in minutes
The team asks in plain language and gets a ranked list back. Negative keywords and budget moves go in the same day, with no manual export.
Why TripleTen chose Plurio
Automated daily media-buying operations
The agent watches every ad continuously, moves budget to what’s working, kills losers early, scales winners, and runs every decision on full-funnel data.
Turnkey setup led by our AI-marketing team
A dedicated AI engineer collects the business context, sets up the first skills and rules, and shares best practices — delivering a working system the team owns.
Value that compounds with every automation added
The agent analyzes thousands of decisions and outcomes to refine the system, so the next cycle becomes more effective.
Max Epifanov
VP of Performance Marketing at TripleTen
Skills built
42
11
Rules always on
60
Hours freed / week
