2.1× engagement rate uplift and 4.6M organic impressions in a single summer campaign.

The Challenge

Sprite's social media calendar was built around brand content — polished, high-production posts that reflected the brand's visual guidelines. Engagement rates averaged 1.2% — strong for a large brand page, but significantly below the 3–4% benchmark for category-leading FMCG brands on Instagram. The summer campaign window — the brand's highest-relevance season — was approaching with no strategy for Reels or trending audio, and no mechanism to respond to real-time cultural moments.

The Solution

Results

The combination of structural changes, AI-driven optimisation via Adtrafix OS, and weekly cadence produced measurable improvement starting in week 3, compounding through month 3:

Why Creator-Style Content Outperformed Brand Content by 2.1×

The shift from high-production brand content to creator-style Reels was the single most impactful change in Sprite's social media programme — and also the most counterintuitive recommendation for a brand with Coca-Cola's production standards. The data behind the recommendation was unambiguous: Sprite's three highest-performing posts from the previous 12 months were all lo-fi, text-on-screen Reels with no professional production. The three lowest-performing posts were all studio-produced brand videos.

The reason is algorithmic. Instagram and YouTube Shorts prioritise content that holds attention — completion rate, saves, and shares — not content that looks expensive. Creator-style content, with its native-platform aesthetic, generates significantly better completion rates than broadcast-style brand content because it looks and feels like what users expect to see from their personal network, not from a brand account interrupting their feed.

The Performance Scoring System in Practice

Every Reels script was run through Adtrafix OS's performance scoring before production began. The score evaluated three dimensions: Hook strength (first 2 seconds — does it create enough curiosity or pattern interrupt to prevent swipe?), Content structure (does it deliver on the hook's promise within 15 seconds?), and CTA alignment (does the ending create a save or share motivation?). Scripts scoring below 6.5/10 were revised before production, eliminating the cost and time of producing content that data predicted would underperform.

Over the summer campaign period, 23 scripts were submitted for scoring. 8 were revised based on scoring feedback. Of the 15 that went to production with their original script, average engagement rate was 2.1%. Of the 8 that were revised based on scoring, average engagement rate was 3.4% — a 62% improvement attributable directly to the pre-publication scoring intervention.

The Competitor Intelligence Advantage

Adtrafix OS tracked 15 FMCG brand accounts in Sprite's competitive set weekly — including Thums Up, Mountain Dew, Limca, and emerging D2C beverage brands. The intelligence surfaced two consistent patterns: brands posting trend-reactive content within 24 hours of a cultural moment received 3.2× the organic reach of brands posting evergreen content in the same period; and carousel posts about "ingredient truths" and "brand myths" consistently outperformed product-led content across all beverage accounts tracked.

Both insights directly shaped Sprite's content strategy. The trend-response system was built around the 24-hour window Adtrafix OS identified as the peak engagement window for reactive content. The "myth-busting" and "ingredient truth" format became a monthly content series for Sprite — one of the highest-performing recurring formats in the account's history.

Key Takeaway

This case demonstrates what changes when AI is embedded into the operating model rather than used as an occasional tool. The results compound weekly — each optimisation feeds the next — rather than spiking once and stabilising. Adtrafix OS running behind every decision is what separates a one-month win from a structural improvement.