How SEMrush Unlocked a 7.1% Performance Lift by Validating Upper-Funnel Media with Rockerbox

At a Glance

• Customer: SEMrush (SaaS / Digital Marketing)
• Challenge: Internal attribution tool struggled to accurately value upper-funnel prospecting and cross-channel impact.
• Solution: Rockerbox Multi-Touch Attribution (MTA) and Marketing Mix Modeling (MMM).
• Result: A controlled experiment yielded a 7.1% lift in performance by reallocating budget to Paid Video and Paid Social + Paid Display.

1. The Challenge: Blind Spots in a Home-Grown Model

As a global leader in digital marketing intelligence, SEMrush manages a massive, complex media mix. To track performance, the company had developed its own internal attribution system. While the internal system was effective for aligning performance teams around conversion insights, it began to show cracks as the company's cross-channel complexity scaled.

Paula Iglesias, Marketing Analytics Project Manager at SEMrush, noted that their internal system struggled to distinguish the true impact of prospecting campaigns versus retargeting.

"We realized that with all the updates... we might not be measuring correctly prospecting campaigns—mostly top-of-funnel campaigns—because we didn't have the granularity to actually see the impact of those campaigns." Paula Iglesias

The team faced specific "blind spots":

• Granularity: The internal system grouped platforms into broad buckets (e.g., "Prospecting" and "Retargeting") without visibility into specific channel impact (e.g., YouTube vs. Meta).
• Attribution Logic: The internal model relied heavily on lower-funnel signals, often failing to credit the upper-funnel touchpoints that initiated the customer journey.
• Privacy & Scalability: As privacy regulations tightened and data sources became "noisier," the internal tool struggled to maintain accuracy across SEMrush’s global footprint.

2. The Approach: The internal system vs. Rockerbox Experiment

SEMrush didn't just switch platforms; they validated the decision through rigorous testing. The marketing and analytics teams designed a controlled, 8-week experiment to compare Rockerbox MTA against their internal model.

The experiment focused on prospecting campaigns across three key channels:

1. Paid Display
2. Paid Social (Meta)
3. Paid Video (YouTube)

The Experiment Setup:

• Hypothesis: Rockerbox’s data-driven model would reveal undervaluation of upper-funnel channels that the internal system was missing.
• Methodology: The team ran campaigns side-by-side. One set was optimized based on the internal system’s recommendations, and the other was optimized using Rockerbox’s attribution insights.

The "Shocking" Insight

The results highlighted a distinct divergence in strategy. The internal system recommended investing heavily in Display. Rockerbox, however, recommended shifting budget toward Paid Video (YouTube) and Paid Social (Meta)—channels the internal system historically undervalued.

"A big surprise for us... was Paid Social, specifically Meta. This channel for us was something that we always had in a very small scale... we didn't have the resources and we didn't know if it was going to make a change. But actually, Rockerbox was telling us that this could be a great growing space for us." — Paula Iglesias

3. The Results: 7.1% Lift and Operational Agility

Following the data, SEMrush shifted their budget allocation away from Display and into Paid Video and Meta. The results validated the Rockerbox methodology immediately.

• 7.1% Performance Lift: The Rockerbox-driven campaigns generated a measurable 7.1% lift in performance compared to the control group.
• Content Optimization: Beyond just channel allocation, Rockerbox provided granular insights into content types. It revealed that specific video creatives were driving conversions, prompting a shift in their creative strategy.
• Operational Flexibility: Unlike the rigid internal tool, Rockerbox allowed the team to rapidly explore attribution scenarios and utilize synthetic touchpoints to fill data gaps.

"The experiment validated Rockerbox’s model not through theory, but through impact, a 7.1% measurable lift in performance that translated directly to better business results." — Paula Iglesias

4. Looking Ahead: A Unified Measurement Framework

The success of the MTA experiment has cemented Rockerbox as a core pillar of SEMrush’s measurement framework. The 7.1% lift helped secure buy-in from executive stakeholders and finance teams to expand the budget for previously untested channels.

The Future Roadmap: SEMrush is now moving toward a Multi-Methodology Measurement approach:

• MTA (Current): Powers day-to-day optimization, pacing, and tactical content decisions.
• MMM (Next Year): SEMrush plans to implement Rockerbox’s Marketing Mix Modeling (MMM) to handle long-term budgeting and forecasting, combining the speed of MTA with the macro-view of MMM.

"We are looking into actually increasing the budget of those channels... not with a redistribution from other channels, but actually getting more money from the finance team to invest in these channels next year." — Paula Iglesias

Why SEMrush Recommends Rockerbox

When asked what she would tell others evaluating the platform, Paula emphasized the ease of implementation and the support team:

"It was super easy to implement. They make it sound easy even though it's not... I think that was for sure one of the things that made the overall process and project very easy to work with."