Save the Children

How the design of a website’s call-to-action banner during a campaign impacted conversion rates

Experiment ID: #203386

Save the Children

Experiment Summary

Timeframe: 02/13/2026 - 02/24/2026

Save the Children’s global reach, especially in Muslim countries, has made fundraising during Ramadan a successful campaign for them. However, not all traffic to the website during Ramadan was driven by the campaign. During previous high-urgency campaigns, we tested and found that a slide-up with gift arrays achieved higher conversion rates. In this experiment, run during Ramadan, we wanted to see whether less intrusive, persistent promotional elements on the homepage would impact conversion rates.

We ran the control (sticky-bar with gift array) against 2 variants:

Variant B: “Crisis alert banner” at the top of the page with a hyperlinked call to action.

Variant C: Sticky “promo bar” at the bottom of the window with a button CTA.

Research Question

By testing two different website tactical elements against an existing slide-up promo during Ramadan, we aim to understand how different promotional assets perform in driving donation engagement.

Design

C: Control

T1: Treatment #1

T2: Treatment #2

Results

	Treatment Name	Conv. Rate	Relative Difference	Confidence
C:	Control	3.7%
T1:	Treatment #1	4.2%	15.1%	91.3%
T2:	Treatment #2	4.5%	22.3%	98.7%

This experiment has a required sample size of 6,789 in order to be valid. Since the experiment had a total sample size of 21,604, and the level of confidence is above 95% the experiment results are valid.

Flux Metrics Affected

The Flux Metrics analyze the three primary metrics that affect revenue (traffic, conversion rate, and average gift). This experiment produced the following results:

0% increase in traffic
× 22.3% increase in conversion rate
× 0% increase in average gift

Key Learnings

This test aimed to assess the impact of replacing the existing slide‑up with less intrusive, persistent promotional elements on the homepage during Ramadan. We hypothesized that a lower‑friction approach would feel less disruptive and support donation engagement. Key Results:

+15% non-statistically significant increase (91%) in overall conversion rate for Variant B and +22% statistically significant increase (99%) for Variant C.
+37% statistically significant increase (98%) in overall RPV for Variant B and +47% statistically significant increase (100%) for Variant C.
-52% statistically significant decrease in click-through rate to the form for Variant B (100%) and -19% non-statistically significant decrease for Variant C (93%).

Overall, conversion rate and revenue per visitor increased for both variants, with Variant C showing the highest lift compared to the control. Click-through rate declined for both variants, particularly for Variant B. This suggests that users may have engaged with the experience differently and completed donations through alternate paths rather than the primary CTA. Although multiple tests were intentionally running during this period, users were not expected to experience more than one test at a time. We identified that some users were exposed to overlapping tests, which may have influenced performance, particularly click-through rates. The conflicting tests were paused, and a new iteration of this test was launched. As part of the update, Variant C was adjusted to a black sticky bar to improve contrast, as the previous color may have blended into the page, impacting performance. The experience was also added to the Gaza country page. Overall performance:

Desktop performance:

Mobile performance:

Recurring performance:

One-time performance:

Experiment Documented by NextAfter

Question about experiment #203386

If you have any questions about this experiment or would like additional details not discussed above, please feel free to contact them directly.