Relying on color alone to convey information can leave parts of your audience behind. Some people have color vision deficiencies and can’t distinguish certain hues, while others may view graphs in grayscale or at a glance miss subtle color differences. Colors alone also don’t communicate patterns, trends, or magnitudes as clearly as patterns, textures, or explicit labels do. By adding patterns, textures, labels, or clear legends alongside color, you ensure that the data remains understandable across abilities and viewing conditions. So this answer is best because it acknowledges that some individuals may need patterns in graphs to understand the data effectively, highlighting why color should not be the sole encoding. The other options overstate color’s universality or suggest relying exclusively on color, which reduces accessibility and interpretability.

In a typical A/B test, participants are assigned to the two variants randomly, usually with about half seeing each version. This randomization makes the groups comparable on both observed and unobserved factors, so any difference in outcomes can be attributed to the variant itself rather than who the users are. A 50/50 split also maximizes statistical power by keeping the group sizes equal, which helps detect real effects more reliably. Allocating by location or device would introduce confounding factors, since differences in results could reflect user characteristics rather than the variant. A fixed or uneven split might be used in special cases, but the standard approach is a random, roughly equal allocation.

Global site tag is a single snippet designed to work with Google products, configuring things like Google Analytics and Google Ads on your site, but it isn’t a general tag manager for third-party tags. Google Tag Manager is a container you place on your pages, and inside it you can deploy a wide range of tags—Google and non-Google—without editing the page code for each one. This makes GTM the flexible option when you need to coordinate multiple vendors or add non-Google tracking pixels. The statement highlights that the global site tag works with Google tools, while Google Tag Manager can handle any HTML or JavaScript tags. In contrast to some misconceptions, GTM doesn’t require extensive manual coding for every tag—the UI lets you configure most tags, with optional custom HTML/JS as needed, whereas the global site tag doesn’t provide that multi-vendor management capability.

To express the portion of all purchases that came from the campaign as a percentage, you divide the campaign-related purchases by the total purchases and multiply by 100. This converts the share into a percent that shows how much of the total purchases were driven by the campaign. For example, if there are 40 campaign purchases out of 200 total purchases, the percentage is (40 ÷ 200) × 100 = 20%. Dividing total purchases by campaign purchases would give an inverse ratio, which isn’t a percentage of all purchases. Multiplying the campaign count by 100 just scales the raw count, not its share of the total. Subtracting from total and multiplying by 100 also doesn’t represent the proportion of the total.