BrandKit

For users, getting started felt overwhelming. Assembling brand elements before their first email delayed time to value. For the business, those delays meant fewer users reached the first send milestone, a key driver of activation and conversion.

The insight was straightforward: users should never have to build their brand from scratch every time they create an email. The existing product had no mechanism for capturing and reusing brand identity. Every new template started from zero.

The opportunity was a system that could generate a brand identity from a URL and apply it across the email experience. Instant brand setup instead of manual assembly. Faster first sends.

From a business perspective, the hypothesis was that reducing brand setup friction would increase the likelihood of users creating and sending their first email. BrandKit was positioned as a time-to-value accelerator.

I worked with product to connect the user friction we were seeing in research with the activation metrics the team was targeting. Together we framed the goal clearly: increase activation by helping users create professional-looking emails faster.

BrandKit required engineering to build extraction for logos and colors from websites, structured brand profiles, and template application logic. I worked closely with the engineering team to map technical capabilities and constraints before designing the flow. This meant the experience was technically realistic from the start.

One key decision was making BrandKit automatic. Instead of asking users to configure a brand profile, we introduced it during onboarding so users would start with a branded environment immediately. This ensured the feature would impact activation rather than becoming a hidden tool users might never find.

Extracting colors was straightforward. Mapping them intelligently was the real design problem. Colors had to go to the right places: backgrounds, accents, text. Done wrong, users would feel trapped in a mismatched design and lose confidence in the feature entirely.

The engineering team and I tested multiple auto-mapping approaches. Some overfit to brand palettes and failed on low-contrast colors. Others were technically correct but produced ugly results in real content.

The solution was a hybrid model. The system auto-maps reliable colors to high-confidence template areas (backgrounds and primary accents) where the logic holds across a wide range of palettes. For ambiguous mappings, users get a simple adjustment interface. The goal was to balance automation with trust, so users felt assisted rather than overridden.

The results showed a clear improvement in early user success. More users created professional-looking emails quickly because branding was already applied when they reached the editor. Activation increased as more trial users hit the first send milestone.

Users who engaged with BrandKit converted at 63.9% compared to 19.6% for those who didn't. BrandKit plus asset uploads yielded the highest conversion at 68.7%. BrandKit alone, without uploads, still converted at 44.2%, more than 2x baseline.

This strengthened a core growth lever: the faster users send their first email, the more likely they convert. Over 124,000 accounts have engaged with BrandKit, and the platform has processed over 107 million asset uploads.

What made this project work was alignment across disciplines. Product framed the opportunity, engineering built the extraction system, and design tied it together into something users could trust on first contact. The cross-functional partnership was essential.

The color mapping system works well on typical brand palettes but has limits. Low-contrast or neutral-heavy palettes still produce mappings that need adjustment. The hybrid approach absorbs most of that friction. Shipping something that works for 90% of users while knowing the other 10% need manual fixes is a calculated trade-off, one the conversion data suggests was worth making.

A fair caveat: users who visit BrandKit are likely more engaged overall. BrandKit engagement is a strong leading indicator of conversion intent, though it may not be the sole driver. The 3.3x lift reflects both the feature's value and the signal of user investment it captures.