Sales onboarding redesign

A project to improve the new-user onboarding for ZoomInfo Sales. We rewrote the value props on every step, quieted the layout, and gave users a way to skip from the get go.

New users landed in an 8-step onboarding before they could use ZoomInfo Sales. The layout didn't help either: the left side, where users did the setup, was often dense with text. The right side included a big illustration, and on some steps a customer testimonial on top of it, neither of which carried much information about what the user was being asked to do.

Value props, which were supposed to answer "why am I giving you this info," were either missing on some steps or too long and vague on others. The step that asked "Who are you trying to reach?" had this as its value prop: "Shorten your sales cycle by finding all your Buying Committee members at once. Identify up to 8 Personas to help you close deals quicker." A long text about Buying Committees and closing deals, for a step whose actual ask is "tell us the roles you sell to and we'll help you find them faster."

This resulted in new users taking too long to reach the actual value that would make them come back: searching for and exporting contacts, sending an email, and so on.

Fewer steps meant shorter onboarding, which directly meant faster time to value, so that's what we pushed for first. If a step wasn't needed before first use, move it somewhere the user would run into it naturally, or relocate steps an admin could do for the whole org into admin setup.

But Engineering and Product pushed back due to scope and business value: removing a step meant it had to live somewhere else, either contextually in-app or in an admin-only flow. Both of those were descoped. Additionally, some steps weren't really about getting users to value in-app. Some were used for gathering data (e.g., asking for the user's job title to improve data and analytics) or drove meaningful traffic to other places (e.g., advertising the mobile app to increase mobile downloads).

Given these constraints, we asked ourselves: what else could we do to improve time to value?

We explored two concepts for how we could make onboarding faster using layout and copy improvements:

(1) Keeping the two-column layout, but with tighter copy and illustrations more specific to each question. This was a quieter version of what was already there.

(2) A more minimal approach that would support faster onboarding completion: centered visuals and minimal illustrations for faster reading, and friendlier, clearer, shorter copy. Lemonade's onboarding stood out in our competitor research, and we used it as inspiration. Internal design reviews favored the Lemonade-inspired concept.

Quieter page layout

We removed almost all illustrations and dropped the two-column layout in favor of a centered content area. Where we thought illustrations could help (e.g., an illustration around the QR code for downloading the mobile app), we used smaller, more specific visuals. We assumed these would make content easier to scan and faster to complete.

We also rewrote the value props on every step with PMMs. Some steps had none; others were too vague and didn't explain why the user should answer. The new ones were short and tied to what the user gets out of answering. We considered adding contextual motivational microcopy (e.g., outcomes, hours saved, social proof), but agreeing on numbers for every step was its own project and was descoped.

Changes to the welcome screen

We made several dedicated changes to the welcome screen because it had its own set of issues.

First, the old flow only showed progress from step 2 onward, which meant users couldn't tell how much they were in for until they were already past the first screen. We assumed a progress bar in the first screen would help alleviate uncertainty and introduce an endowed progress effect, because current steps already filled their part in the progress bar.

Second, in the old flow, the welcome screen was unskippable. To improve user control and time to value, we added a skip button to that step.

Third, again taking inspiration from Lemonade, we added a small tone change: a waving emoji next to the greeting to make the first screen in the app feel friendlier.

Fourth, while we couldn't drop the role question itself, the new welcome screen shows a role auto-filled from ZoomInfo data we already had on them, with Yes and No buttons to confirm or correct it.

Consolidating sub-flows

Since removing steps was out of scope, we consolidated related screens into one where it made sense. In the old flow, defining the Ideal Customer Profile was divided into three separate screens (location, industry, company size). In the new flow, we combined them into a single fill-in-the-blank sentence on the same screen. This didn't remove any setup users had to do, but it did make the wizard shorter (6 vs 8 steps).

Time to value and completion rate didn't improve. Step reduction had already been descoped, and later projects confirmed: when other teams removed steps, time to value improved. In retrospect, we should have done user research before the scope was locked. It was skipped for time, but it would have given us the evidence to push harder for step removal and make the case for redefining the scope before development started.