Accelerating Developer Activation: Reducing Onboarding Drop-Off by 45%

Overview and Outcomes

Apollo is a leader in GraphQL tooling and the creator of GraphOS, a platform for building, managing, and scaling GraphQL APIs. GraphQL is used to simplify how applications communicate with data sources — it gives developers a single, flexible endpoint to fetch precisely the data they need, which dramatically improves performance and developer experience.

GraphOS acted as the operating system for our data graph — it gave engineering teams a unified platform to manage schemas, monitor performance, handle versioning, and secure API access. For enterprise evaluators, it reduced complexity and made adopting GraphQL at scale much easier.

When I joined, only 8% of users moved past the initial onboarding screen, and conversions hovered at 0.1%. My goal was to turn this around by delivering an onboarding experience that educated, activated, and supported developers at every stage of their journey.

Initial onboarding data revealed that >50% of new users lacked an API endpoint and were just exploring GraphQL. We needed to understand user intent to shape future onboarding experiences. After V1 launched, we moved into V2. With growing demand for enterprise features, the sales-led model couldn't scale. Developers wanted to self-evaluate GraphOS without talking to sales, but had no way to explore premium features on their own. Despite improvements, most users still dropped off due to a lack of immediate value. Requiring an endpoint to get started was a blocker. The need for contextual guidance, progressive disclosure, and early value surfaced repeatedly. These insights are what led to our V3, referred to as "Unified Onboarding."

‍

Role & Scope

As the lead product designer on this multi-phase initiative, I owned end-to-end design strategy across Apollo’s onboarding and adoption experiences. Over two years, I partnered cross-functionally with product, engineering, marketing, and developer experience teams to ship scalable, high-impact improvements—from early user segmentation and self-serve trials to a fully unified onboarding system.

My work spanned user research, information architecture, feature integration, and conversion optimization, directly contributing to a 300% increase in enterprise trial signups and a 55% boost in user engagement.

• Led design for a fully reimagined onboarding experience, introducing contextual guidance, new IA, and a modular landing experience based on persona needs.
• Integrated multiple Apollo services (e.g., Odyssey tutorials, demo graph, REST connectors) directly into onboarding, increasing self-serve activation.
• Worked closely with engineering to design and QA feature rollout for paid-tier trial pathways and brand system updates.
• Championed for the organization to integrate Amplitude as our data & analytics platform. (which we did!)
• Drove the cross‑functional ICP narrative (“Charlie at 3M”) across 3 PMs, CS leadership (EMEA), principal SA, solutions eng, eng director, and Director of Global Enterprise SE. 
• Pushed for language clarity (industry terms first), surfaced trust/compliance content, created forked journeys (demo/no‑endpoint paths), and apollos first custom path onboarding experience

‍

Key Outcomes

• 45% drop-off reduction in onboarding flow  
• Engagement rose from 8% to 55%  
• Enterprise trial signups grew from <1% to 4% weekly  
• Product demo accounted for 30% of all onboardings  
• User research base grew from 0 to 1,000+ in the first 3 months of launching our opt-in

Discovery

As a precursor to the work, our growth team ran an onboarding survey that collected a little over 13k responses between mid-August and october 4th 2024. We needed to understand user intent to shape future onboarding experiences

‍

‍

We were seeking to understand more around our persona segments (who are these people), their jobs to be done (what goals do they want to achieve?), and their journey thus far (from learning GraphQL all the way to having a graph in prod).

This survey accomplished:

• Capturing user goals, experience level, and intended use  
• First segmentation of users based on motivations  
• Introduced light personalization to onboarding emails  

‍

‍

The most important takeaway from the survey focused on developer use cases (aka, why are you here?), with almost 50% of survey respondents sharing their top use-case was running a query.. meaning they are coming with the intention of doing something.

• Initial onboarding data revealed nearly half of sign-ups had no graph;
• Most just wanted to run a query or explore GraphQL.
• We had no real path for people who didn’t have one—so activation stalled.

‍

Ideal customer persona

Our target user is a tech lead at a large company with lots of APIs. She needs a quick, low-risk way to see value and a path she can share with her team. Users within our ICPs role are typically dealing with  ‘spaghetti’ REST APIs and looking for a way to consolidate APIs with governance.

‍

Enterprise focus

‍

‍

With growing demand for enterprise features, the sales-led model couldn't scale. Developers wanted to self-evaluate GraphOS without talking to sales, but had no way to explore premium features on their own.  In V1, everyone was silently routed into Serverless by default (free plan) which created a lot of friction for enterprise users looking to trial ApolloGraphQL.

Apollo’s revenue is driven by enterprise contracts, so letting enterprise evaluators self-start mattered. In V1, everyone was quietly routed into the free Serverless flow. That hid the things enterprises actually evaluate—private setup, security posture, and advanced controls—and it created day-one friction with security and platform teams.

This worked laid the foundation for self-serve Enterprise goals:

• Redesigned org creation flow with pricing tier selector  
• Self-serve path to begin a 30-day enterprise trial  
• optional  bonus: Opt-in user research data collection  
• Updated product visuals and tone after an extensive rebrand

‍

‍

With the onboarding survey informing our personas, signup now starts with a clear plan choice, which ties back to our enterprise goals. This required users to select a hosting method which didn't previously exist.

Because most revenue is enterprise, we put the 30-day trial at signup so champions can self-start. This enables evaluators to experience value immediately instead of being silently dropped into the free plan that can’t prove what they came to assess.

So why did we start here?

• Right tool, right away: Enterprise users aren’t evaluating “free”—they’re validating private, secure, scalable. Starting in the wrong plan creates false negatives.
• Faster proof-of-value: Champions can show a working demo to leadership without Sales in the loop.
• Cleaner signals: Plan choice + feature usage tell us who’s truly enterprise, improving lead quality and prioritization for Sales.

The value of that:

• Try the real enterprise experience (private setup, your cloud/region) on day one.
• Show progress to your team quickly (no procurement wait).
• Clear, consistent steps after signup regardless of plan.

‍

Early signals

• Trials became one of the strongest lead gen drivers for Sales  
• Teams evaluating GraphOS expected to build a proof of concept within onboarding  
• 300% increase in enterprise trial signups (now 50% of all sales leads)  
• Weekly trial signups grew from <1% to 4%  

Why This Mattered

Apollo’s onboarding experience was rigid and one-size-fits-all. Regardless of background or intention, every user was funneled into a flow requiring an API endpoint, excluding over 60% of incoming users who lacked one. Drop-off was massive, and there was no clear ownership around resolving this friction.

Goals

Primary: Get new users to their first “this is valuable” moment quickly and prove a viable self-serve model.

• Users: reach value faster (run something that works, understand what happens), remove the dead ends.
• Users: Support non-GraphQL users and technical evaluators  
• Business: Reduce onboarding drop-off and increase engagement  
• Business:Increase quality leads and proof-of-concept completions  
• Enterprise-level: Enable product-led enterprise trials  
• Enterprise-level: Improve enterprise adoption without reliance on sales

Contributions

Across all three onboarding phases, I led product design strategy, from problem definition and ideation through launch and iteration, collaborating cross-functionally with product, engineering, marketing, and developer experience stakeholders.

‍

Onboarding Survey (V1)

• Designed and shipped a multi-step onboarding survey that segmented users by role, intent, and experience level—laying the foundation for future personalization.
• Identified friction in early user journeys by leading exploratory research on new signups and their goals.
• Synthesized 13k responses to identify patterns (e.g. over 50% of users didn’t have a graph) and created user segments.
• Collaborated with developer experience (DX) teams to prioritize and scope a product demo based on high user demand for "learning-first" pathways.
• Partnered with marketing to update onboarding emails and education flows using insights from survey data.
• Influenced the prioritization of a GraphQL product demo based on insights from the survey.

‍

Self-Serve Enterprise Trial + Brand Refresh (V2)

• Simplified plan selection in onboarding to align users with the right tier based on needs and intent.
• Redesigned the org creation experience to support product-led trials for enterprise tiers.
• Created a research opt-in path during onboarding that grew Apollo’s UXR research pool from 0 to 1,000+ users in 3 months.
• Partnered with sales to ensure product-led trial flows aligned with pipeline goals and enterprise buyer behavior.

‍

Unified Onboarding Parts 1 & 2 (V3)

• Led design for a fully reimagined onboarding experience, introducing contextual guidance, new IA, and a modular landing experience based on persona needs.
• Integrated multiple Apollo services (e.g., Odyssey tutorials, demo graph, REST connectors) directly into onboarding, increasing self-serve activation.
• Worked closely with engineering to design and QA feature rollout for paid-tier trial pathways and brand system updates.
• Used Amplitude to define success metrics and evaluate onboarding engagement and drop-off rates—improving activation from 8% to 55%.

Research Insights

To understand the barriers to adoption and activation, we combined qualitative interviews with developers, unmoderated user testing, and behavioral analytics via Amplitude. Our research spanned both first-time and returning GraphQL users, as well as internal interviews with sales, support, and developer experience teams. Key insights included:

• Majority of users lacked graphs or production-ready data: Over 50% of new signups had no graph to connect—yet our onboarding initially required one. This revealed a significant intent mismatch and led to the creation of education-driven flows like the Spotify product demo.
• "Aha" moments were missing for new users: Developers unfamiliar with Apollo struggled to understand how to reach value. This insight fueled a series of onboarding changes designed to reduce friction, clarify benefits, and surface key features earlier.
• High friction from unfamiliar concepts: Terms like supergraph, router, and containers introduced unnecessary complexity for new users. We addressed this with more intuitive copy, contextual tooltips, and restructured flows that gradually layered in complexity.
• In-product discoverability issues: Even experienced users failed to find power features like schema contracts without a sales walkthrough. This showed us the IA and in-product visibility of our core offerings were limiting self-serve growth and needed to be restructured.
• Misaligned plan selection during sign-up: Enterprise-ready users were mistakenly starting on Cloud due to lack of clarity around tiers. This insight helped prioritize onboarding paths that aligned user needs with the right product tier early.
• Strong interest in continued learning: Many developers were entering the product to build skills—not just ship features. This influenced our decision to embed direct pathways to Apollo Odyssey, our tutorial and certification tool, from onboarding.

Delivered Work

‍

Onboarding Survey (V1)

• 3-step onboarding survey to capture user goals, experience level, and intended use  
• First segmentation of users based on motivations  
• Introduced light personalization to onboarding emails  

‍

Self-Serve Enterprise Trial + Brand Refresh (V2)

• Redesigned org creation flow with pricing tier selector  
• Self-serve path to begin a 30-day enterprise trial  
• Opt-in user research data collection  
• Updated product visuals and tone  

‍

Unified Onboarding Parts 1 & 2 (V3)

• New UI framework with progressive onboarding paths  
• Sidebar redesign replacing top navigation  
• Direct pathways to product demo, REST backdoor, boilerplates, Odyssey  
• Educational tooltips and improved language  
• Added CTA to select a Dedicated (paid) tier during onboarding  
• Introduced visual language updates aligned with new brand  
• Created navigation parity between enterprise and free-tier experiences  

‍

The Demo Experience

Not everyone arrives with an API. The work prior is what led to the approval for design and DevOps to invest in & build the spotify demo - which accounted for 30% of all onboardings.

‍

‍

‍

• See value fast: Click Demo → run a sample request → see the result in under a minute.
• Zero setup: No credentials, no config, safe read-only environment.
• Make it tangible: Simple explainer shows how one API pulls data from several services (no jargon).
• Choose a next step: From Demo, jump to connect your own API, use a boilerplate, or start the 30-day Dedicated trial.
• Consistent across tiers: Same flow for free and Dedicated—no relearning, no dead ends.

Phase learnings and outcomes

‍

Onboarding Survey (V1)

Learnings

• 50% of users had no graph  
• 70% were using Apollo for personal learning  
• 24% of those wanting to query an endpoint were focused on skill-building  
• Users lacked foundational context to continue with onboarding

Outcomes

• Guided future onboarding strategy and surfaced need for a no-code product demo  
• Informed marketing and DevRel updates to content and email flows  
• Sparked development of the Spotify demo, now used in 30% of onboards  

‍

Self-Serve Enterprise Trial + Brand Refresh (V2)

Learnings

• Self-serve enterprise users had 8% higher success rates than free tier  
• Trials became one of the strongest lead gen drivers for Sales  
• Teams evaluating GraphOS expected to build a proof of concept within onboarding  

Outcomes

• 300% increase in enterprise trial signups (now 50% of all sales leads)  
• Weekly trial signups grew from <1% to 4%  
• Research opt-in base expanded from 0 to 1,000+ within 3 months  

‍

Unified Onboarding Parts 1+2 (V3)

V1’s job was to remove early friction and confusion. I clarified the landing page, added a short ‘what’s about to happen’ overview with links to security info, rewrote the URL step in plain language with simple help text, and surfaced the key network settings early so the first try actually worked. I also added a real path for people without a URL. V1 still fell short. Different intentions still needed clearer routes. Some wanted to start with what they already had; others wanted an isolated, private setup. One path couldn’t do both well.

V1 proved we can lift engagement, but it also showed a single path forces different users to fight the UI.

V2 started by  asking intent so users pick the right path and don’t bounce.

From the first screen, you can try a live demo, connect what you already have via REST, start from boilerplates, or open tutorials. Tooltips keep help where it’s needed. The structure is identical across tier. For mid-market evaluators, we added a clear CTA to start the Dedicated (paid) tier during onboarding to start testing a new market for pay as you go (usage based), mid-market tiers.

Introducing the decision fork removes guesswork and made the page feel relevant to our users. It also sets cleaner expectations for time and effort.

Learnings

• Clearer paths increased interaction with neglected product levers (e.g., Odyssey)  
• Developers responded positively to tier-specific guidance  
• Unified onboarding validated adoption as a key investment area  

Outcomes

• 45% drop-off reduction post-launch  
• Engagement rose from 8% to 55%  
• Triggered 2024 roadmap adoption initiatives and team formation  

Reflections

• You can’t build onboarding in isolation. Partnering with DevRel, Sales, and Eng ensured better narrative alignment and a shared source of truth.
• Personalization was a game-changer. Even light segmentation made a massive difference in engagement.
• Proving value early and reducing friction opens the door for wider adoption, especially in enterprise tools where first impressions matter most.