The Evolution of Auth0's (Ops-first) Research Practice
It’s hard to believe it’s been over a year since I last wrote about what we’ve been working on at Auth0. We’ve heard a lot from Research @ Auth0 team in that time, including Increasing ResearchOps Impact Through Automation, Getting Started with Inclusive Research, Onboarding to an Ops-First Research Team, and The Evolution of Auth0’s Insights Repository in case you missed them 👀
Recently, though, a lot has changed for both the team and the company, and I wanted to talk about what’s happening and why. Before we get too far into that, let’s provide some context about where we’ve come from and where we are now 📆
The Early Days
If we go all the way back to the beginning of 2020, I joined Auth0 to prove a hypothesis: that building a Research function by focusing on Research Operations first was the best way to do so (especially at a hypergrowth startup). Over the past 3 years, I think we’ve successfully proven that hypothesis:
- In 2019, Auth0 completed 3 research studies. In 2020, that number grew to 12, 19 in 2021, and 27 in 2022 🚀
- The number of people/teams who engage in research has grown from 6 to 42 📈
- Our insights repository has 550+ entries and our Research Program has 1800+ participants 💪
- Our playbook includes 56 pages, 26 videos, and 9 tools, and 380+ Auziros have completed our Research Onboarding course 📚
Where we are now
3 years is basically forever in the tech world; doubly so in a startup setting. Auth0 today is not what Auth0 was in 2020:
- The company grew from a ~500-employee Series E startup to a ~1100-employee pre-IPO unicorn 🦄
- In May 2021, the acquisition by Okta was completed, and we began a gradual and ongoing merger.
- The team that works on Auth0 now operates as an independent business unit within Okta where our sole focus is on Customer Identity 👤
Heading into 2023, it was time for a change. We need to evolve our research practice to meet the needs of our business with renewed focus and clarity. We’re not a startup anymore: we have a very good idea of what we should be working on, but we have a fuzzy picture of what the solutions should look like. Democratized Research no longer serves that need. It’s time for something new 🔭
Democratized Research excels in situations where you need to cover a large surface area at a shallow depth with limited resources (i.e. basically every startup ever).
Shifting Our Focus
Based on the needs of the organization and the current resources we have available, we’re migrating from a Democratized Research model to a Centralized-Agency model. This represents the best match between what we can offer and what the company needs 🤝
N.B. If you’ve never struggled with having inadequate resources (budget, people, etc), you can skip most of this section. For the rest of us in the real world, read on to learn about what this means and why we’re doing it.
What’s Changing
Personnel. Throughout the past 3 years, we’ve invested heavily in democratization and in ResearchOps. This was evident in how we staffed the team: 2 dedicated ResearchOps teammates plus a Research Strategist. This made sense when we were supporting dozens of People Who Do Research (PWDR), but going forward we’re shifting most of our staff over to dedicated Researcher roles. After the transition, we’ll have a team of four Researchers with one person staying in Operations 👥
Process. When running an ops-first, democratized research function, the majority (if not the entirety) of the research that happens is self-service, driven by your PWDR. What this typically means is your Research Roadmap is virtually non-existent: you can service every request, but your ability to control which requests come in is limited 📋
Going forward, we not only get to decide which projects we take on (versus which are delegated to self-service; more on that later), we have to have a way to let people know what order we’ll be picking projects up. Enter our triage and prioritization process. Basically, as requests come in, they either immediately go into our self-service pipeline or wind up somewhere on our backlog. The position on that backlog depends upon a variety of factors, including business priorities, scope, team capacity, etc. This backlog lives in Jira where anyone at the company can view it 🔍
If the requester doesn’t like their anticipated delivery date, they can either plead their case to have it reprioritized, or they can work with us to reduce the scope until it falls back into the self-service process. This way, nobody’s ever completely stuck 🔓
We’re also making some changes to how we do research, including implementing a peer review process for our Researchers. As a democratized practice, it’s hard to have solid peer review because strategy and expertise were centralized under one person. Now with a team of researchers, it makes sense to provide opportunities to give and receive feedback directly from each other, both strengthening the research we do and helping develop the team 🌱
What’s Not Changing
Any time there’s change happening, I like to give people a clear picture of what we’re planning to change, and sometimes more importantly, what’s not changing. Our Ops-First approach has a few major facets, and for the most part, those aren’t going anywhere. Notably:
- Rodrigo Dalcin (our Ops Lead) will continue to meet with each Product team every month to keep tabs on them and share updates from the Research team 📰
- We’ll continue our monthly knowledge-sharing rituals: Research Deep Dives (60-minute workshops or lecture-demonstrations on a specific topic within Research) and Research Readouts (5-minute summaries of all the research that’s completed in the past month) 🧠
- We still have a single point of entry for engagement with Research: our Insights Request Form 🚪
- We’ll still have a self-service research process, though with re-scoped levels of support to balance the needs of our Research team with the needs of our PWDR ⚖️
Does this mean Democratization is Dead?
Not at all! Nobody said Democratization was the be-all, end-all answer to how to structure a Research function. I still stand by the belief that it’s the best way to start building a team (unless you walk into a situation with a huge charter and a lot of buy-in at all levels of the business), and we’ll continue to reap the long-term benefits (standardized processes, strong Research culture, existing toolkit, clearly defined responsibilities, seamless onboarding, etc) for years to come.
If anything, the natural evolution of building Ops-First is to move on to another model: prove the value of Research by supporting a broad team with limited resources while preparing your team for scale.
That being said, we also aren’t shutting down the self-service branch of our research engagement process. We won’t have the bandwidth to service every request, but continuing to support self-service means we can empower teams to tackle some of the smaller/simpler projects themselves. If you’ve ever found yourself frustrated because you feel like you’re stuck doing low-impact work, democratization can help.
Why are we doing this?
It’s February 2023. Like many companies in the technology sector, we’re facing reality. In an ideal world, we’d grow the team while keeping our current level of service for Democratized Research, but with limited resources, the most responsible thing to do is to shift our focus 🔀
It’s not all gloom and doom though: there are a variety of benefits to this approach. One of the biggest drawbacks of Democratized Research is your Research team has relatively little control over the Research Roadmap. If you aren’t the ones doing the research, it doesn’t necessarily make sense to say no to anything. By centralizing, we gain a lot of control over what research gets done, including any strategic initiatives we want to take on ourselves (i.e. Consumer Auth Archetypes, which we’ll be launching later this year) 🏆
Centralizing also allows us to build a deeper understanding of customers and end-users in a few key areas. We may lose some visibility over things we’ve decided are less relevant to the current business goals, but for the topics that are in the spotlight, we can increase the volume of research and tighten the focus of it, leading to a more thorough understanding of the problem space 🔬
We’ve decided to go with an Agency model because it provides the right balance for us: reap the benefits of increased focus and depth while also maintaining flexibility to work on the top priorities and the largest needs. As much as I’m a proponent of Continuous Research, most teams don’t need to do research constantly, so we can deliver insights to one team and let them act upon them while we pick up work on something else 🔁
What’s next?
I’m confident this is the right move for our organization right now, and we’ve already begun to implement some of the changes (notably, personnel changes, updating our documentation, communicating all of the differences, and engaging with teams under the new model), however that’s not nearly the end of it. After all, as Mike Tyson said…
everybody has a plan until they get punched in the mouth.
Or, more eloquently (and originally) put by Helmuth von Moltke: no plan survives first contact with reality. So we’ll make adjustments as we go and learn from what’s working and what isn’t. And, of course, we’ll continue to share the journey here as well as in other places. Speaking of which…
Check out Great Question’s People Who Do Research virtual conference on February 8th, where I’ll both give a talk about the history of Research @ Auth0 and sit on a panel about building an Ops-First Research Practice. We’re already well over 500 attendees, so get your tickets now. See you there! 🎉
