Debunking the Speed vs. Quality Myth
Tim Duong leads engineering, product, and design recruiting at Lovable. He walks through the two problems he had to solve in his first two weeks on the job: engineering leaders spending 60% of their week interviewing, and no shared definition of what "senior engineer" meant. He explains how pass-through-rate data found the exact stage costing the process candidates, and how the time that data saved got reinvested in the kind of candidate experience that gets people to say yes.
Speaker
Key Takeaways
- Cutting time to fill and protecting quality of hire can move together when a team rebuilds its actual hiring process, as Lovable did in two weeks for its senior engineering hires.
- Interviewer time is a capacity problem worth planning against: Lovable's engineering leaders were spending 60% of their week on interviewing, and a three-month capacity plan brought that down toward a 25% target.
- Pass-through-rate data can point to the exact stage losing candidates, as Lovable found when it traced a drop-off to a late-stage AI coding interview and moved it earlier in the process.
- Fast, structured feedback compounds: over 70% of Lovable's interviewers submit scorecards within 24 hours, using Ashby's AI notetaker to keep every scorecard clear and useful.
- Hiring managers who source directly alongside recruiters raise both collaboration and top-of-funnel quality, and gain a stake in the candidates they helped find, as Lovable saw pairing its hiring managers with sourcers in Juicebox.
Session Overview
Tim Duong spent twelve years recruiting at Google before joining Lovable to lead engineering, product, and design recruiting. The pace shift was immediate: a process change that took six to nine months at Google gets decided in days or weeks at Lovable. Neither speed is wrong, he argues. Each fits what the company and the moment call for.
His first two weeks set the tone. Leadership asked him to build a new senior engineering hiring process from scratch, with a two-week deadline. He found two problems waiting. Engineering leaders were burning an unsustainable share of their week on interviewing, and the company had no shared definition of "senior engineer" beyond the title itself, a calibration question he says is still being worked out.
Data did most of the diagnostic work from there. Tim lives in Ashby's AI assistant daily, using pass-through rates to find where candidates were dropping off and why. One example: an AI agentic coding interview sitting late in the process turned out to be a major drop-off point. Moving it earlier took a simple change and returned real time and resources.
The same philosophy shows up in how the team works together. Most interviewers now turn feedback around within a day, helped by Ashby's AI notetaker. Hiring managers source candidates directly alongside recruiters using Juicebox, which Tim says raises both collaboration and top-of-funnel quality. Because Lovable builds its own tools, the team now ships hiring kickoff and debrief tools in about an hour, work that used to take weeks.
Tim closes with a story: a design candidate who flew in from New York with his family for final interviews. The team built neighborhood tours, restaurant recommendations, last-minute childcare, and conversations with employees who'd made the same move. The candidate later sent personal notes to everyone he'd met, and joined three weeks before this talk. That room existed, Tim argues, only because the team had already won back speed of hiring on the scorecards, analysis, and top-of-funnel work underneath it.
Chapters
- (00:03) Intro: from Google to Lovable
- (01:07) The speed vs. quality myth
- (01:56) Building a hiring process in two weeks
- (02:19) Fixing interviewer capacity
- (03:14) Defining what "senior" means
- (04:00) Using data and AI to find bottlenecks
- (05:48) Team collaboration and same-day feedback
- (08:15) A candidate experience story: the move to Stockholm
Q&A
Q: How do you build a hiring capacity plan from pass-through-rate data?
A: Start from the hire target and a stated pass-through rate, then work backward to the interview volume that target requires and the interviewer time that volume costs. Tim used that math to justify a three-month plan that brought engineering leaders' interviewing time from 60% of their week down toward a 25% target.
Tim Duong, Talent, Lovable: "60% is unsustainable. But we pushed forward anyway. One great hire unlocks a ton of capacity, raises the bar, and moves everything forward. So my pitch to leadership was pretty simple: 'You're at 60% now. Trust me, give me three months, and by September, you'll be at 25%.'" (02:49)
Q: How do you use pass-through-rate data to find a broken stage in your hiring process?
A: Interrogate pass-through rates stage by stage to test why candidates drop off, then act on what the data confirms over what feels true anecdotally. Tim used that method to trace a drop-off to a late-stage AI coding interview, then moved it earlier once the data backed up the hunch.
Tim Duong, Talent, Lovable: "A concrete and very simple change that we made as an example: at Lovable, we have an AI agentic coding interview at the latter stages of our process. We quickly identified that this was a big drop-off point anecdotally. But once the data really showed that, we simply just moved this interview earlier in the process. Super simple change, real impact, tons of time and resources saved." (04:39)
Q: How do you cut time to fill without lowering quality of hire?
A: Treat speed and quality of hire as goals a better process can improve together, then reinvest whatever time that process saves into the parts of the hiring process candidates actually notice. Tim rebuilt the company's engineering hiring process around data-driven fixes and put the time saved into deeper candidate experience work.
Tim Duong, Talent, Lovable: "A common assumption in recruiting is that speed and quality are a trade-off. Pick one. But at Lovable, we believe you don't have to choose between them. You can do both. What I'm about to walk you through are the plays that we're making around speed, the plays that we're making around quality, and how we're investing that time back into candidate experience." (01:37)
Q: How can hiring managers and sourcers work together to improve top-of-funnel quality?
A: Have hiring managers build searches directly alongside sourcers in the same tool. At Lovable, hiring managers use Juicebox to build searches with their sourcing counterparts, and Tim credits the pairing with raising collaboration and improving the quality of candidates entering the funnel.
Tim Duong, Talent, Lovable: "Our hiring managers and sourcing team work incredibly closely together. So hiring managers are using Juicebox to build out searches and identify candidates alongside their sourcing counterparts. This has done a few things. Obviously raised collaboration, but it's increased and improved top of funnel quality as well, and also given extra skin in the game for hiring managers who want to see those gems of candidates that they found themselves do well through the process." (06:49)
Tim Duong — Talent, Lovable
Hey. So I'm here to debunk the speed versus quality myth, but first, a quick intro. So I'm Tim. I spent the previous twelve years working at Google recruiting across EMEA and North America, building out various hiring functions. And then a few months ago, I joined Lovable, where I currently lead engineering, product, and design recruiting.
The pace at Lovable is different. I've gone from one of the most process-driven companies to one of its fastest-moving ones. At Google, a process change can take six to nine months, and at Lovable, you often have the space of a few weeks or just a few days to make changes. Neither is wrong. It's about what the company and what the moment calls for.
A process change at Google could impact thousands and thousands of candidates within the first week. And at Lovable, we haven't quite hit that sense of scale just yet. But one thing is the same across both: everybody wants top talent. So a common assumption in recruiting, speed and quality are a trade-off.
Pick one. But at Lovable, we believe you don't have to choose between them. You can do both. So what I'm about to walk you through today are the plays that we're making around speed, the plays that we're making around quality, and how we're investing that time back into candidate experience. Okay, I'm going to take you back to my first couple of days at Lovable.
I was given the grace of one day to onboard, which was very nice of them, and then on day two, the ask from leadership was, "Hey, Tim, can you go out and build a brand-new hiring process to hire the best senior engineering talent on the market?" Of course, I said yes. My second question was, "How long do I have to implement this?"
And the answer was a maximum of two weeks. So high stakes, high agency, zero time to waste. We had a few challenges, the first of which was interviewer capacity. Hiring at Lovable is the most important thing that we do. We truly believe that and mean that. But a lot of our engineering leaders were spending sixty percent of their weeks on interviewing and hiring.
Sixty percent. That is unsustainable. But we pushed forward anyway. One great hire unlocks a ton of capacity, raises the bar, and moves everything forward. So my pitch to leadership was pretty simple: "You're at sixty percent now. Trust me, give me three months, and by September, which is now, you'll be at twenty-five percent."
So short-term focus, long-term gain. Challenge number two, and this is a very important one. So at Google, we'd spent years defining and refining what leveling was and what seniority was in engineering. We had the rubrics, we built out all the leveling docs, the lot, and now I'm sitting at Lovable and literally all I have is this title, senior engineer.
What does that actually mean? So part of unlocking both speed and quality is making sure that everybody's aligned and calibrated on what a senior engineer is. And once calibrated, we see it, we know it, and we go and get it. So we had to build and define what a senior engineer was as a top, top priority, and truthfully, this is still a live question amongst the company.
We haven't fully nailed it yet, but I guess that's part of the fun of learning and iterating. So how have we been able to deliver processes with speed and quality? First is data and AI tooling. Personally, I'm in Ashby's AI assistant every single day, and it's been incredible for providing detailed and quick analysis of our processes.
Interrogating pass-through rates at each stage, not just where bottlenecks are, but why they exist. Where are candidates succeeding? Where are they falling off, and what's driving that? And when you start working back from the end goal, the numbers become absolutely key for iteration. So a goal of, say, ten senior hires at a ten percent pass-through rate, which is what it was, that's a lot of interviews.
So you're starting to ask, what does this mean for interview capacity, for training interviewers? The maths here drives every single decision. A concrete and very simple change that we made as an example: at Lovable, we have an AI agentic coding interview at the latter stages of our process. We quickly identified that this was a big drop-off point anecdotally.
But once the data really showed that, we simply just moved this interview earlier on the process. Super simple change, real impact, tons of time and resources saved. And that's the whole philosophy. You don't need a perfect process from day one. We needed to move, we need to measure and continuously improve.
And iteration here is the process. Second is how our team works together. Sourcers, recruiters, hiring managers, coordinators, everyone's tight, everybody's aligned, and honestly, it's a lot of simple back to basics relationship stuff, but with AI doing the heavy lifting. Each role here has a speed and a quality lever, and AI amplifies both.
So a few things that we have done to move the needle. One is same-day feedback. So over seventy percent of our interviewers submit feedback within twenty-four hours, which is music to my ears as a recruiter. We build writing time at the end of interviews so people can submit, but everyone uses Ashby's AI notetaker, which has been amazing for creating efficient, clear feedback for every candidate.
Second is collaboration. Our hiring managers and sourcing team work incredibly closely together. So hiring managers are using Juicebox to build out searches and identify candidates alongside their sourcing counterparts. And this has done a few things. Obviously raised collaboration, but it's increased and improved top of funnel quality as well, and also given extra skin in the game for hiring managers who want to see those gems of candidates that they found themselves do well through the process.
And lastly, continuous iteration. We continuously evolve the process constantly, and this is driven by the feedback and the data that I mentioned before, and that's been a key unlock. Once people understand that nothing's set in stone, they stop working around the process and really start to improve it. And of course, with us being Lovable, we build a lot of the tooling in-house ourselves.
So our recruiting and people team have built various tools using Lovable which span both the speed and the quality play. Everything from hiring kickoff tools to debrief tools, talent mapping dashboards. We've even created an app that we send to candidates with Stockholm recommendations when they come on site with us. And some of these things now take an hour when they used to take weeks to build or simply just not exist.
So a fast, clear process says this team's organized, they're decisive, they're respectful of your time. And a slow and chaotic process says the opposite before we've even got to offer stage. And for senior candidates who are almost always passively looking, the process is often what tips the scale. And when a process works well, candidates really feel it. Let me give you a real-life example of how this all came together once.
So we had a design lead fly in from New York for interviews in Stockholm. They brought along their wife and two young kids. Longtime New Yorkers, had never lived anywhere else as a family before. So this wasn't just about getting the interviews right or landing on the right compensation, it was a whole life decision.
So for the interviews, we did things like put together neighborhood recommendations. We took them on a neighborhood tour of Stockholm for potential places to live, sent them a ton of recommendations of places they should eat and things that their kids could do around the city. And our coordinator even scheduled last-minute childcare so this candidate's wife could explore Stockholm whilst they interviewed.
And the candidate spent a lot of real time with the team. Of course, they did their technical interviews and competency interviews. But we were also really intentional about setting them up with parents at Lovable and also people who had made the same move from New York to Stockholm to talk through logistics of moving, the culture shift, and of course, where to find the best pizza in town, which is very important for any New Yorker.
And after their visit, this candidate sent a super detailed memo doc with individual notes to every single person that they met along their journey, covering their experience at every single stage, how each call felt, what resonated with them, what they appreciated, and how valued they felt throughout their interview journey.
Genuinely thoughtful stuff. And that's not just a recruitment process. That's how you help guide someone with a family, with a life, and a real big decision to make. And the best part of that is they joined Lovable in Stockholm three weeks ago. Whoo. Thank you. And because we move fast on things like scorecards and AI analysis and top of funnel, we had room to build the experience so candidates like this can really feel it.
Of course, none of this is possible without the foundation underneath, the speed, the communication, the infrastructure, hiring manager relationships. But if you get that right, you can deliver incredible quality candidate moments like this. Thank you.
Recommended Sessions
The New Blueprint for Building Talent Teams
Reece Batchelor, Talent Acquisition Manager at Wayve, explains three assumptions about AI and lean teams that didn't hold up, and the team structure he built once he saw the real cost of getting them wrong.
View session
Quality of Hire at Enterprise Scale
Tim Chattaway of Sony Sports Technology and Laura Barnes of Flock compare how their teams measure Quality of Hire at scale: rebuilding scorecards around justified conviction instead of middle-of-the-road ratings, catching a struggling new hire inside the first 30 days, and feeding the data back into how the next role gets written.
View session