#FirstPrinciples with Dexter Zhuang — The Show Notes

An anecdotal deep dive on growth, retention, user behaviour, and experimentation mindset.

#FirstPrinciples with Dexter Zhuang — The Show Notes

By Ayush Sharma

We ❤️ scale. Our love for building products that impact millions all at once has been pretty evident throughout our journey. #FirstPrinciples, a series of product webinars we host, gives us the right opportunity to learn from industry experts about building products that constantly grow.

We recently hosted Dexter Zhuang, a product veteran who worked with the behemoth that’s Dropbox.

Dexter is currently working with Xendit, a fellow Indonesian payments company. In this #FirstPrinciples session, he spoke about the four years he spent with Dropbox as a Growth PM and the lessons he learnt from the journey. During Dexter’s time at Dropbox, the company saw exponential growth and a successful IPO at $10B valuation. When an organisation elevating at this scale, there would be no dearth of lessons to be learnt.

Here are the four key lessons from Dexter’s time at Dropbox:

Lesson #1 — Retention is the best indicator of a healthy product 🤝

Dexter believes there’s a lot of chatter about growth in the startup ecosystem, and not enough about retention.

He defines retention as the number of users still active within a certain time period you define.

Strong retention reflects a strong product-market fit. According to Dexter, there’s no point in pouring money for growth if there’s poor retention. Growth-hacking techniques at the top of the funnel without fixing underlying weak retention will lead to a “leaky bucket” outcome in which users simply pass through the product.

Citing Dropbox’s example, Dexter shares that the freemium model, frictionless sign-up, sticky engagement model, and in-product upsells enabled good retention.

But, how does one measure retention? 🤔

A good way to do so, is to talk in terms of the retention curves.

Source: Sequoia

Declining Curve: Hints poor product-market fit, eventually leading to only a handful or zero users. In this situation, it is important to focus on changing the product to find a value proposition for a core set of users and then expanding from that set.

Flattening Curve: Suggests the business is viable and people continue to return to it. The next step from here would be to raise the retention to a higher point.

Smiling Curve: Indicates a great product showing users eventually return to the product after initially churning away.

The logical thing to do if the retention falls under Flattening or Smiling Curve is to focus on growth. However, if the product is inclining towards a Declining Curve, the focus should be on retention, before moving to growth.

Dexter suggests a good way to ensure retention is through acquisition loops. Referrals worked really well for Dropbox as users who signed up invited more people to join after seeing the value in the product. Healthy retention for a freemium model like Dropbox ensured users paid overtime after their free trials got over.

Lesson #2—Nudge user behaviour (thoughtfully) towards getting more product value 💭

Once you have retention figured out, it’s time to focus on driving growth and a good way to do this is to nudge the users thoughtfully.

Growth = Product Value – Friction

Describing the role of a Growth PM, Dexter says, “Just because features exist within your product, it does not mean users will actually use them. As a Growth PM, you are responsible for accelerating the rate at which users adopt new features, which usually means removing the friction.”

Some nudges that worked for Dropbox were:

  • Launching free trials reduced barrier to entry for users to experience value
  • Contextual in-product upsells tested out different product actions to find the ones that clicked.
  • Making notifications more visible made a big impact if users hadn’t noticed them.
  • Improving information hierarchy, navigation, and discoverability.
  • Visual changes (colour, animation, abstractions) that demonstrated urgency were powerful.
  • Pricing messaging, like contextualising the price.

Lesson #3—Adopt an experimentation mindset 🔬

Quoting Todd Jackson, VP – Products at Dropbox, Dexter said, “ From running two-three experiments per month in 2012, Dropbox now runs close to a hundred experiments per month. This is driving more than a hundred million dollars in incremental revenue every year.”

Experimentation is a huge part of the growth infrastructure.

Think like a scientist — everything is hypothesis testing and constant iteration towards the truth.

Dexter suggests creating sandboxes for each team so they have autonomy to pursue their respective hypotheses. It is okay to fail. That said, it’s important to create a diversified portfolio of bets.

~50% low hanging fruit, ~30% big bets, 20% optimisations.

Host recurring meetings for teams to share their learnings and insights from experiments. Learn from each other.

Lesson #4—Build the machine that builds the machine 🛠

Quoting the Evernote founder Phil Libin, Dexter says, “Follow the rule of three. Everything in your company breaks as soon as it triples in size.”

Make sure you’re rethinking the process while scaling rapidly to ensure all teams are up to date.

Talking about the importance of investing in platform-tools, Dexter emphasises the importance of automating things to save precious engineering time. He urges it’s always better to look for bottlenecks while scaling and find ways to automate them.

Rehearsing for the next stage

While focussing on what’s happening at present, it’s easy to lose sight of the next stage, says Dexter. For instance, at Dropbox, they built quarterly forecasts, measured the numbers rigorously and introduced an escalation process around those numbers. The key is to keep building and refining the growth model.


You can catch the recording of the session here and find Dexter’s entire presentation here.


Have something you’d like to share on #FirstPrinciples, but never given a presentation? We’ve got just the thing for you.

Introducing ‘Lightning Talks’!

We’re opening three slots of five minutes each before the guest speaker at #FirstPrinciples begins their session. If your talk is picked, we’ll provide a mentor to assist you in putting together a clear and engaging presentation.

Interested? Apply here. 🖖