Building Your Product Growth Strategy with ZALORA

The rise of an online audience is a force to be reckoned with in Southeast Asia, and ZALORA is just one of the many brands and businesses that has capitalised on this crowd to scale up their efforts in building a successful channel strategy to engage them. With over a whopping 45% of its sales coming from mobile devices through the app or mobile site, more of them need to focus on enhancing the user interface by translating data and feedback into actionable steps for product innovation.

Having amassed more than 7 years of expertise in developing platforms across SEA from their early beta, bootstrap mode all the way to maturity, May Chin, the Head of Product Growth & Analytics in ZALORA, further extends her ownership and strategic decision-making in these areas.

ZALORA is one of the leading fashion and lifestyle eCommerce players, with a robust customer base of over 50 million monthly visits. By leveraging on data analytics to stay on the pulse of diverse consumer needs and demands across the region, the brand is perfectly positioned to predict key insights into latest technological trends within the industry.

In this exclusive Q&A interview, she delves into the sizeable value when every product growth team treats an idea as a controlled experiment – measuring its tangible impact by gathering quantitative and qualitative data on user experience so they could grasp the way it addresses an actual pain point customers have. Customer-brand engagement is a two-way street, when generating data-driven insights on how customers are interacting with the product features, can we then roll out products that hit the mark.

1. Instead of chasing the latest tech trends, how does Zalora shift its focus to identifying and solving real user pain points?

My personal view is that following all the latest tech trends and best practices can often be, perhaps counterintuitively, rather dangerous. From my experiences and the literal 100s of experiments I have seen firsthand, what works for one company can never be reliably predicted to work for another, no matter how compelling or obvious the idea seems. The reality is you will never really know how a given idea or feature will work unless you experiment on it yourself with actual users in a real-world setting. Because of this, the biggest purists in the experimentation world even go so far as to say that best practices shouldn’t even be a thing that exists! I don’t go as far as that in my day to day, but to alleviate this, every single new idea is launched as a controlled experiment, so that tangible impact can be measured and controlled for. Where time permits, we also partake in qualitative user research where possible so that every hypothesis tested already has an initial round of validation. And of course, we rely heavily on quantitative behavioural data which enables 100s of micro validation loops even within a month. All this in aggregate helps us to adequately ensure that every idea is addressing an actual pain point that exists and is somewhat validated to a certain extent before we expend any kind of technical effort into it.

2. Can you describe the process ZALORA uses to translate validated assumptions into actionable product decisions?

This differs from team to team, but for my team in particular, every single assumption is first broken down into smaller, more addressable sub-components. There are a few reasons for this, (1) it makes the problem less abstract and more approachable, (2) it helps us to isolate impact better as each hypothesis is singled out, (3) we can launch features faster which enables for a higher velocity feedback loop with our users. The output of this breakdown would then essentially be a laundry list of hypotheses which we would then need to prioritise. Each hypothesis will then be discussed with our engineering team members for us to get a better grasp of the technical nuances of each. With this, we now have a more complete view of both the projected impact & projected effort of each hypothesis which we can then use to aid our decision making process in which idea makes the most sense to pursue first.

3. How does data analysis and user research play a role in solidifying these decisions?

Data analysis and user research play a dual role at two separate but equally important parts of the decision making process. The first is at the ideation phase. None of our ideas are borne from nowhere, each stems from either a data trend observed or a key piece of user research so we know our solutions are addressing problems which actually exist. The second role it plays is in the post-launch measurement phase, where we use both qual and quant data points to assess how successful the feature actually was, and to inform what we will do next.

4. How do you create a culture of experimentation where teams are empowered to test and iterate on ideas quickly? What metrics do you use to measure the success of these experiments and inform future investments?

Creating a culture of experimentation often just takes time and a few evangelists in the company who are actively educating and expounding on the benefits of experimentation. The reality is that it can often be an uphill battle to instill this mindset. Even now, 10 - 15% of my work week is spent on this education piece to ensure teams are executing experiments and thinking about it in the right way. As for success measurement, we mostly look at the conversion impact of each experiment relative to a control experience. We also have certain guardrail metrics in place such as profitability to control for the scenario where a strong topline conversion impact is realised but at the cost of our bottom line performance. For the most part though, we mainly look at conversion impact; as long as we have a new feature which is getting users to purchase more, this is usually considered a win unless the feature has other non-obvious subjective costs to it, such as harming our brand identity.

5. How does Zalora optimize its existing platforms to better serve customer needs?

We are constantly looking at data every single day to understand how our customers are interacting with our features. This data also includes external sources such as social media sentiment, app store reviews, NPS survey results, etc. We then synthesise all of this data, prioritise accordingly, and constantly ship out new iterations of our product to delight our users on an ongoing basis. We also try to address less obvious customer needs such as faster page load times and quicker API response times through continuous technical improvements. Users may not necessarily notice or grasp the impact of such enhancements, but they are certainly felt subconsciously and ultimately trickle down into healthier conversion rates as well.

6. How does this approach to tech investments translate to real-world business results for ZALORA (e.g., increased user engagement, improved conversion rates)?

Everything we do ultimately has improved engagement or conversion rates as the end goal. Achieving improved engagement or conversion rates, in turn, is usually a direct result of an excellent user experience and continuous technical enhancements which allow for a fast and seamless shopping experience. It can often be difficult to attribute real-world business results to such initiatives, which is why we always try to experiment as much as possible. We have even, in the past, AB tested different image compression percentages, and different CDN providers, to see which one hit the sweet spot between the fastest loading time and therefore fastest conversion, versus the lowest cost to administer and develop.

Hear from May Chin, at Equarius Hotel, Singapore, on 14th May, 12:55 PM: Keynote Interview: Redirecting tech investments, with a focus on business value – How can you move from assumptions to decisions when building winning solutions and products with minimum waste and maximum SAVE? Find out more here!