The Backstory

It was the mid of 2022. Our food delivery business was almost profitable. The company’s vision was to expand it to new avenues. A new team was setup — New Service Offerings (NSO) to build, experiment and find product market fit for new categories.

From there on, I led the design of this stream for around a year. We built 3 major products — Pre-order breakfast, Occasions, Super Saver.

To be honest, initially I was a little scared. I had no experience designing zero to one products. I was thinking What if something we build doesn’t work? How would we bounce back? Worst, what if it still doesn’t work?

I discussed this with my manager Saptarshi Prakash. He was aware of my side projects and reminded me how this wasn’t very different. That was some assurance. I also scraped the internet for videos, articles, etc. that could teach me about zero to one products.

Defining People Outcomes

One of the products we built was Pre-ordering breakfast. Order share of breakfast and morning traffic were consistently low. The first step was identifying the people problem. We wanted to define the user needs, their pain points, current workarounds, etc.

The Challenge — Limited Data

As this was a new product, we didn’t have much quantitative data. Our plan was to reach out to consumers directly. This was an opportunity to conduct user interviews, prepare questionnaires and analyse the insights.

We discovered 3 problems around breakfast meals:

1. Time crunch in the morning: Because of which people skipped breakfast or switched to packaged food.

2. Affordability: Ordering value-packed, small-portion meals wasn’t very effective with the delivery charges.

3. Limited variety: No easy way to find regional cuisines, healthy food like protein shakes, juices, etc.

Defining An Initial Target Audience

The problems were defined. And we had a good idea about eventually solving them for a large population. But for zero to one products, the goal was slightly different.

Start with a small audience of early adopters. Get them to love it, and then expand to more people over time.

The hypothesis was — Pre-ordering breakfast one night before would solve for time crunch in the morning. Although I was a little skeptical about this. Planning next day’s meal is an effort and people might forget about it. However, the team thought let’s validate it in the real world.

The Minimum Viable Product

We’ll come back to Pre-order breakfast. Another product we built was Swiggy Occasions. When ordering food in bulk, say for a party, we identified problems like:

1. People were looking for a variety of brands, cuisines, price range, etc. Finding such restaurants or caterers was hard.
   
   

2. There was no accountability on the quality. People had to rely on word of mouth or previously hired caterers.

The Challenge — Time to Insight

Finding PMF meant building and iterating until we achieve the people outcomes. We wanted to pick a narrow target audience and design an end-to-end experience for them. A focused Minimum Viable Product. 

Speed and development effort were important. We didn’t worry too much about things like creating a new type of listing or fancy animations. At the same time, we didn’t want to ship a subpar or confusing experience.

A technique we used to create UI was using MxN images. This was a powerful way to quickly test any layout, content or illustration style. We could ship any update with zero tech effort or app release.

For instance, almost everything on the onboarding screen is an image. Yeah that’s right — even the bullet points, and CTA button. 

Validate or Learn

Coming back to Pre-order Breakfast. We launched a pilot in two regions of Bengaluru. Metrics like orders per day, retention, drop-offs, were defined. The idea was to keep track of any big changes in user behaviour. After a couple of weeks… orders per day weren’t what we were hoping :(

At this point, you have to be ready for quick responses or turnarounds. We reached out to a few users to understand what they thought about it.

“There weren’t enough options available. I don’t like South Indian food that much, and there was no way available to order other things.”

“I usually think about my breakfast in the morning only, that’s when there’s a need for ordering. At night, I forget to schedule the order.”

Lack of variety was a valid concern. We explored a few options keeping tech effort in mind. After some discussion, we aligned on showing different restaurants as tabs. We shipped this in literally 2 days!

Drop-offs in the funnel reduced, but we still didn’t have as many orders as we wanted. This was very disheartening! I talked to some people in the team and realised that we can’t always show that a product is successful. Sometimes, we just need to learn from it, and that’s also a good thing.

We also launched a pilot for Swiggy Occasions in Bengaluru. At first, we got a decent number of orders, especially for big parties, which don’t happen very often. We saw a huge spike during the festival of Onam, which showed us that people liked it.

We also made improvements based on user feedback. For example, we added the option to order on the same day, which we couldn’t do initially because of our supply. Right now, we’re making it even better, using festivals to make it more popular, and bringing it to more cities.

Bringing It Home

We used the same mindset for our third product— Super Saver. For value conscious or new users, online food ordering is perceived expensive. Our idea was to onboard restaurants which provide a combo of discounted price, free delivery and zero packaging charges.

Usually these users didn’t order frequently. The incentive for restaurants was increasing the number orders. We launched a pilot of Super Saver in a couple of locations in Bengaluru and Gurgaon.

The pilot did exceptionally well. In both the locations, we saw orders per day increased by ~30% for this segment. Once it seemed to find a product market fit, then came the next set of challenges.

Challenge 1 — Designing A High Quality Experience

As we moved quickly in the PMF phase, we had to make some tough choices and trade-offs. Now, the challenge was to create a really good experience and build things the right way.

Value seekers were usually reluctant to order. So we wanted to boldly callout the value everywhere. The idea was to weave a brand story around the product. Thus born the name, Pocket Hero!

We kept a central character at different touch points in the flow. Big thanks to Ezaz for that. This also included a lot of collaboration with product marketing. Thanks to Shalini for helping with the copywriting.

Designing 0→1 products you get to wear a lot of hats. Ownership to put creative ideas, define brand guidelines, set the tone with copywriting, etc.

We A/B tested both the approaches. The test group seeing Pocket Hero got more orders per day and higher conversions. Although the difference wasn’t huge, we felt that Pocket Hero could provide a platform that we’ll could build upon over time.

Challenge 2 — Overlap With Other Products

In the PMF phase, we don’t stress about it. But later on, the challenge is to figure out how the product fits into the whole app experience.

Let me give you an example. For people who order less frequently, usually once a month, we were redesigning our food page. Bringing noticeability to offers, popular outlets, etc.

Almost half of low frequency users were actually value conscious — the target audience for Pocket Hero. After a lot of cross-team discussions, we decided to build two versions. One retaining the Pocket Hero communication, and another for the rest of the users.

What’s Next?

At this time, I am transitioning to a different team with some new responsibilities. At NSO, the plan remains same. Experiment with different ideas and try to find PMF, making the current products more valuable to existing users, and unlocking them for the next set of users.

You never know, something awesome might come up soon. Before concluding, I just wanna mention the best takeaway from this journey:

Product discovery is about strong opinions, weakly held.

On one hand, we try to work with the best possible hypothesis. On the other hand, we need to be a little skeptical — constantly open to feedback.