Blog

In 2017, Donald Trump had won the US Presidential election, and echo chambers were the talk of the town. This inspired friends and colleagues Vipul Chaudhary and Jani Pasha to work on building a news aggregator platform called The Soup.

The duo spent close to five months building that product, but never paused to speak to the target users to understand their reading habits, how they consumed news, and where they got their news from.

“Eventually, when The Soup didn’t work out, we decided to reach out to people. We realised most people in Tier II and Tier III towns weren’t getting information about their areas from existing apps. These insights led to Lokal,” Jani says.

Most people heavily relied on local newspapers to get updates in their own language. However, the content was limited – the number of pages were few and the news would reach readers the next day.

“Meanwhile, the rest of the world had instant news updates at their fingertips. This is when we decided to build a platform that would make relevant content digitally accessible to this population in their native language. From a single-use case, we have now moved to a multi-use case format. Addressing the issues of 200+ districts, we have successfully built a hyperlocal platform that enables masses to interact, exchange information, and unlock digital opportunities,” Jani says.

The hyperlocal content startup, headquartered in Bengaluru, offers local news, classifieds, matrimony advertisements, job listings, and more in regional languages.

From WhatsApp to Lokal

Things started on WhatsApp. In February 2018, the team started a WhatsApp group to distribute local news sourced from a stringer in Kodad, a town in Telangana. That one group turned into many quickly as more people kept joining in.

“In May 2018, we released the first version of the Android app. From a single WhatsApp group of 10 participants, Lokal had grown to become a significant player in the hyperlocal ecosystem,” Jani says.

The founders had visualised Lokal as a local updates platform when it was launched. Hence, it was a simple chronological feed. The first product was a result of a lot of user and market research.

The simple, single scrollable feed of news and information available in a language users were comfortable with received positive feedback.

“Through our solution, users were not only getting news in vernacular language but also local updates. Users liked our first product primarily for two reasons: it was a convenient solution as it was a single, scrollable feed similar to YouTube to check all the latest content from their location, and they could quickly the latest updates in their language,” Jani says.

Within a year, due to increased demand and user understanding, they added two more hyperlocal use cases: job listings and matrimonial profiles.

At every product development stage, the team focused on unmet user needs. They would test out new features in smaller areas, in ways that did not disrupt the established user flows.

“Once we were sure, would roll the features out to a wider audience. We believe the speed of developing products is paramount as with more iterations you learn more and create a truly scalable product,” he says.

Building the MVP

To build the MVP, user understanding and problem statement were crucial to “keep the solution aligned”.

“Our learnings through the process were that you can think of automating after you have addressed the challenge. An example is our moderation of content where every content piece would get moderated manually before it went live. Now, 95 percent of our content is auto-moderated according to our community guidelines,” Jani says.

He adds Local has two sides: creators and consumers. From the MVP in small regions, the team identified that demand for the product was high but supply would be the key to keeping the demand growing and engaged.

In the initial phase, the team put a lot of effort into getting the supply right and that is when they identified the need for technology-heavy tools for creators, a moderation process,  and auto-approval.

“We also realised that a simple feed with content won’t work for diverse use cases and problems at the hyperlocal level. As we scaled up, we identified new user needs: jobs, matrimony, and real estate. Thus, we curated an architecture that assured multiple and diverse use cases critical to our growth. At an early stage itself, we recognised that a manual moderation layer would not be able to support the pace of our growth,” Jani says.

This is why the team ensured that they build the product with a user-feedback loop and ironed out any potential gaps in the solution. As use cases increased, they also doubled down on personalising content for the users.

“Staying in touch with the user through on-ground meetings and constantly monitoring reviews and support channels helped us identify the gaps in our product and simultaneously find the next innovative solution,” Jani says.

Today, one can easily map the entire online journey for an urban user, right from accessing banking services to ordering vegetables.

However, Jani says non-English users from Tier 2 and beyond locations do not have the same luxury of access. To solve this, Lokal is working towards becoming a one-stop solution for all hyperlocal needs, in your own language. We strive to learn a bit more about the internet consumption behaviour of this segment every day, and plan to keep adding new use cases as we discover more,” he says.

In June 2019, as Lokal crossed one-year mark, they had a Minimal Viable Product providing updates, jobs, and matrimony to an active user base of 500,000 users across 15 districts of Telangana and Andhra Pradesh.

Lokal became multilingual in January 2020 when it was launched in Tamil Nadu.

“At present, we are a 200+ strong team and the Lokal app has 20 million+ downloads along with a monthly active user base of six million+ users from across 180+ districts in eight states,” Jani says.

Taking on the challenges

One of the eureka moments the team had in the process was that no product in India had a creative flow that let users create different types of content like updates, news, classifieds, blogs, etc. The team worked to involve and generate diverse formats of content at the hyperlocal level from users.

Lokal encountered several challenges such as lead time for launches, depth vs breadth, and ease of UX, but rallied together to resolve them.

“One of our primary challenges has been reducing the lead time on our state launches. We believe a successful state launch has two critical aspects – translations and optimal supply of content in all use cases. Building systems that would help us quickly expand geographically was a challenge. But with time, we have figured out a playbook to reduce our lead time to less than two weeks with the launch of Lokal in eight states,” Jani says.

When it came to overcoming the depth vs breadth problem, the team realised that being a multi-use case product is challenging in the initial building stages. Maintaining the balance – whether to expand on existing use cases or test out newer use cases – was a major challenge.

The product team took calculated risks backed by user research, data, and undying conviction. Building the systems in configurable ways allowed for quick experimentation.

Jani says Lokal has been able to do short experiments like gaming, quizzes, Q&A forums, etc., and build larger use cases like jobs, matrimony, classifieds, and more in a span of less than 1.5 years.

“Lastly, discovery and navigation of multiple use cases and differentiated user experience was also a major challenge for us. Since Lokal is based on the premise of user-generated content, it was challenging to build systems for both, creators and consumers, especially with such varied types of content – news, jobs, classifieds, personal wishes, ads, etc,” Jani adds.

Working on different playbooks

Jani says the Lokal team has let go of several inhibitions and belief systems in their journey from idea to product.

Earlier, it was believed that the market they operate in was tough to crack in terms of revenue. As they understood user points and behaviours and built products for them, the startup was able to scale, in terms of revenue and growth.

“Our playbook approach has helped us adapt to user needs in different states along with being agile in our launches, leading to tripling the revenue within the last year. One of the biggest learnings for us has been that building for the next billion users requires significant attention to the user experience,” Jani says.

He adds that “one cannot simply lift and shift a product from Tier-I cities”; it needs to be localised and made relevant for the target audience.

An interesting observation is discovering how digital behaviours and interactions have become a daily practice for users. Be it scanning QR codes, making online payments, swiping interactions, online feedback forms, or content creation tools, users are open and curious to adapt to digital products. 

The next steps

Some of the key innovations the team is working on now are:

Integrating technology into language: Content optimisation is key for a platform like Lokal, where the team understands and analyses consumer needs and consumption patterns and designs the right content accordingly. Owing to the feature-rich corpus in English, one can get a number of features, models, prebuilt sentences, tagging, post tagging, etc. But to conduct natural language processing in local languages, there is no completely cohesive corpus. This is one of the biggest challenges.

Lokal has approximately 10,000+ user-generated content pieces getting added and tagged daily. This involves segmenting a particular content piece in the correct category for a correct location. “This helps in leveraging the content to build technologies so that we can have a significant language corpus. On account of the amount of original content the team receives and generates on the platform, they have the ability and capabilities to build a local language corpus,” Jani says.

Localisation as per use case: The technology helps to scale at the top level and also personalise the feed for each and every person to receive district-specific content. This format can be replicated across languages, states, and towns. Each and every use case has multiple implications for localisation. For instance, the matrimonial category is not location-bound, and therefore the radius for this service is much larger than a job search. Jobs are localised to a district. All these algorithms are set as per user preference and vertical understanding.

Scaling up for growth: Lokal has a bottom-up approach. “We launch at one location, get a density of people, follow our use cases, and move on to the next. We require technology to replicate and scale the model. The technology we have built is for scale and customised feeds. As a result, we moved to microarchitecture wherein each and every feed along with every use case is a service in itself,” Jani says.

Lokal aims to become a super app, which will work as a one-stop solution for all hyperlocal needs of Tier-II and III non-English speaking users.

“While we continue to build on our existing use cases, we are also laying the foundation for the next set of innovative solutions. We function in an ever-changing paradigm and being a multi-use case app, we understand that the one-size-fits-all approach will not work with our diverse user set. This pushes us even further to understand the specific needs of a user and ensure they are able to derive maximum value from our platform,” Jani says.

Personalisation using AI and ML models and extending to each use case will be a major focus area for Lokal. The team also plans to focus on enabling social commerce opportunities for SMBs, by aiding them to offer enhanced customer experience.

“With an exceptional growth rate of active users and district-level penetration in over eight states, Lokal serves as a highly beneficial tool for businesses to acquire, communicate, and advertise in untapped regions.”