How it evolved from a search engine to a social city guide
Dunro is a place for discovering local businesses around you in Iran, there are more than 500K local business data available at Dunro, mostly from Tehran and then from 5 top cities of Iran.
As the product designer I was responsible for creating a whole new digital city discovery and social experience, from ideation to production, across iOS, Android and Web.
When I joined Dunro, it was a simple search engine for local businesses, which had inherited its data from its parent company, Bartar Media Group who is an offline advertising giant in Tehran. By the time there were about 120K local business data inside the platform which was mostly inaccurate or outdated.
The revenue was meant to come from showing banner advertisements! beside the very raw revenue model which was not a very clever one, we had a huge problem, most searches result in wrong, out-dated or no data at all.
Although we had a button for adding new places, but we needed more users to contribute to data, and to have more users we needed more accurate data.
I started to gather insights from similar businesses, I studied the product evolution of services like Yelp, Foursquare and some other competitors. We lacked so many infrastructures, basic features and most importantly we needed to solve our not-enough, not-clean data to solve that chicken-egg problem.
I had a lot of meetings with our stakeholders, CEO and Bartar Ads department to help them clear, refine and modernize Dunro's product vision and business model.
They obviously wanted to build a great place for selling their advertisements, but had no clue about how to gather people around Dunro.
I made and showed them a concept design of the possible future of Dunro, I was leveraging curated editorial data, social connections inside Dunro, and multiple smart and targeted yet unobtrusive advertisement methods on the concept design.
I was successful to bring them along my idea.
My idea was to start with offering quality and curated content, and using paid manpower to add initial data to attract more users. I also wanted to build basic social foundations to let people more interact with Dunro.
We quickly started to recruit our content marketing team to create curated content about the city, venues and attractive experiences.
We also took advantage of the Bartar distribution crew which was consisted of 1000 men marching into every neighborhood, street and ally to distribute weekly ad journals. They were assigned to add every local businesses they see during their patrol.
I started the first step of redesign by changing the information architecture to make room for editorial contents, I also designed a rate and review system to let people interact with Dunro.
By releasing this version, our user retention rates started to increase, but we needed more to do.
I believe that products like Dunro needs to build a need for users, specially in Iran which the majority of users are new to Internet and still haven't tied their entire life to it.
We conducted a series of focus groups and one by one interviews with users and received their opinions, needs and priorities.
We identified that people would only use Dunro when they need to search a specific business. Most of the time Google will show them better results instead.
They have never thought of contributing in Dunro's data, they simply didn't see any incentives to do that.
I followed the BJ Fogg's behavioral model. We wanted our users to actively add new places, suggest edits and leave comments and rate on venues. By translating our targets to behaviors we could shape the product to meet the requirements for those behaviors to happen. So we needed to motivate them, make those actions easy to do, and finally trigger them.
I categorized the idealistic core motivations Dunro can offer into two types:
So I begin to do a competitive analysis on successful gamification platforms in Iran like Swarm and Waze to validate my hypothesis.
Then I started to sketch ideas, making different ways of interaction between users and the app:
Finally we decided to bring these features to 10 percent of our users:
By the time our new Rate and Review system, the overhauled Add New Place and Edit Suggestion flows have been designed and implemented and after some user tests, and doing some tweaks on them, they were ready to release.
We chose random 10 percent of our users and rolled on the update for them, since we didn't have time and needed to act fast, that was our only option to test our new features.
We waited to see the reaction, for the first weeks it was actually tremendous! user engagement was lifted 47%. Then the negative sides happened. Although we were expecting that fraud would happens and we had tried to set smart rules to confront them, but users been more creative than us. So we got a lot of new insights about how people are reacting and how they beating our limits.
So we started to learn from them and make our rules more smart, rational and also dissuasive.
After carefully curating every detail, we launched the new Dunro for everyone. It was followed by an extensive marketing campaign encouraging people to make a better city by collaborating, expressing themselves and being a better citizen. We also ran a podcast, named DunroBuzz covering topics about exploring the city, food, culture etc. which got so popular.
After 6 month from our public release our daily active users rate has been doubled. We had 500K downloads and our daily new places went from 5-6 a day to 400-500 each day.
We had given more than 18 million coins to our users in just 6 months. Our rate and reviews increased by 28% and we earned 1B Rials with our advertisements program which has about 3 times increase.
Dunro Bazaar was offering discounts and savings on different platforms (like ride hailing apps, Food ordering apps, Cinema tickets etc. after about 9 months those offers lost their attraction for our users and we learned that those incentives was not powerful and durable enough. So we decided to involve local businesses.
Check-in feature didn't get enough attention and by further interviews we found out that users still doesn't find any reasons to do that.
Our junk data and fake new places rate was gradually increasing that led us to design a superuser program to use our best users' power to audit them. You may read the case study here.