Circles Last Blog

Circles: Shutdown

We have decided to shutdown Circles due to regulatory issues. Our journey lasted for a year and we are extremely grateful and thankful to our users, friends and family for their constant support.

Things we learnt:

• Work on problems which you can personally relate to and are passionate about. In our case, even though discounts and savings add great value to end user, we were personally not that attached to the discounts space (but we still gave it our all since we loved creating a product).
• Product design is extremely important and it takes a lot of hard work and patience to get it right but it matters only when you have nailed the core solution.
• Distribution is absolutely critical and ideally it should be an inbuilt attribute of the product itself. Just building an app for anything is not enough unless its for fun.
• If you build a product without visibility or clear plan to make it 10 to 100X better than substitutes or competitors then you will always be firefighting. Improvement can be measured in terms of performance or cost or both. It might not happen right off the bat but be aware of scenarios when these conditions will be met.

Things we are proud of:

• Circles got us incubated at CIIE on IIM Ahmedabad campus under IIMAvericks Fellowship program.
• We won GC Mital Entrepreneurship Award in 2019 which is awarded to one graduating student in the batch.
• We won Y Combinator SUS 2019 grant which was awarded to 41 startups out of 25000 startups globally.
• PayU offered (informally) to acquihire us.
• We were pleasantly surprised by the widespread interest from the VC community but we made a conscious choice to not go after funding till we were absolutely sure of a path to profitability (turned to be the right decision).
• All the work was done by me and Abhiram with additional support from few interns. Nobody lost any jobs due to Circles shutdown. We believe in the power of smart and small teams.

But still, why shutdown:

• We’re certain that Circles adds value to users by enabling them to use credit cards they do not own. A critical use case for users was availing offers on e-commerce websites on credit cards they didn’t already own. These offers are funded by marketing budget of banks or e-commerce companies to gain customer loyalty but Circles solution has conflict of interest with these players.
• We think technological hack is not possible due to high security and encryption on payment gateway pages. Even if we do figure out a workaround, we cannot offer uniform experience to users across platforms.
• Most importantly, Credit Cards have very high fraud and security risk associated with them. Terms & Conditions of use and security standards (PCI DSS) do not allow for credit card sharing. Thus, making it extremely difficult to build a solution which can work at scale for millions of users. Our best solution was to tie up (become payment method) with other payment gateways but that cannot happen due to regulatory issues.

If anyone is really eager to still attempt to solve credit card sharing, these are things we tried but didn’t work:

• WhatsApp groups and excel sheet database on IIMA campus. It works only when trusted intermediary is involved or two people already know each other. It is slow and involves a lot of repetitive steps. We did it as MVP.
• Built Android App and did anonymous matching of cardholder and discount requestor (similar to how Uber matches you with unknown driver). Trusted intermediary needs to hold payment and resolve disputes. Lots of data needs to be exchanged between two parties to complete transaction which makes it slow and tedious because cardholder won’t risk sharing card details and has to complete transaction for other person. Has privacy and trust issues. Still slow. High transaction drop rate. Unit economics unviable due to payment gateway charges and GST. Not good enough to scale.
• Built another Android App but this time to do matches and discovery between people who already know each other or their 2nd degree connections. Our hypothesis was that if only credit card details are exchanged and not all the details related to purchase then it would be much faster. We assumed that an end-to-end encrypted channel and the ease of sharing card details with others would convince users about the value-add. Trust of users is not swayed by a secure and convenient chat — its dictated more by in-person or on-call interactions. For someone else to truly crack this use case, they’d have to figure out a way to share card details without the requestor getting to know the card details being used for the transaction.

So, what next:

We’re not sure yet. We’ll probably take a few weeks to decide what to do next. We’re always game for ideas or brainstorming, so feel free to reach out.