Customer Interactions Shaping our Business
So far, the reception to Sentisweets has been amazing. We’re doing really well on Instagram and we’re growing slow and steady. The most promising aspect about it is that it’s so easy to explain- an Etsy for cakes.
We wanted to focus on independent bakers, especially in the wake of the “quarantine-banana-bread” trend. The idea passed the smell test, and I was able to quickly scrap together an MVP product. As I mentioned in my previous post, the marketplace homepage looks aesthetically pleasing and customers could easily order their desserts because we wanted communication between customers and creators to be universal.
We focused on phone numbers, because phone numbers are universally accepted and used. If we forced users to make accounts early, my rationale is that we would never be able to attract an audience. So we needed the ability to order desserts be as easy as possible. The best solution to this would be to allow creators to post their contact info and their Instagram accounts directly to the marketplace.
All of this should allow the lowest barrier of entry on both sides of the marketplace. It should allow creators to post quickly and with confidence that we wouldn’t be stealing their client base (like Grubhub is being accused of), and giving a direct line of communication- allowing consumers to directly text the creator- would increase the likelihood of customers ordering.
However, despite having a pretty solid Instagram following, a steady trickle of creators registering, and plenty of positive feedback from everyone about the idea and interface, we still were not seeing any posts.
I wasn’t sure why. Everything seemed to make sense. Our strategy was so simple, it was almost foolish. Yet no posts.
Finally, I decided that the time had come to get to the bottom of this. There was so much positive feedback from people about our sample desserts- the ones I had posted myself. The thing is that positive feedback doesn’t really help solve problems.
What I realized though, is that everyone that I asked for feedback was from the consumer-end of the marketplace. I was not asking the right question. The question that I had inadvertently been asking people was, “would you buy from this page?” Their answer was yes. But none of the people I asked were creators- they did not have any experience selling food over social media. Our goal was to make it easier for individual creators to sell food on the internet and yet we had no data about what creators liked, and what they didn’t like about the status quo.
After that, it was a simple matter of having conversations. We learned creators liked Instagram. They found the process of posting an image to Insta to be easy, intuitive, and useful with its filters and correction tools.
What they didn’t like was how their posts got buried in peoples’ feeds, and the fact that there was no uniform, direct way to post and sell items.
This helped me recognize our core value proposition for Sentisweets. We wanted to fix what they didn’t like about selling food on Instagram. What was even more important was realizing what our value proposition isn’t. We weren’t an image hosting/posting platform, and we were really bad at it already.
Our posting process was like Reddit, where we ask users to post their pictures somewhere like Imgur, copy the link, and paste it into our submission form.
Terrible. It sucked. Even with a “Preview” button, creators were unable to solve the problem of broken links, direct vs indirect links, and so much more. We should have realized this earlier. If they had this sort of knowledge, they wouldn’t have made an account on Sentisweets. They would have made their own website.
I scratched my head for days, trying to figure out how to solve this problem. I experimented with image APIs like Cloudinary, added more detailed instructions, but nothing worked. It was all still too difficult.
What really helped was the conversations that I had with creators. At one point, I told them to just DM me their details, and I would post it on their behalf. It helped our posts grow, but obviously that isn’t scalable. But I realized that creators were sending me screenshots of the Instagram pictures. They already liked those pictures, and they felt that there was no need to mess with perfection. (Although, obviously the screenshots were lower resolution images than if they had sent me the real image).
I realized I was reinventing the wheel. Like them, why was I messing with perfection? Was there a way that they could directly integrate their Instagram pictures into the site. I experimented with multiple ways, such as asking them to copy-paste the Embed HTML into the submission form, but that was so stupid and difficult. I looked into the Facebook OEmbed API, but even that was finicky and it ruined our page aesthetic.
Finally, I found a StackOverflow article explaining how to get images from an Instagram link. Instagram makes it really difficult to find the link, but there was a stupidly easy and foolproof way to do it with almost literally 1 line of code.
So we fixed our interface. Now, users could submit their images just by copy-pasting their shareable Instagram link to the submission form.
Within 12 hours, after notifying our creators about this update, our posts tripled. Now, we still have a very modest userbase, but at least now, creators have an easy, reliable way of posting their items.
So that’s step 1. Onto the next.
