After much discussion over the past few days, we've decided to close down operations at the end of the month (September 2023).
We set out to make an AI-enabled spreadsheet experience that works "like magic", and we know this news is disappointing to many of your found our tool useful — we're disappointed too, and your support meant everything to us.
We'd like to thank our customers, friends and investors for their support, and we sincerely hope to have the opportunity to serve you again in the future.
We are no longer accepting new customers or upgrades.
Existing users' plans will continue to work until the end of their respective billing periods, then cancel automatically. There is no need to do anything from your end.
If you'd like to cancel your plan immediately and receive a pro-rated refund, send us an email at [email protected] and we'll organise this for you.
Past invoices can be found through the customer portal.
There are largely two reasons: technical and personal.
Flowshot started out in life building bespoke e-commerce AI models for several enterprise customers.
Using our stack of in-house models, we spun out what became one of the first AI add-ons for Google Sheets. We spent thousands of hours optimising our AI stack for near-instant generations with zero downtime — it really felt like magic.
However, we received a lot of requests from our users to switch over to OpenAI's GPT-3 model (at the time), and we eventually made the transition.
While the quality of generations improved, we quickly found our users were constantly hitting rate limits. We attempted to open a dialogue with OpenAI and applied multiple times for increased quotas, without ever receiving any response.
Our user base and revenue grew exponentially, but unfortunately, so did the amount of complaints about rate limits. We doubled down on trying to make contact with OpenAI, without any success.
During this time, dozens of competitors had appeared, and transitioning back to our own in-house models or to another provider seemed impractical.
Eventually, the ongoing technical issues and customer complaints wore us down, and we (Chris and John) came to terms with the fact that both of us were burned out.
Although this wasn't first time in our 2+ years of operating the company, we ultimately agreed that pushing through as we have in the past didn't seem like a sustainable approach — and it wasn't fair to our customers, our friends, and ourselves.
An in-depth retrospective is on the way, detailing the genesis of Flowshot (and what came before it), the challenges we faced, and the reasons for closing down operations.
Looking back, we learned the importance of controlling our technology stack, rather than being locked into a third-party vendor — despite the possibility of quick wins.
We're taking some time off, however we plan on working together again on new products in the future.
Thank you again — really.
Although this chapter is ending, we look forward to applying the lessons learned to make an even bigger impact in the future.
Have a fantastic rest of the week and we'll follow up with another blog post soon with a full retrospective.
- Chris and John