sammy

foreword

this series is going to be a weekly log of what i am doing in order to get my startup ideas off the ground. the purpose of this is to refine my ideas, reflect on experiments, and share with others to get feedback and hopefully help someone doing something similar.

background

for some quick background, my name is sammy, i studied mechatronics engineering, and am now working at a small hedge fund in toronto as a software engineer. outside of my work, i have worked on a number of projects including:

ParaSight: a computational malaria diagnostic tool
ApertureMD: ai-native electronic medical records
Sonic Slides: an ai tool to rapidly edit your PowerPoints

at the time of writing this, i am exploring fashion tech with two close friends. we have primarily explored inefficiencies during the garment design process, as well as bottlenecks in fashion production.

operating under conditions of extreme uncertainty is a defining characteristic of startups. working in an unpredictable environment is challenging because you lack clear signals about whether your actions are pointing in the right direction. to mitigate this as much as possible, i will be adopting a more scientific approach to my ideas. inspired by the Lean Startup, i will ask myself:

what do i want to learn? what do i need to measure to learn that? what can i build in order to get that measurement?

right now we are looking at fashion production coordination; this describes the process of managing all of the moving timelines and information flows to get a product from a tech pack (like a CAD file for clothing) to the final sample. it is an arduous process that involves a lot of manual tracking, manual input, ad-hoc spreadsheets, email handling, and many more tedious tasks.

one fear we have is that we may have over-indexed all of our customer interviews on this one specific role in the entire fashion process. while many of the junior production coordinators we spoke to mentioned how annoying this process can be, we questioned whether it was something that the decision making executives actually cared about.

now that i have laid the groundwork, we can get into the first experiment we are running.

experiments

hypothesis

do leaders and executives in production/operations/logistics perceive production coordination as an urgent, strategic pain rather than just an operational annoyance that has become the industry norm?

the signals of urgency will include things like:

do they engage at all when shown the problem or solution?
do they treat it as a strategic improvement (speed, cost, scale, risk, etc) instead of just a part of “their team’s workflow”
do they respond in ways that show interest in the subject matter

method

the team and i are going to send cold emails directed at the executive problem level. additionally, we are going to attach a small demo that presents a potential solution at the biggest pain point for production coordinators: tracking and email management.

the process

we created a profile of the type of fashion brand we want to reach out to. this includes mid-sized business that does revenue on the scale of $10M-$100M. we figured these kinds of businesses would have more flexibility to look into tech solutions.

we primarily used Exa’s websets to scrape as many emails as possible. then used a combination of LinkedIn and RocketReach to find the emails of specific people that Exa was not able to locate.

using Exa to find executives at Aime Leon Dore.

next, we built a quick demo in Next.js that demonstrated how we would use ai to react to inbound emails, and how we would manage them accordingly. we would attach this in the email we would be sending.

watch the video on the original post

the last step is to just blast these emails out. we created a standardized email, with a touch of customizability at the beginning of the email. that way we can move quickly while being personable.

result

hopefully in the next post i will have results to share about this experiment. the goal is to set up some calls with executives, or on the flip side get our signal that no one really cares about this problem. either way, stay tuned for the results next week.