Disclaimer

This is not a technical talk

but I hope you'll find it interesting

"No Bullshit" Guarantee

Failures is what we learn form

Say no to the term "Big Data"

This talk comes with a

About me

Gabor Ratky

CTO at Secret Sauce Partners

2003

2007

2009

About Secret Sauce

Founded in 2010

SV startup based in Budapest

B2B apparel e-commerce company

The early days (2010-2011)

We had no idea... literally

We were 2 founders, 4 engineers

Let's build multiple products in the first year

We identified large, underserved markets

We talked to people to hear their pain points

We validated that it was a real problem that most people had

We built a minimal product that could solve the problem

Fit Predictor (2011)

Only 10% of clothes are sold online​

Return rates are very high (20-40%)

Major problem for retailers and customers

We weren't going to be the first ones to try

Let's use the retailer's existing data to make size recommendations

Chicken and the egg problem

Came up with an initial algorithm (R&D)

Launched the product before writing any code

Worked with a small retailer to build the product

Fit Predictor (2012-2014)

US launch in January 2012

We do usability testing continuously

The only thing that matters is our impact

Being a third-party vendor is hard

We work with custom feeds, do custom integrations

Apparel retail companies are (usually) not technology companies

To scale our business, we need to onboard partners fast

Trade-offs that made sense early caught up with us later

StyleFinder (2014-)

Our R&D team developed visual recognition
algorithms that extract features from images

We can identify high-level features of clothes

We can build amazing experiences on top of it

Market and user feedback has been amazing

Launching later this year with one of our existing partner

We minimize the effort to launch as we don't know its impact yet

AMP (2014-)

Vision: we help retailers merchandise apparel

Two of the many problems we can solve

Data and website integration is the same

We can standardize our feeds and use common
services for analytics, data ingest, user tracking, etc.

We are moving towards a more loosely coupled system as we grow

Treat all data as a realtime stream of logs (feeds, events, messaging)
and let any service consume it irrespective of time.

Today

5 years later, we're 20 people, 10 engineers

Exceptional and diverse team

The data we have is mind blowing

We have an amazing vision that we're executing

Parting thoughts

Working at a growing startup is demanding

It is also a lot of fun and full of opportunities

The only thing that matters is people

If you find you grow the most when you're
being challenged, you should work at a startup

Trust your instincts on whether its a match