Customer-First �Company Building

Gaurav Oberoi

Why am I giving this talk?

What am I going to talk about?

With a special focus on winning over customers

Chapter 1: �Finding the Idea

My thesis

My thesis

There are step function improvements in AI, how can I build a killer B2B SaaS product using these new capabilities?

​

Joined the Allen Institute for AI incubator to tinker

What does the process look like?

1. Thesis (problem, for who, why buy, market size, how to get them)�
2. Lots & lots of customer calls (call script, list building, weekly quota)�
3. Build a prototype at the same time�
4. Some marketing sizing�
5. Traction? Keep going. Else back to 1.

Idea 1: Generate product photos instead of shooting them

Idea 1: Not technically feasible… but I got on TV!

Idea 2: Use satellite data to give businesses data

Idea 2: Hard to find value besides military & agriculture

Idea 3: Detect blood vessels & veins in ultrasound

Idea 3: Become R&D for a big OEM? Not my jam

Idea… 6?: Extract data from contracts

How did it happen?

• Inspiration from Emad’s wife: “we are doing this manually in procurement!”�
• Tech demo → can we actually do this? It’s pretty hard…�
• Start calling & testing with customers:
1. Finance
2. Insurance
3. Legal
4. Call center logs

​

Chapter 2: �Grinding to Product-Market Fit

Our legal PoC with Wilson Sonsini was a hit

And gave us the wind to raise

But we had cool tech, not a real product…

And law firms were not a great fit! But corporates were…

We started to test our thesis on commercial buyers

1. Called 3-5 potential customers a day�
2. Spent a lot of time validating the problem:
1. How do you manage your contracts today?
2. What happens if you don’t?
3. What does “manage” mean, can you show me?
4. …�
3. Then started testing solutions:
• First just mocks
• Then live product�
4. Always made sure they paid: time, reputation, or money

Got a decent first product, but it was not easy

• Lots of work to build the AI backend�
• Lots of work to build a good SaaS tool�
• LOTS and LOTS of iterating with customers �
• Consistent sales, but not hyper growth → okay product-fit

Chapter 3: �Scaling your business

The next stage really came from market pull

• Customers pulled us away from analyzing contracts�
• The real hair-on-fire problem: contract is SLOW, make it FAST�
• The market pulled us into a full-fledged workflow platform
• Lots of minor requests: “can I send an approval?”, “can I store a draft?”
• Lots of questions in our sales pipeline: “love this, but can you also help with negotiation?”
• Lots of ROI signal: “darned Legal slowed down my sales deal and I missed the quarter!!!”�
• That’s when we tripled the business in one year

​

Success came from building a customer feedback loop

1. Every deal won or lost → analyze deeply why, so we can do it better�
2. Every customer support ticket → why, and what are we missing?�
3. Strong company culture of:
1. Listening to our Sales team, and being honest about what we will or won’t do�
2. Tying our Product, Sales, and Customer Success teams at the hip, with weekly syncs�
3. Measuring success: “if we ship this, we must see usage / Sales / retention, or it’s a #fail

Parting lessons

Lessons learned

• Market size really matters → reduces risk and determines growth�
• You’re always testing market-fit, just as much as product�
• You can FEEL Product-market-fit almost like a physical force�
• Acquiring customers is HARDER than building a product for most companies�
• Iterate early and often WITH customers → don’t build in isolation�
• Persistence matters → it’s hard, buckle up!�
• You have to enjoy the ride → find balance, or you will go nuts�
• Don’t do it alone → I don’t know how solo founders (or their partners) do it

Thank you!

Q&A? Some ideas…

1. How do you find customers to call?
2. What are elements of a good discovery call?
3. What if customers tell me they love it, but I’m not getting traction?
4. …