,
25 tweets,
6 min read
Read on Twitter
@VernacularAi’s 🇮🇳 story, reaching $1MM SaaS ARR with 3 customers in 8 months in India.
This is our journey of how we turned the tables and fought our demons to up our B2B game. 👇
1/n
This is our journey of how we turned the tables and fought our demons to up our B2B game. 👇
1/n
Our vision was clear: Build the infrastructure to empower billion Indians to communicate
We were building chatbots for businesses in Indian Languages.
Timing was perfect, hype was right. We built a stellar team, found great investors and had excited customers
2/n
We were building chatbots for businesses in Indian Languages.
Timing was perfect, hype was right. We built a stellar team, found great investors and had excited customers
2/n
A few customers later, we realized something.
While the market was ✅, the medium was ❌
Keyboards/scripts in Indic langs are too complex ⌨️
Voice is a rather intuitive/faster medium 🔊
Also, our customers had started asking to use our tech to automate their call centers
3/n
While the market was ✅, the medium was ❌
Keyboards/scripts in Indic langs are too complex ⌨️
Voice is a rather intuitive/faster medium 🔊
Also, our customers had started asking to use our tech to automate their call centers
3/n
As @justinkan puts it, a B2B customer has two questions, and we gave our best shot to answer them before changing gears.
1⃣ Did it solve the problems faced by our customers?
2⃣ Will they pay for it?
4/n
1⃣ Did it solve the problems faced by our customers?
2⃣ Will they pay for it?
4/n
1⃣ Yes
2⃣ Yes
The overhead associated with running a contact center is India is HUUUUGE.
High attrition rate, super high and changing customer expectations, etc.
5/n
2⃣ Yes
The overhead associated with running a contact center is India is HUUUUGE.
High attrition rate, super high and changing customer expectations, etc.
5/n
You feel the pain if you have managed one.
We did the math and found that even a small percentage of automation in contact center (CC) can have a significant impact.
6/n
We did the math and found that even a small percentage of automation in contact center (CC) can have a significant impact.
6/n
We didn’t want to make too many changes initially. We theorized that we could use a commodity speech to text engine with our chatbots and make things happen.
7/n
7/n
Turns out, voice isn’t chat. 🗣️ != ✍️.
Watch out if someone says otherwise. 👀.
It was time to go all in.
We decided to focus all our energies on building voicebots for contact centers.
8/n
Watch out if someone says otherwise. 👀.
It was time to go all in.
We decided to focus all our energies on building voicebots for contact centers.
8/n
Pivots have to happen out in the open @stevesi
One fine day, we put our foot down and decided to get out of the chatbot business. What did it mean for us?
- Going from $300k+ ARR to 0 instantly 📉
- Discontinuing engagements with 10+ customers 🚫
9/n
One fine day, we put our foot down and decided to get out of the chatbot business. What did it mean for us?
- Going from $300k+ ARR to 0 instantly 📉
- Discontinuing engagements with 10+ customers 🚫
9/n
It was one of the hardest decisions I have taken in my small entrepreneurship journey.
In hindsight, it was a critical decision that turned out to be rewarding.
10/n
In hindsight, it was a critical decision that turned out to be rewarding.
10/n
It was also important to communicate with the team on why discarding our year-long effort makes sense. (Sunk Cost Fallacy)
You can be wrong on a lot of decisions if you are really right on a few huge ones - @sama
11/n
You can be wrong on a lot of decisions if you are really right on a few huge ones - @sama
11/n
We have been able to do $1Mn ARR in 8 months with a pipeline to follow T2D3 track, golden rule of SaaS scaling.
And all this is done by a team of 14 in Tech/Product/Sales in <$250K, not to mention the solid IP we have built in terms of tech/data
12/n
tcrn.ch/2JikUXE
And all this is done by a team of 14 in Tech/Product/Sales in <$250K, not to mention the solid IP we have built in terms of tech/data
12/n
tcrn.ch/2JikUXE
Here is what worked out for us👇
1⃣ Laser sharp focus
2⃣ Incredible hustle
If that sounds obvious, the real requirement is to get the entire company in this mode, all the time. Achieving that helped us ramp up. ⏩
13/n
1⃣ Laser sharp focus
2⃣ Incredible hustle
If that sounds obvious, the real requirement is to get the entire company in this mode, all the time. Achieving that helped us ramp up. ⏩
13/n
Focused execution was key, knowing exactly what to achieve when, our "vector clock" 🧭⏲️ approach.
Realised that domain knowledge is important even though every call center is the customer. We focused on only big enterprises in 2 verticals instead of every one out there.
14/n
Realised that domain knowledge is important even though every call center is the customer. We focused on only big enterprises in 2 verticals instead of every one out there.
14/n
Voicebot required us to build the core, multilingual (10 langs) speech to text engine. And a lot of work on Spoken Language Understanding.
Again, our customers wanted us to use these technologies to do umpteen ♾ other things, but we remained focused on the voicebot.
15/n
Again, our customers wanted us to use these technologies to do umpteen ♾ other things, but we remained focused on the voicebot.
15/n
While we got our first customer early, the real challenge started after we deployed our product.
The billing became stagnant and we realized sitting in office and trying things out wasn’t working out. We decided to get to the root of the issue.
16/n
The billing became stagnant and we realized sitting in office and trying things out wasn’t working out. We decided to get to the root of the issue.
16/n
40% of our team started working out of customer’s call center to understand their inner workings. We even became their agents for days and answered their calls.
17/n
17/n
2 weeks and ~100 experiments later, we were able to turn the tables.
Not only did we end up increasing 📈 the billing significantly, we also streamlined their CC operations.
18/n
Not only did we end up increasing 📈 the billing significantly, we also streamlined their CC operations.
18/n
For deal #2, we were competing with a big org 🦈 2000x our size
The customer asked for a POC, and we didn’t do justice to it. The feedback was not something you want to see on a Monday morning.
19/n
The customer asked for a POC, and we didn’t do justice to it. The feedback was not something you want to see on a Monday morning.
19/n
We asked for a 2nd chance and initiated a company lockdown. We went from a 9-9-6 to a 9-12-7 company for the next 3 weeks.
20/n
20/n
The next version was exponentially better and the sheer magnitude and speed of improvement made it a no-brainer for the customer and they ended up signing a 3-year deal with us.
21/n
21/n
With deal #3, we got the approval from the decision maker. But it was “stuck in legal” for a time too long for our comfort.
22/n
22/n
Through one of our contacts, we came to know that the DM is talking to another vendor. At 1 AM the same night, I booked a 6 AM 🛫 to meet him.
Get on a jet - @jasonlk
23/n
Get on a jet - @jasonlk
23/n
Met him at 3 after waiting in the lobby for ~6 hours. 72 hours in the lobby (on the same chair :P) and a couple more meetings later, I returned with the signed contract.
24/n
24/n
If I were to give 1 takeaway from my experiences, it would be this.
You don’t get anything if you stay in your comfort zone.
You have to get out and fight.
FOR EVERY. SINGLE. THING.
EVERY DAY. ALL THE TIME.
25/25
You don’t get anything if you stay in your comfort zone.
You have to get out and fight.
FOR EVERY. SINGLE. THING.
EVERY DAY. ALL THE TIME.
25/25
