Founder Interview
How ParallelDots Grew to $500K ARR with 15 Enterprise Customers in Under a Year (Interview with Angam Parashar)
- Interview Date
- July 30, 2018
- Interviewee
- Angam ParasharFounder
Company Metrics at Interview Time
ARR (2018)
$500K
Enterprise Customers (2018)
15
Total Funding Raised
$2,000,000
Team Size (2018)
30
Historical Snapshot
These numbers were reported by Angam Parashar during the interview recorded in July 2018 and are a historical snapshot, not current figures. See ParallelDots’s current numbers.
Key Takeaways
- 01ParallelDots reached $500K ARR in 2018 with zero revenue as recently as July 2017
- 02The company signed 15 enterprise customers in its first ten months of enterprise sales
- 03New enterprise customer additions ran at two to three per month over the prior three months
- 04Average contract value for enterprise customers was $25,000 to $30,000 per year, reaching up to $50,000
- 05Customer acquisition cost was $1,500 to $2,000 per enterprise customer, with payment charged in advance
- 06More than 12,000 developers had signed up on the platform cumulatively by mid-2018
- 07The sales team closing enterprise deals consisted of just two people
- 08ParallelDots raised $2,000,000 in an equity round in 2016 from US-based investors
- 09Enterprise churn was zero across all 15 customers over the ten-month period
- 10The company was founded in 2014 and spent its early years focused on AI research and technology
Company Metrics at Time of Interview
| Metric | Value | Source |
|---|---|---|
| ARR (2018) | $500K | Founder interview, July 2018 |
| Revenue (2017) | $0 | Founder interview, July 2018 |
| Enterprise Customers (2018) | 15 | Founder interview, July 2018 |
| Paying API Developers (2018) | Around 100 per month | Founder interview, July 2018 |
| Developers Signed Up (cumulative) (2018) | More than 12,000 | Founder interview, July 2018 |
| New Enterprise Customers per Month (2018) | 2 to 3 per month | Founder interview, July 2018 |
| Average Contract Value (Enterprise) (2018) | $25,000 to $30,000 per year, up to $50,000 | Founder interview, July 2018 |
| Customer Acquisition Cost (2018) | $1,500 to $2,000 per customer | Founder interview, July 2018 |
| Enterprise Churn (2018) | 0% | Founder interview, July 2018 |
| Sales Team Size (2018) | 2 | Founder interview, July 2018 |
| Total Team Size (2018) | 30 | Founder interview, July 2018 |
| Total Funding Raised | $2,000,000 | Founder interview, July 2018 |
| Equity Round (2016) | $2,000,000 | Founder interview, July 2018 |
| Year Founded | 2014 | Founder interview, July 2018 |
Growth Breakdown
Revenue
ParallelDots reported $500K ARR in mid-2018, up from zero revenue in July 2017. All growth came within the prior ten months as the company shifted focus from AI research to enterprise sales.
Customers
The company had signed 15 enterprise customers by the time of the interview, adding two to three new enterprise customers per month over the prior three months. More than 12,000 developers had signed up on the platform cumulatively, though the vast majority used the free tier.
Team
The team grew from five people in the early years to 30 at the time of the interview, with most staff based in India. The enterprise sales function was handled by just two salespeople.
Funding
ParallelDots raised $2,000,000 in an equity round in 2016 from US-based investors. At the time of the interview, the company was in advanced conversations with investors about a potential new raise and had received a term sheet, though negotiations were ongoing.
Growth Strategy
Enterprise API Solutioning
Enterprise customers have different datasets, so ParallelDots often added "a bit of customization" to its NLP and computer vision APIs for them, which Angam called "solutioning the API for these enterprises so that they see enhanced performance." He said this also helped the company make higher revenue.
On-Premise Deployment for Enterprise
Enterprise customers were concerned about data security and preferred on-premise deployments. Asked whether all of its deployments were on-premise, Angam answered, "Pretty much, yes."
Annual Upfront Contracts
The company charged enterprise customers on an annual licensing basis and collected payment in advance. Contracts usually ran for three years, against an acquisition cost of $1,500 to $2,000 per customer and a usual ticket of $25,000 to $30,000 a year, up to $50,000 at times.
Freemium Developer Tier
ParallelDots offered 10,000 free API calls per month to developers, building a base of more than 12,000 signups. Most developers stayed on the free tier, and Angam said the sign-ups made ParallelDots "one of the most popular NLP/NLU APIs out there."
Inbound Investor Interest
Angam said the company was not actively raising, but its traction over the prior year had brought "a lot of inbound requests" from investors, and it was talking with a few of them.
Best Quotes
“So ParallelDots is an artificial intelligence platform for developers and enterprises globally.”
“More than 12,000 developers have signed up on our platform today. So we are one of the most popular NLP/NLU APIs out there today.”
“So in a month, we have around 100 API developers who pay us. That's what the numbers are. But the real revenue comes from the enterprise as as you would guess.”
“We have signed up 15 enterprise as of now.”
“We charge them on an annual licensing basis. Typically, we charge them... The usual customer ticket size would be around $25,000 to $30,000 going up to $50,000 at times.”
“Our current ARR is around $500,000. That's where we are.”
“So July 2017, haven't really... Until then, we haven't really signed any enterprise. So all the growth that we have witnessed so far has been in the last ten months.”
“For enterprises, the churn has been zero. I have a small data point for just 15 customers that you're talking about. The churn has been zero, and it's been over the previous ten months only.”
“Our acquisition cost per customer is around anything between $1,500 to $2,000 usually.”
“We are looking to raise around $5,000,000”
What Happened Next
At the time of this interview in July 2018, ParallelDots had been signing enterprise customers for about ten months. It was not actively raising, but inbound interest had it talking with investors about a raise of around $5,000,000; it had a term sheet that was still being negotiated and no closed round. Angam said the company was "definitely in losses" while closing two to three enterprises a month with a sales team of two. Visit the ParallelDots company profile on GetLatka for current figures and any updates since this recording.
View ParallelDots’s current profile and metricsFull Transcript
Chapters
- 0:01Introduction and Angam Parashar's Background
- 0:18What ParallelDots Does and How It Makes Money
- 1:28Developer vs Enterprise Customer Segments
- 2:31Enterprise Customer Count and Contract Structure
- 3:30Revenue and Growth Trajectory
- 3:56Company History and Early Burn
- 5:03Funding History and Equity Round
- 6:53Current Fundraising Plans and Valuation
- 8:05Enterprise Churn and Customer Economics
- 8:59Customer Acquisition Cost and Payback Period
- 9:38Famous Five Rapid Fire Questions
Introduction and Angam Parashar's Background
Nathan Latka
00:01Hello, everyone. My guest today is Angam Parashar. He's the founder of ParallelDots. He's also a graduate of Indian Institute of Technology. After his graduation and before starting the company, he worked as a consultant at Opera Solutions, a niche big data consulting firm. Angam, are you ready to take us to the top?
Angam Parashar
00:17Absolutely. Go ahead.
What ParallelDots Does and How It Makes Money
Nathan Latka
00:18Okay. What is ParallelDots and how do you make money?
Angam Parashar
00:22Sure. So ParallelDots is an artificial intelligence platform for developers and enterprises globally.
00:31When I talk about artificial intelligence platform, what we essentially do is that we provide text analysis, NLP/NLU APIs, as well as computer vision APIs. It's a simple plug and play API which people can just sign up on our website and start using it.
Nathan Latka
00:48And is it a SaaS business?
Angam Parashar
00:51Yes, it is a SaaS business. There are two primary users that we are targeting. One is the developers, as I mentioned.
00:59These could be students, indie developers, app developers, whatever work they are doing. There are many use cases for them. The other one, the target users are the enterprises for us. So when you talk about developers, it's essentially a pay as you go model for them. We charge them on a number of API calls that they make. More than 12,000 developers have signed up on our platform today. So we are one of the most popular NLP/NLU APIs
01:26out there today.
Developer vs Enterprise Customer Segments
Nathan Latka
01:28How many paid customers are you working with today?
Angam Parashar
01:31So number of paid customers in developers. Developers usually use the freemium model where we provide per month around 10,000 API calls for free. So most of the developers around, I would say 90% of the developers, 95% of the developers would follow around that. So we don't usually make a lot of money from the developers. So I would say 5% of our developers would pay us.
Nathan Latka
01:52Okay. So what's the number? How many paying customers are you working with right now?
Angam Parashar
01:55So in a month, we have around 100 API developers who pay us. That's what the numbers are. But the real revenue comes from the enterprise as as you would guess. Right? In terms of... Let me talk a little bit about the enterprise as well. So in enterprise, the kind of challenges that they face is different. They are very much concerned about the security. Do not
02:15feel
Nathan Latka
02:15Name like one or two.
02:16Just give us a picture first and then work backwards. So name an enterprise working with you.
Angam Parashar
02:22I cannot disclose the name of the enterprise we are working... Enterprises...
Nathan Latka
02:26There's no enterprise that you have permission with that you can disclose. They're all secret.
Enterprise Customer Count and Contract Structure
Angam Parashar
02:31Yeah. I would not like to disclose the name of the enterprise that we have signed up. Having said that, we are working currently with... We have signed up 15 enterprise as of now.
Nathan Latka
02:38Oh, There's a very... In case of our... It's it's a very few amount so far.
Angam Parashar
02:43Oh, yes. Yes. It's a few amount of enterprises so far. Yeah.
Nathan Latka
02:45Okay. And what's their revenue model? They they pay per month?
Angam Parashar
02:49They pay... We charge them on an annual licensing basis. Typically, we charge them... The usual customer ticket size would be around $25,000 to $30,000 going up to $50,000 at times. That's what a typical...
Nathan Latka
03:02Monthly or per year?
Angam Parashar
03:03Per year. That's an annual contract size.
Nathan Latka
03:06That's great.
Angam Parashar
03:06And our contract usually lasts up to the duration of the contract is usually for three years. So we have signed up around 15 enterprises, as I mentioned. Our current ARR is around $500,000 That's where we are.
Nathan Latka
03:22That's great. So $500,000 divided by 12, you're doing about $40 to $42 grand per month currently?
Angam Parashar
03:29Yes. Yes, absolutely. You're right.
Revenue and Growth Trajectory
Nathan Latka
03:30That's great. And what's the growth rate look like? So if you're doing $42,000 today, how much were you doing a year ago in July 2017?
Angam Parashar
03:36We are signing anything between two to three custom enterprises per month for the last three months. So July 2017, haven't really... Until then, we haven't really signed any enterprise. So all the growth that we have witnessed so far has been in the last ten months.
Nathan Latka
03:53So you had no revenue a year ago?
Angam Parashar
03:55Yes, we had no revenue.
Company History and Early Burn
Nathan Latka
03:56And when did you launch the company? What year?
Angam Parashar
03:58So we started this company in 2014.
Nathan Latka
04:022014? Yeah.
Angam Parashar
04:04We were experimenting with two different products, And after a long time, we eventually reached here where we are showcasing. So all the growth that we have witnessed has been the past ten months.
Nathan Latka
04:13So Angam, you're rich. Somebody was paying the bills for four years when you had no revenue. Who was it?
Angam Parashar
04:18Our burn was low, so that helped us.
Nathan Latka
04:20How low?
Angam Parashar
04:23So most of the team is based out of India. So back then our burn was very low. We were a team of five people. So the burn was pretty low, as you can guess. Now we have a substantial burn, but then we also make decent revenue, so we can cover that. We're also funded.
Nathan Latka
04:37What is substantial burn?
Angam Parashar
04:39So I cannot disclose the burn again, but we're definitely in losses. It's more than what we make today, but we're also funded. With the kind of growth trajectory that we are witnessing closing around two to three enterprises in a month with the current team focus of the current sales team, is just two people. So
Nathan Latka
04:39And what's
04:56the total team size?
Angam Parashar
04:58We are 30 people at the moment. Three zero.
Nathan Latka
05:01Three zero?
Angam Parashar
05:02Yeah. Yeah.
Funding History and Equity Round
Nathan Latka
05:03And how much have you raised?
Angam Parashar
05:04We have raised $2,000,000 in venture capital.
Nathan Latka
05:062,000,000. And was that from folks in India that you raised from VCs there or folks in The States or Europe? Where'd you raise from?
Angam Parashar
05:13It's it's folks from United States.
Nathan Latka
05:15Okay. And was that a convertible note or an equity round?
Angam Parashar
05:18Equity round.
Nathan Latka
05:19Oh, equity round. Very good. Why you decide to do equity right at the bat instead of doing a seed on a convertible note or something?
Angam Parashar
05:25So at that point of time, we were looking for an equity round because we wanted to have more... When we raised the money, we did not have a lot of clarity around revenue numbers that we had. And our team's DNA earlier was very much based on the AI research rather than on the business side of it. So our focus was always to publish the right papers, build the right technology rather than going out to the market.
05:46So at that point of time, the money that we raised was more on the technology side rather than the business side. So we decided, hey, let us do equity. Let's do an equity round.
Nathan Latka
05:54I see. I see. Okay. And what year was that that you raised that capital?
Angam Parashar
05:582016.
Nathan Latka
05:59'16. Okay. And do you mind me asking what valuation did you raise that
Angam Parashar
06:03I'm sorry, I cannot disclose that.
Nathan Latka
06:05Okay. What multiple of revenues at that point was it? We're talking like 10x or 4x or?
Angam Parashar
06:11At that point of time, as I mentioned, our revenue was nil. So the valuation was... If you talk about the multiples of revenue, it doesn't make any sense actually
Nathan Latka
06:11So for
06:21how did you go in and say, I want this valuation when you had no revenue? What did you kind of base it on?
Angam Parashar
06:28It was simple actually. We had a lot of technology to show and we had some previous failed production to show. So this is what we had. We wanted to build the business as we go ahead. So those are the three things I had. I had a very strong AI team with me. I had a very strong technology with me, and I had some strong experience of how to build the products. So all these three things really
06:52helped us to raise this round of money.
Current Fundraising Plans and Valuation
Nathan Latka
06:53And are you raising additional capital right now?
Angam Parashar
06:56Oh, yes. We are not actively raising it, but there's a lot of inbound requests now given the traction that we have for the past one year. Now let's see what happens. We are talking with few investors. Let's see how it goes.
Nathan Latka
07:08How much are you looking to raise?
Angam Parashar
07:11We are looking to raise around $5,000,000
Nathan Latka
07:13Okay. And are you... I mean, generally in a in a... I mean, I guess you could call this your series B. People are typically selling five or maybe 10% of the company. Right? So I mean, is it fair to say you're trying to get like a 45,000,000 pre money valuation, something like that?
Angam Parashar
07:25I don't think we'll get that kind of valuation with this much revenue. So... Which is why I I I mentioned that. Let's see how it goes. The kind of valuation that I'm looking right now is more of a of a 20,000,000 pre money rather than 45,000,000 pre money.
Nathan Latka
07:3920 pre means what you'd be if on 5,000,000, you'd sell what, 20% of the company?
Angam Parashar
07:43Yeah, that would mean selling 20%
Nathan Latka
07:43Yeah,
Angam Parashar
07:43of the
Nathan Latka
07:45very good. And do you, I mean, you have a lead investor? Do you think you'll get it?
Angam Parashar
07:49We are in conversation with quite a few of them. Docs are in advanced stage. I would like to, of course I'm hopeful.
Nathan Latka
07:57Do have a term sheet yet?
Angam Parashar
07:59Yes, we have a term sheet yet. We are still negotiating. It might fall apart. It might just go ahead. Let's see what happens.
Enterprise Churn and Customer Economics
Nathan Latka
08:05Yeah, very good. Okay, walk me through some of the kind of customer economics. So you have a lot of developers on the platform, but your real bread and butter are these 15 enterprises using you. What's churn today?
Angam Parashar
08:15For enterprises, the churn has been zero. I have a small data point for just 15 customers that you're talking about. The churn has been zero, and it's been over the previous ten months only. So in developers, the churn rate is pretty high as you would guess. But developers for them, the use cases are very project driven. They have a certain use case. They solve it probably for the students. They're doing a master's thesis, their bachelor's thesis,
08:37they use it, they go away. So that is what happens with developers. But for enterprises, they have a huge scale at times. So the pay as you go model doesn't really work for them. Apart from that, they're very careful about the data that they want to send you. So they usually prefer on premise deployment. There's another very interesting problem which-
Nathan Latka
08:37Hold on,
08:56are all of your deployments on prem?
Angam Parashar
08:58Pretty much, yes.
Customer Acquisition Cost and Payback Period
Nathan Latka
08:59Okay. So with that in mind, what are you paying to acquire a new customer?
Angam Parashar
09:04Our acquisition cost per customer is around anything between $1,500 to $2,000 usually.
Nathan Latka
09:11Okay. Well, that's pretty good. If they're paying you $30,000 ACV, you're getting that money back instantly.
Angam Parashar
09:16I would also like to answer on the third point actually, which is enterprise have different datasets. So often there's a bit of customization which is involved, which is what I call solutioning the API for these enterprises so that they see enhanced performance. So this also helps us make higher revenue.
Nathan Latka
09:32Yep. But just to be clear, again, payback period is almost instant, correct?
Angam Parashar
09:35Yes. We charge money in advance. Yes.
Famous Five Rapid Fire Questions
Nathan Latka
09:38Yep. Yep. Very good. Okay. Very good. Let's wrap up here with the famous five. Number one, what's your favorite business book?
Angam Parashar
09:44My favorite business book?
Nathan Latka
09:44Yep.
Angam Parashar
09:44My favorite business book would be the recent Peter Thiel's book that I read.
Nathan Latka
09:52Zero to One.
Angam Parashar
09:54Yes. I I loved it.
Nathan Latka
09:56Good. Number two, is there a CEO you're following or studying right now?
Angam Parashar
09:59Jeff Bezos.
Nathan Latka
10:00Number three, what's your favorite online tool for building your business?
Angam Parashar
10:04Online tool? Oh, Google Analytics.
Nathan Latka
10:06Number four, how many hours of sleep do you get every night?
Angam Parashar
10:10Five.
Nathan Latka
10:11And what's your situation? Married, single, kids?
Angam Parashar
10:14Single.
Nathan Latka
10:15Single, no kids?
Angam Parashar
10:16Single, no kids.
Nathan Latka
10:17Alright. And how old are you, Angam?
Angam Parashar
10:19I am 28 years old.
Nathan Latka
10:2028. Last question. What do you wish your 20 year old self knew?
Angam Parashar
10:26That's a good question. I think just go ahead, take risk in life. Don't worry about things.
Nathan Latka
10:31Guys, take more risks. Don't worry. Again, launched his company back many years ago in 2014. Now has a team of 30 people, 2,000,000 raised. Again, really creating, I think, you call simplified essentially machine learning for anyone. They're making a lot of their money. Well, they have a lot of developers using them, but the real money is coming from enterprise and enterprise services, them solutioning and architecting these services in an on prem format for these enterprise customers,
10:55currently serving 15 of those folks doing about $40,000 per month in revenue. Each of those people paying again about $30,000 ACV's. They had no revenue a year ago, so good first year kind of traction here. Too early to talk about unit economics, they're spending about $1,500 to acquire customers with a team of 30 based in India. Currently raising or, you know, who knows what'll happen, but potentially raising $5,000,000 on a 20,000,000 pre money. We'll see what
11:16happens. Angam, thank you for taking us to the top.
Angam Parashar
11:19Thank you, Nathan.