This was one of the first interviews I conducted, as part of the research for my book on product market fit (PMF here onwards). Asad is cofounder & CEO, LambdaTest, building one of the most technically innovative India startups. As Satya Nadella recently put it after meeting Asad: “LambdaTest is doing for test automation what Kubernetes did for container orchestration – creating that next level of efficiency around test automation so that people can actually focus on testing versus test orchestration.”
LambdaTest was one of the most non-consensus bets we took at Blume (we led the seed round in late ‘19). When we first met them in July ’19, PLG (product-led growth motion) was not the buzzy catchphrase it was, the two founders did not have a technical background (that said both have an incredible tech understanding though), they were based in Noida, not Bangalore, and so on. Yet, they have built an incredible product, one that has garnered a lot of user love globally. Now, they are adding on an enterprise motion with the HyperExecute test execution platform.
In this convo, Asad talks about three broad themes. The first is his take on PMF, including his rather quirky definition (curiosity and reaction), how they determined that they had PMF at LambdaTest, his view on why PMF should be seen at feature level as much as at the product level (a point that Nishchay of Jar also alluded to) and why a $1m ARR is not always a sign of PMF. The second theme he covers is his growth playbook – here he covers why extremely innovative products often struggle for adoption early on, his playbook for growing from $1m to $5m & beyond, and why a high churn doesn’t worry him too much. The third and last theme was his perspective on PLG including how founders pursuing a PLG motionshould invest in content months ahead of the product launch, why PLG startups have a faster learning curve, and why they are more capital efficient.
This is a dense, insight-packed convo, so be prepared to read it slowly, and even twice. Enjoy!
(The transcript has been edited to make it more readable. There will of course be errors; please excuse me for them. The interview took place in person at Asad’s office in Noida, on 8th August 2022.)
Asad’s definition of PMF
Asad: PMF (Product Market Fit) for us is that when you give the first cut (of the product) in hands of the user, the very first PMF is: are they curious about it? The product can never be perfect, but if they have any questions and if they come back to you & ask you to solve something, it means that they did not write off the product; something worked for them that’s why they want it better even though you released one piece (early MVP = Minimum Viable Product) and the customer paid 100/1,000 bucks.
It’s good for you because when someone is writing you off, they don’t talk, they don’t want to waste their time. If they are talking to you, they’re coming back and spending some money, it means something is working, they like the whole approach of the product. This is one step; probably they want it better, that’s the first step of the PMF.
When you decide to launch something, what should be the first piece of the product? That first piece is super important because that is the zero to one part, and it determines whether your product and co may live or die. They may write it off because their needs are completely opposed to what you built. So, figuring out that small piece as to what it could be and the market matters.
Our belief was that if the product is highly innovative, and it is the first time the user is seeing it, then it will be slow on adoption. Start with a feature or product that a user or VC can relate to faster, then innovate on Day 2 or 4. It will be easy for you to showcase the product to the audience you have. They have trust in you and can lead to faster adoption of product, leading to a fast-selling product and a product which achieves PMF fast.
So, imagine I ship the HyperExecute (their enterprise product) first and PLG (Product-Led Growth) one later, the adoption for HyperExecute will be very very slow.
If you are building from India right now there are a lot of capital issues & also heavy growth pressures. So, I’ll not suggest that you launch with an innovative product. It might create confusion. You may get written off by the users just because they couldn’t understand the product. So, giving them something that they can relate to, a 2.0 version, is better, say an extended or better version of something existing, e.g., say you are launching an electric vehicle for the first time, the first of its kind; imagine how people will write you off. Imagine the first EV when it was launched and how people reacted, and now compare it to the reaction when the 10th EV was launched. The reaction was very different the first time because the first time they were curious, like, how does an EV run on a battery etc., but by the 10th product, the manufacturers had learnt a lot & gave you a better experience, a better version of the product, so adoption was faster. Every entrepreneur wants faster adoption. So, when you launch something, which the user can relate to fast, this is the 1st PMF.
If you are getting feedback, it’s a good sign; it means that the customers are interested in your product even though it is not working perfectly. This is the basic idea of PMF.
Now, people have different benchmarks around PMF; someone says 100 customers of ticket size of $15-20, someone says two large enterprises of $20,000; someone says better PMF is when you have customers from North America & Europe, not just the domestic market, for SaaS particularly; people have different benchmarks, and they can keep counting.
Growing from $1m to $5m and beyond
Asad: But I also believe that referral is not PMF, e.g., if an unknown customer shows interest, it’s definitely a true PMF. There is no bias involved. And I also feel that the customer who will use it without any prejudice, the user behaviour, adoption, all this gives you some indicators that something is working. Now you keep getting the feedback & you keep building on top of it. If something is working you keep it; if something is not working you unplug it.
When you reach $1 million ARR & you figure out that this $1m ARR product, can we take it to $10m or we need something more. Because that was an MVP that took you to $1 million ARR. So we decided that we need another product with a higher ACV so that we can land & expand. You are actually bringing some customers for a small ticket product but you don’t have any other product to upsell & grow faster. So, the second PMF comes from an adjacent product. Ask your customers what is needed next and many of them keep giving you unsolicited advice. That said, customers create confusion as well(!) People come with very niche demands, or for a customised experience etc. You should always listen to the larger market. Ask: if I build something, say one piece, can I sell 100k pieces of it, is the market out there or not? (Be careful about building too much for one customer and getting trapped in their unique demands.). So, that is the thin line of the product management which the product manager has to take care of. The demand will be huge & you’ll have to pick the fast moving, fast selling features.
You will keep giving them excitement, you cannot build everything. Building everything is a services mindset, it will take you to a slow death; you will end up bloating the product that no one understands that.
To reach $5m ARR, you introduce another product which actually your customer demanded, and you felt this is a larger market and then you focus on the (org structure and all) but after $5 million maybe these two product lines are enough to take you to $100 million. You keep adding more features and keep pulling out features which are not working and keep fine tuning it.
After $5 million ARR, I believe the best thing becomes the experience, scalability & reliability. If focus is too much on the features, then your scale and reliability and experience goes to the back-seat. You cannot keep building the features forever at the same speed. We believed that stability, reliability is very important. Your down time, quality, test exploration all these are super important. What do our customers come for? Do they come for better UI? Do they come for better indications? Do they come for high quality tests? What do they come for? So, all this comes into product market fit.
You keep embedding all these into the product. Then it comes to figuring out aha moments, how will you scale aha moments, figuring out highly used features, how will you innovate those highly used features, how will you make them version 2.0/3.0/4.0 (how you will improve these features), like Stories on Instagram, right, how will you create more and better and better experience every month for the Stories feature. Because it is working nicely, so the Stories we’re seeing today and Stories we are seeing after five years will be very different. That is a dedicated job, to keep doing it. High adoption features should be innovated massively going forward.
That’s how the fundamentals of product management work. Focus on customers, learn from them. Every aha moment is not a market. Every innovation is not a market. Innovation which mass market loves is the right segment or market to focus on. I mean sometimes you go crazy saying it’s a very hard problem I solved. That is not the market, that is not the actual problem of the customer. They will say they loved it, they will say you did a great job but is that a business? That kind of thin line the product managers have to manage.
Sensing PMF at LambdaTest
Sajith: This is interesting. Now in a talk to the Blume founders recently you mentioned how LambdaTest started creating content before the product was launched, how Jay Singh (his cofounder) did about hundred cold calls to get customers to try out the product. When did you feel there was PMF during this journey?
Asad: When our cross-browser product was working, we got 100 customers. But that is not the product that we wanted as our forever product; that was not our main product or main priority, that was our foot in the door. The moment we launched the adjacent product and the ACV went up from $180 to $1,000-1,200 for one licence yearly & when we started getting our first 5 customers over there…two theses worked – the 1st product got 100 customers; that was low hanging fruit. It means my GTM / foot in the door is ready & then we launched the adjacent product after learning from the customer. I got 5 customers where we’re getting $5000 to $20,000-25,000 yearly. Combined is the platform. So, we believed that after getting 5th or 6th customer in the automation task, we found that PMF came in. The customers were screaming too much, opening a lot of tickets, it means that customers really wanted it. That’s how we achieved our PMF.
A Bay Area VC told us that 65,000 signups and $250,000 revenue & 100plus customers are the benchmark metrics for PMF. But as a founder I don’t believe in this alone. I have a grocery store & only wheat sale happens. There is everything in the grocery store but people only come for wheat. If so the grocery store is not a PMF, wheat is the PMF. So, the founder has to understand how many features of the product actually truly got the PMF.
Metrics for judging PMF at LambdaTest
Sajith: So, in the early days what were the metrics that you tracked razor sharp, till you achieved PMF? One interesting thing you told me is that customers screaming too much is a good sign, like tickets being opened.
Asad: At the early stage if you don’t have too much bandwidth & the capital to track everything, then go slow, but I don’t recommend this. I really recommend to track everything. It helps to get to know the issues faster & fix it. We did this mistake. In the early days we were majorly focusing on adoption – I released one feature, and looked at how many users actually logged into the system and how many of them use that feature? That adoption plus some metric around it, how many hours of test is happening, what user test is happening, how that adoption is coming on the bug login, how our products being used to log the bugs, then how many users are logging the bugs, how much consumption is coming on the VM side, all these metrics we actually track, user metrics. We really believe that the faster you do, and understand these metrics, the faster you get to know issues & the faster you fix it.
Sajith: These are user / usage metrics. What about revenue metrics, did you track it?
Revenue is not a signal for PMF
Asad: I will honestly tell you, I can tell you that with even $1 million ARR you cannot say that you have PMF. Revenue is not the only sign for PMF.
Sajith: Imagine a company where a founder or VC comes to you and says with $1.5 million ARR or something like that, it has PMF. What would be a sign that it doesn’t have PMF according to you? Like what would tell you that even if it’s $1.5 million revenue it doesn’t have PMF?
Asad: I’ll go deeper in the user metrics & figure out how many cohorts there are, user cohorts actually. Who are these people and why are they using it? If I have a sweet shop, I will ask who are the consumers – the diabetic patients or non-diabetic, I will figure it out. You are saying that everybody is buying it. So, we have to figure out the cohorts by ourselves; that there is some behaviour that it sells mostly in the Middle East because they need it or it sells globally, it sells in the oil companies or it sells in only the finance vertical. Whether it’s a global product, vertical agnostic, geo agnostic, lifestage agnostic etc. But you have to figure out actually what is working for these guys even though they’re speaking the truth. The VC has to go deep down, and figure out actually what is working for them. Is it 1 customer paying $1m or 100 customers paying $1m or 1000 customers paying $1million? I will bet on 1000 customers paying $1 million over the company who is taking $1 million from 1 customer. I can even give 5 times more valuation to the company with 1,000 customers paying $1m than the company that has $1million from 1 customer.
I’ll be very careful if a couple of customers are paying $1 million, because this clearly indicates to me that these guys are working for a couple of customers and fulfilling all the demand. There is clearly some service component involved here, so I’ll be careful. So, you have to deep dive. Just ask simple questions, how many users are using the product, how many users are using which features or the revenue which came from where, what is the nature of the companies who are using it etc. Just get the data & analyse it. This gives you a clear-cut opportunity to understand the $1 million business. But it gets more difficult when you are betting on $100-200 thousand revenue companies. It is too early. There as an Angel I have a different behaviour; VC will definitely have a different behaviour. As an Angel I only see 2 things:
- How flexible the founders are to pivot & do they really understand who they look upon to build the business? Where they could end up. If they have 1 feature, is there any agility… will these founders will figure out quickly to convert that one feature to a platform?
- How much open thinking they have? It should not be cult following (of some practice). So, the founder behaviour becomes very important, how far this founder can think and can pivot fast & move faster. How faster things resonate to him. If you’re telling him something then you see the reaction. Second is the market, if the market is not big enough then founder cannot solve it. So, being in the right market & being the right founder are the 2 metrics, the only two things that I actually go for when I bet on the founders in the early stage.
Sajith: Right founders & right markets.
Asad: Right founders and right market can do anything.
Churn
Sajith:At the early stage how important is churn or how not important was it?
Asad: Churn definitely is a very important metric. But don’t shoot the company if the churn is happening. You have to understand the psychology of the customer and why they are churning. They churned since the customers couldn’t wait, say the company doesn’t have capital & that’s why they cannot deliver the features fast and that’s why they are churning or they were too small to pay that price e.g., sometimes $99 is a pain for the agencies to pay. So, there are 3-4 reasons; if 70% of my customers churned for a certain duration, & I couldn’t figure out why they churned then that is a dangerous situation.
If they know why they churned & if they have the right approach to solving the problem, I guess it’s okay. Because that inbound funnel is working, more customers will come. But if they don’t know why they are churning it’s a dangerous sign. If the founders are able to answer why they are churning, they are the right founders to bet on they will figure out how to fix it. But sometimes you see that the founders don’t have any idea & they are trying to hide something, that they don’t have any idea why they are churning. This is not a good sign.
The PLG (Product-Led Growth) motion
Sajith: So, now I just want to go a little bit into GTM, you used PLG and as a founder how do you decide on the GTM motion? Was your picking PLG a logical thing for you? Because browser stack was using PLG you had to? How did you think about this?
Asad: There are 2 reasons: 1) PLG doesn’t require too much capital 2) Faster acceleration comes through PLG. Enterprise requires foot on the ground plus a mature product to go the market with. I found that to build a mature product and put sufficient feet on the ground, you require lot of capital. So it’s situational. Every product cannot be PLG & it’s not necessary that every enterprise (software motion) can be PLG, or PLG can be an enterprise. You just don’t have to do it forcefully. You just have to make sure that the product allows you, your market allows you.
For us, it was a low hanging fruit because our competitor was also doing it. It validated very clearly that it is working for them and it will work for us. So, we started as a PLG. We had limited capital; it was better for us to keep non-paid marketing channels and build something for sustainability. I mean your PLG can be sustainable for the long term. But enterprise motion is completely dependent on the people. So, the more horsepower you have the better you can track. Enterprise is a proven channel. You have more people and senior people on board they will go & sell but they will create a lot of noise. Compliance is not there but we need it, Security is not there but we need it, all these demands will come. So, maturity will take time. But for PLG; we also need it for 1 more reason – lots of customers will teach you how to build a product. Few customers will not give you the right PMF. Imagine 1,000 customers v/s 5 customers for $1 million. It’s better to build a thousand customers product company because you get the PMF fast.
Sajith: Learning is faster with PLG motion. Got it. Say for PLG motion, let’s say a founder is saying, I have to crack this, let’s say some project management software or something he wants to launch, say $20 ACV per month. What is the first thing they need to be doing right now, company will get launched in 6-7 months. That day you spoke about content etc. So, what are the channels and how should the founders think about setting up a good PLG motion what are the basics?
Asad: Basics are around your SEO strategy, content strategy, going deeper into community and not hard selling anything. Just evangelising, talking to them in reddits, subreddits; you go & share your thought process, what you think about the product, issues etc. I’m working on the testing product, I will go in testing reddits & share my opinion. I will build my credibility as a thought leader. You do this in Quora, Medium etc. You start doing this but you don’t (hard) sell, because your product is not ready so what will you sell?
Identifying those places where you can ship the article and you get some kind of fraction immediately; so, all this preparation takes time. It takes almost 6 months to figure out what are the right key words, right structure, technical SEO, non-SEO, everything to get into shape & then you launch the product (if need be with an access invite etc.) in the smallest cohort. Then when you start getting some product market fit, once you get some kind of feeling that PMF is coming, then you open for all.
Imagine if you launched it and now you start the wheel (on content), think how much time it’s going to take. It’s not a 1-day job. You cannot push more than 10 articles in a month, maximum 50 threads a month you cannot do more than 50 reviews in a month. You have limited capital, limited resources, limited everything. So, if you’re not doing this prelaunch content work for 12 months then how will you do it after launch?
Many times, founders, they don’t have an idea. Even we figured it out slowly, we have taken our own time to learn that how PLG will work? So, it took 3-4 months to conceptualise, how the website structure will be, how the SEO keywording will come, what kind of review sites we will target immediately etc. So, to shorten the time of planning and building strategy we started 6 months before.
Sajith: You advise that, start with content 6 months before the product is launched or you even code.
Asad: Also additionally, one important learning for me was that PLG motion for a market with a smaller TAM, say under $30b or so, never works forever. The motion gets stagnant after sometime. So PLG startups with a smaller TAM market need a hybrid approach of PLG bottom up and brutal top down motion. That said when your market is large say $100b+ like Figma, then I think PLG works like a charm through out and can be never ending. You can focus on PLG alone and accelerating it. But remember from $1m to $10M, hustle and hard work works nicely, after $10M we believe culture, sales machine and math will work.
Sajith: Got it, Asad, we should take a break now.