13 lessons for driving growth at Startups

Recently I had the opportunity to speak to around 20 startups as part of Conquest 2020 organized by BITS Pilani. In our interaction, I had shared with them 13 of my top learning I got from driving Growth in Indian Startups like Shuttl, OYO and Snapdeal. We had an engaging discussions on all these learning and how they could use these lessons in their startup. A few of them requested to make that presentation into a blog so that it would be useful for a other startup founders and growth leaders as well. So here you go. Would love to hear your thoughts on the same. Happy Learning!

Note — These learning are not in any particular order.

1.Starts with a Crazy Growth Engineer in the team! 🤪

Message for every startup founder. Get a crazy Growth engineer early on. You will thank me later :)

There will always be a difference in way of working between Growth and Engineering teams. Growth would focus more on quick experimentation and fast pivoting if things aren’t working out. Compared to this Engineering team would prefer to be more planned so that they schedule their sprints properly. That’s where a Growth Engineer comes in. A growth engineer can unblock the Growth team at multiple levels and helps them to move much faster with their experimentation to figure out what’s working and what’s not. Below image details out “Who” qualifies to be a Growth Engineer and what to expect from the person. A Growth engineer believes in making the impossible, Possible!

At Shuttl, we had a Crazier than crazy Growth Engineer (Shout out to Mudit Agarwal!). He alone was able to unlock 2–3 new channels for the Growth team to acquire customers at a much lower CAC. Whenever anyone of us thought about any crazy idea on Growth, Mudit was our go to person!

2. No PMF = Misaligned incentive 💰

PMF or Product Market fit is an often discussed topic in the startup world and there is a ton of material available on how it is defined and what’s the best way to measure the same. However, for every startup, this article by Rahul Vohra (Founder, Superhuman) should be the starting point to understand PMF.

I wouldn't delve into the details around how to measure PMF for your startup but if I have to summarize what PMF looks like, I would just ask early users how they would feel if they don’t get to use the product and the answers would give us the clue.

In a lot of startups, if the PMF is not strong enough, there would be some other misaligned incentive for users to stick with the solution provided by the startup. The moment that incentive goes away, customers also move away along with that incentive.

However, what’s important to understand is the right time for a startup to go for scale. The above 2X2 grid explains when should a startup plan to scale. One thing comes out very clearly. Once PMF is proved, the only option left for the company is to scale the idea or else there would be some other startup which would come with a slightly different solution and walk away with the market. This would lead to a slow death for the startup instead of fast death if the startup decides to scale without a PMF. Sustainable growth for the startup comes when there is strong PMF and there are proportionate investments behind scaling up the idea! Scaling also ensures that the growth flywheel runs well. Will talk about the same sometime later.

3. Respect Experiment design: Avoid Hacks 🧭

As a growth marketer, whenever we have tried experimenting with manual hacks, those experiments haven't scaled much and in the process, probably we ended up killing the bright idea as well. As Growth marketers we are always excited about doing experiments to prove our insights / ideas. However if we do not follow the experiment design like setting up a control group or follow the principle for measuring the experiments properly we won’t arrive at the right conclusion. Also, a lot of time we change the experiment design to suit our narrative. That’s a trap. Stay away from it.

Sharing an example to clear out my point. At Shuttl, we had always run A/C buses in most of our cities as we believed that was a key element in our offering for customers (With Pune being an exception where we run only Non A/C buses). Sometime in early 2019, we thought of experimenting with Non AC services as we believed that non AC vehicles had higher availability in the market and would help us scale faster. However, our back end systems were designed for having only a single offering to customers (either A/C or non A/C) as we didn't have capability of showing differential pricing for different types of vehicles on a single route. Since we wanted to experiment customer adoption of non A/C services quickly and without spending engineering bandwidth on building the capability to have multiple offerings on the same route, we decided to go ahead with a manual way to set up the experiment. Instead of having lower pricing for non A/C vehicles, we charged the same amount for all the vehicles (A/C or non A/C) with a cash back for people who opted for Non A/C vehicles. Since we have a subscription service where people upfront pay for almost a full month, this “pay for AC first and get cashback for using Non A/C later” was a bit complicated for the customers. Needless to say, we couldn’t really get any indication around whether the experiment worked for us or not. To the extent that some of us practically ended up writing off non A/C offering. This gave me an important learning about experimentation. Manual hacks without proper experiment framework don’t help much in evaluating an idea.

Finally, it’s extremely important to have a well designed A/B setup on the product itself. Most of the time all of us will struggle to get this prioritized because you can’t really measure the impact immediately. But if you want to build your organization with a proper experiment framework, we would have to get it prioritized.

4. DOA : Don’t Carry a dead body!💀

DOA stands for dead on arrival. Projects fail. We all need to accept the same. As growth people, we are always optimistic about the future and the potential of any project we are driving. Even if the early signs are not very encouraging, we keep adding more time, energy and effort on the project. This is a Sunk cost fallacy. We need to be dispassionate about these projects and “Cut the cord” sooner.

Let me give you a couple of examples. Flipkart Ping came in 2015 and they spent tons of money on engineering and then despite low traction continued spending on Growth. Early signs were not so positive for the project but the company continued investing behind the project. With their might, Flipkart could afford this. But most of us can’t!

Google Plus came in 2011 and was kept alive till 2018. How many of us have actually used this service? Since early days they struggled to get momentum but continued investing behind it. Google was slow in accepting their failure on the social front. 7 years in a long time to continue to spend on a project which was “Dead on Arrival”

Remember, carrying dead bodies is an unnecessary strain on your finances.

At Shuttl, I have personally been guilty of doing it with more than 1 project. We realized in the early days itself that the project didn’t make sense at all but we continued investing in it. However, as a Growth Marketer, I realized my mistakes soon and tried rectifying them. We were fortunate that this misjudgment didn't cost us much and we could course correct faster but at times any delay in identifying DOAs might be fatal for the organization.

5. Depth >>> Width! 👓

Temptation to grow really fast is real and all startups go through the same at varied levels. This leads them to shift their focus on expanding into new geographies, new functionalities, new business sooner without even understanding the true potential of existing geography or business. This not only leads to distraction but also a much shallow understanding of business. Premature expansion into new geographies or businesses lead to lesser focus on solving the pain points of existing customers and hence will always be a big strain on the overall resources available.

At Shuttl, since very early days, the team was focused on solving it really well for only one geography and one business line. For the first 3 years, the entire focus was on building consumer business for Gurgaon cluster (Delhi to Gurgaon) only. Only when we felt confident that we had solved extremely well for Gurgaon, we went out and immediately expanded into 5 new cities and grew rapidly. Focus on depth allowed the team to understand the problem statement much better than other competitors (globally) and helped us in not only quickly expanding to similar geographies but also to adjacent business (Enterprise / Intercity etc).

Another good example of depth before width is “MilkBasket”. In the initial days their model was to focus on verticals in Gurgaon and the way they approached this was really unique at that point in time. They would go into one society (or large complex) at a time and try to drive penetration of their product there. The entire marketing / growth focus would be on driving penetration in that one chosen society for the time being. Only when the society would reach a certain penetration level, they would go into any other society. And this worked really well for them. Their model is still very unique and robust and they are one of the most loved brands in the societies in which they are present.

6. Once PMF + Depth done — Focus on Non Linear 🚀

Lately I have been reading about Bamboo trees. There is something very unique about Bamboo trees which can be well explained through this quote. “The first year it sleeps, the second year it creeps, the third year it leaps.”

There is a lot a Startup can learn from Bamboo. Sleep phase indicates the startup’s fight towards “Product market Fit”. Once PMF is sorted, the focus needs to be on “Depth”, similar to the Creep stage mentioned above and then finally it’s time to leap i.e. “Grow Non Linearly”.

Once your product has achieved PMF and has spent time to get into the depth to understand the nitty gritty of the business, getting Linear growth would not be too difficult. While there would be continued focus on growing the business linearly (the Business as usual stuff), there has to be some bandwidth carved out to explore non linear projects to accentuate growth.

The best example to substantiate the above point in my mind would be that of OYO. Most of us would be aware about the tremendous growth OYO has seen in the last few years. Within the first 2–3 years, OYO became a household name and a brand synonym with hotels. However people would think about hotels or OYOs mostly when they had to move from their home city to any other city or for vacations. There were marketing campaigns and people who were focused on driving this “Linear Growth” for the Brand. However, there were parallel efforts to explore “Non Linear Growth” opportunities and there was a team dedicated to work on the same. Best example of non linearity in growth for OYO which comes to my mind is that of OYO evangelizing “Newer Use Cases” for their hotels. A lot of time, effort and other resources were spent to create this new market / opportunity for OYO. Once some of those use cases got popular, they started to contribute significantly towards the growth of the company. For example, the Unmarried couples category went on to drive almost 30%+ business for OYO during its growth days. Needless to say, if not for these nonlinear ideas, OYO probably would have been a very different company.

7. Saying “NO” is a Superpower 🦸‍♂️

As founders or Growth leaders, there would always be pressure to drive growth by experimenting with newer projects. There would be tons of ideas which will keep bouncing within the organisation. However, not all projects or ideas have potential to take your startup towards its stated goal. Every new project organisation works on needs resources which are scarce in any startup. Also, if the org is working on one project, it might mean it won’t be able to work on another project which might have much higher potential. So as leaders we would have to cut through all “amazing project ideas” and filter the ones which would necessarily take you closer to your objective. Every time a new person joins the organization, she will come up with ideas which have been either tried or passed on earlier. They might go on to insist on retrying on those failed experiments again. There will be pressure to repeat the mistakes which happened earlier. We would need to be objective and learn to say “NO” to such ideas.

At Shuttl, there have been at least a couple of instances where we went on to try to invest resources behind projects which had failed earlier. We learnt this lesson the hard way. But you don’t need to learn this from your own mistakes. Remember, saying No is a super power!

All startups need to prioritize their product road-map as well and in the process they need to say no to a lot of feature requests. Above is a brilliant framework designed by Amit Singh (Co-founder, Shuttl) which can help you prioritize your product pipeline. Remember, as Sam Altman had once said — “Until you’ve built a great product, almost nothing else matters”.

8. Complaints shouldn’t Overpower Passion 🙉

Let me start with an example from Shuttl. Around 3rd quarter of 2018, there was a phase when we were getting lots of customer complaints around operations and issues with the App. Most of the team members would spend the majority of their time to understand those complaints and to find solutions for those complaining customers. Some of us even started listening to calls on a daily basis to be on top of the issues. It was an extremely demotivating period for the entire team and their shoulders were down.

One day, just to reassure customers, we decided to have a field activity (Ground Zero) where all employees will go to our pickup points during operations and speak to our customers. We were expecting our interactions to be full of complaints as for the last few weeks our complaints were really high. However, contrary to our expectation, almost every customer we met was full of gratitude towards the service. We came recharged and motivated with hundreds of stories around how customer’s lives had changed because of Shuttl.

“I took a new job in Gurgaon because I knew Shuttl was there to solve my commute problem.”

“I could join back my office after maternity leave only because of Shuttl.”

“I use my travel time everyday to complete my course on data science so that I can get a better job.”

It was overwhelming to see that people were not only forgiving but also had gratitude towards the company. When we thought about our earlier perception, we realized we had recency bias and probably a small number of complaints were getting magnified because we were so close to the action. I am no way recommending that we shouldn't hear customers complaints but at times we need to zoom out a bit and see the larger picture. In your startup journey , it will be difficult to have even a single day without customer complaints. Along with ensuring that your startup works toward the right experience for your users, you also need to ensure that your daily involvement doesn't digress you from your core objective. Don’t let complaints overpower your passion.

9. In God we Trust, all others bring data 🙏

“Without data, you are blind and deaf in the middle of a freeway.” — This is a quote you would come across often. There is no doubt around the value of right data towards making key decisions within the organization. However often just going by data might not be enough as it may not reveal the right information. That’s where one-on-one communication with customers to deeply understand the problem statement comes into picture. Data combined with direct customer insights is that lethal combination to drive key decisions in the organization.

Sometime in 2012, I was working in the Usage & Retention team at Airtel. Our team at that point was trying to understand the trend behind usage drop in winters particularly towards the late evening. We looked at data for the last 2–3 years and we saw a similar trend for every year. We started believing the drop in usage in winters as a seasonal trend and somewhere started correlating it to shorter days of winters and reduction in reasons to speak to others. We spoke to a few of our sales team members as well to understand the same and they didn't have any particular new insight to help us understand this better. Once we decided to go ahead and speak to few customers to understand this better so that we could truly appreciate the “lesser reason to call others” as a reason for drop.

Speaking to those customers opened up a completely new dimension for us. We realized that a lot of customers would use small recharges on a daily basis for their talk-time. Those were the days when digital recharges had still not become mainstream and customers largely depended on retailers. In winters, a lot of those retailers would close their shop sooner because of cold weather and those customers would have no means to recharge their phones once they finish their talk-time. This was a big insight for the team. They realized that the drop was not because of “lack of reason to call others” but because of reduced availability of recharges at night. This led the team to launch a talk-time credit / loan product which did phenomenally until digital recharges became big. Had they just stuck to the data, they would have missed out an opportunity to drive better usage during winters.

10. User Growth >>> Traffic growth 💹

When we speak of growth, a lot of people confuse it for Traffic growth. They assume that the core job of a growth marketer is to get more traffic. In reality, traffic growth is only one part of the equation. Things start with traffic growth but ultimately its the user growth which takes the organisation closer to it’s objective. As marketers, we need to focus more on user growth instead of only Traffic growth. Understand your product deeply to get sense of what drives the conversion. Keep a track of the entire funnel for your product as it gives you early signs of what’s going wrong and why prospects are not converting to paying customers.

App Signup => Route search => Book First free ride => First boarding => Purchase Trial pack => Use Trial Pack => Purchase Main subscription pack

Above is the typical funnel for customers at Shuttl and at every step we find few customers dropping off from the funnel. Every funnel drop indicates something not working for the customer and hence the drop. For e.g. — If a customer has signed up on the App and not booked the First free ride despite finding the route, the reason could be lack of right offering (time of service, travel time etc) for the customer. Similarly customer booking first free ride and not boarding indicates operational challenges in identifying the boarding point or the vehicle.

Every funnel drop will give insights towards things that still need to be fixed in your business. Keep a track of drop at every funnel and see your user base growing.

11. Never trust a Digital Marketer ⚠️

Well, not all digital marketers but definitely most of them. The ones who reach out to you with a promise to “Turbo charge” your user acquisition and change your R.O.I. overnight. They usually represent Ad networks or affiliates but at times come from large Platforms as well. But why should you mistrust them?

Because, their only task is to sell their inventory to you and will mostly reveal partial / incorrect data pertaining to your campaigns. Their core objective is to showcase their platform / ad network as high R.O.I. channel so that you continue to spend money on the channel. Overtime, the digital marketing industry has got engrossed into multiple fraudulent activities by ad networks / affiliates to show R.O.I. for brands. Everyone in the ecosystem is currently trying to figure out ways to gain attribution of your customers to show R.O.I numbers for their channel. But alas, most of them are nothing but sheer waste of money.

I am not at all recommending you stop trying out digital marketing campaigns. Just don’t blindly believe in what they say or have to showcase their platform’s capabilities. Believe only your own data and judgement.

At Snapdeal, we ran a big budget video campaign on one of the large platforms for a specific category. When the team came and presented, they mentioned that our reach was higher, Share of voice had improved, awareness was over the roof and significant increase in consideration. However, our data suggested NO impact as our sales were absolutely stagnant. Make your own judgement which data you want to believe. My suggestion — Don’t blindly trust a digital marketer.

12. Frequency, Frequency, Frequency 📅

One of the under appreciated facts about most successful startups is that they were able to get their customers hooked by making their product a habit forming product. This book by Nir Eyal is a great reference for how to build habit forming products and can be a great starting point for any startup founder or product manager.

While having a great product is the right starting point, there are other frameworks as well to drive frequency for your product. I call this framework the “User state Model”. Idea here is to first split your entire user base into various user states as mentioned in the picture above and then aim towards a high number of your best quality users (core users). Best quality users are defined as ones which are doing attributes which you want all your users to do (e.g. 3 orders every month, 100+ Page views etc). Arrows show the movement between various user states with green marks indicating positive movement for growth and red indicates a movement which is typically negative for overall growth.

As a growth marketer, your objective is to increase the count of users qualifying as Core Users (overall favorable inter state movement) by moving customers from other buckets to the core user bucket. There is a natural movement of customers from one bucket to another at a regular interval. Track that movement for the last 3 months to create a baseline and then target an improvement over the baseline. For example, if typically 10% of new users end up becoming Core users in a given 30 days time, then your target should be to improve this % to a much higher number. Similarly if typically 5% of new users move to the Dormant bucket, your objective should be to get to a number lower than 5%. Only if you are able to improve the Green percentages and reduce the red ones, you will be able to increase the core users in the system. This is not a simple task to achieve. Every inter state movement needs to be tracked and worked upon as a project and there has to be someone responsible for the project (and that specific inter state movement). Within core users as well, there needs to be a project to ring-fence your existing core customers (% of users retained as core users) from competition and a project manager driving the same.

Favorable inter state movement across the board will show magical improvement across on the overall frequency and LTV for your customers. Personally I have used this model in a couple of organisations and have got massive +ve R.O.I. from the same.

13. Don’t be obsessed with competitors ⚔️

Finally, the 13th lesson. Let me start with an example. We all know about McDonald’s and how fiercely competitive is the QSR space they operate in. However, their story has great learning for all startups.

Constantly faced with competition, McDonald’s has simply focused on its customers and essentially minded its own business. McDonald’s has been a public company for over 50 years, so the details of its finances and strategy are all in the public domain. Its standard operating procedures are easy to obtain, anyone can get a job there and see how it all works, there are tons of books about McDonald’s including ones written by Ray Kroc himself. There are no NDAs. Everyone and anyone can copy it, and thousands have tried. McDonald’s has defended itself against all these competitors by just making sure that the experience its customers get is great, again and again.

As a startup, it very tempting to keep focusing on competition in your space. We are so keen to analyze what they are copying vs what can we copy from them. Keep evaluating what is their core strategy and how is that one up compared to theirs. However, we forget that the best way to keep the competition at bay is to keep doing the basics right. Every time.

When faced with hyper competition, just be like McDonald’s.

Ahhh, the article is finally over. Just like you drain the water from pasta after cooking, all my energy has got drained out in writing this long article. Hope you enjoyed it as much as I enjoyed writing it.

Do ping me if you found the content useful for your Startup. Cheers.