Wake Up Coffee - Startup Edition

I have heard this in multiple investor forums that the most important element of a pitchdeck is the team. A good team can raise money without even making a presentation. But, lets not delve into the exceptions.

For every investor, the team is important. The key questions in their minds is whether the team has the right skillset to build the startup? How does the team work together?

A startup pitch is very similar to story telling. Every entrepreneur is selling a dream; a dream the entrepreneur himself believes in and a dream that is achievable. 

One of the key element is the growth projection. Many people miss the subtle difference between projections and targets. An investor is not as much bothered about what projections you have. He is more interested in the targets you set for yourself and how you are going to achieve them.

Projections are a very reactive approach to growth. Where you are talking about what will happen based on the things you do. Here the focus is more on doing what is planned and hoping for the best in terms of hitting the milestones.

Any startup is a combination of two things - a product and a distribution network.

The product provides value to the users. While the distribution network brings the product to the users. While building the product is a necessity. It is not the only necessity. Before building a product, startups should invest time on building distribution networks.

Time an again, distribution networks have emerged as winners in every industry. This holds true for technology too. Most companies build the product and then burn a lot of money to buy distribution networks through ads or rewards. This happens mostly when the product and distribution network are not aligned. You achieve the holy grail when your distribution network is tightly integrated into the product itself. Like the Dropbox invite system, or Hotmail's email footnote. But sadly, this is not the case for most products.

Most organizations think of distribution as an after thought. First, they focus on product and only after they have a product ready do they start to look for distributing the product. Now they have a product that does not fit well into the distribution network or vice versa. This later prompts them to think about how to make changes to force fit these two things leading to more confusion and waste of money.

Mobile metrics are vastly different from web metrics. But most people have not realized this yet. We still are trying to superimpose the web metrics onto the mobile platform with a few changes. The same number that looks good on the web might have a completely different meaning on the mobile platform.

Lets focus on certain key mobile metrics that one should focus on - 

Ratio of Daily Active to Current Installs

This helps you understand usage frequency of the app. A app for daily consumption will have a higher number compared to an app meant for weekly consumption. This metric is not meant to judge the performance of the app, but more to do with understanding how people are using the app.

Ratio of Current Installs to Total Installs

This number is a direct measure of targeting. Does the marketing team target the right people? Do people install it because they want it? or are they forced to install the app? Many e-commerce players are relying on spammy mobile user acquisition channels to promote their apps. One can easily spot organic vs inorganic vs spammy installs using this ratio.

D1/D7/D30 retention rates

This helps one understand the stickiness of the app. Do people really find the app useful and do they continue to use it for a long period of time? A low D1 retention can be attributed to incorrect targeting. But, a low D30 retention is a clear sign of people not finding any utility for the app.

Ratio of Sessions per week per user/Sessions per day per user

This ratio helps gauge the retention and user acceptance (an indicator of  PMF - Product Market Fit). The more closer to the number of days in a week (7), the better your Product Market Fit.

This is a Product Management perspective of E-commerce on mobile App. Considering the latest trend of massive mobile adoption and Internet usage. Everyone is moving the direction of Mobile. So, now let us talk about an E-commerce app and its user cases.

Constraints of e-commerce App

It is not a daily use app - People don't visit e-commerce Apps daily. How to retain a user, who sees you as an app that takes up space in your mobile with limited usability? He can easily figure out 100 other apps (social, games, utility, etc) to replace you with.

Mobile is not designed for research - Mobile phones are not designed to help user either "read a lot" or "keep lot of pages open" or even compare two phones. So, users keep postponing their decision or not taking a decision at all. Categories like electronics, health products, etc require users to do a lot of research.

Horizontal vs Vertical category design.- If you are a horizontal e-commerce player, then you are bound by a single user interface across different categories. Different product categories require different visualizations on mobile (this is true even for the web, but more so on mobile)

Transaction behaviour - Your mobile phone is always next to you. So you are not required to complete any task at one go. You can take time. This can be clearly seen in chat usage, where users reply to chat intermittently and continue doing other tasks at hand. This behaviour is something most Apps have not been able to leverage. Most Apps simply choose to ignore the behaviour.

Opportunities on Mobile for e-commerce

We all know the increasing adoption, connected network and smart phone usage. These are generally true for all mobile apps. We are here to talk specifically about opportunities for e-commerce apps.

Different User stories - E-commerce needs to look at different usage patterns of consumers that lead to sales. Thinking about the user stories gives me Goosebumps. There are 100s of ways a mobile phone can lead to a sale compared to the web. We just need to explore them and take the risk to actually test them

Better personalization - A mobile phone is personal, it is not shared by a group of people. So, a user's history of using the app can easily be leveraged to better understand the user and provide more personalized look and feel to the App.

Has anyone realized the subtle changes being made on Twitter and how they are becoming more and more like Facebook?

The current state of Twitter - 

 The changes To improve engagement. The 140 character tweets have a lot of information embedded in them. It can contain links, images, videos, blog posts, news articles. All this information is lost in a simple textual snippet. Expanded tweets with Images definitely improve engagements.

Better UI and focus on User profiles - Twitter has historically given lesser focus on user profile and I would not say it is the most good looking profile page. Improving this feature will definitely improve user satisfaction.

It is still a tweeting service - Unlike Facebook which is more about my friends. Twitter is first about content and then about people. I still don't follow my friends on twitter. I use twitter to track trends, get the latest news and in general be in touch with famous brands or personalities. As long as this differentiation is maintained, I think, twitter will continue to have high engagement.

I still don't have a personal reason for using twitter - I use twitter more as a professional platform to learn and engage with topics and start-ups. Personally, I have no reason to announce to the world what I am doing or thinking. Twitter is yet to give me a solid context for posting a personal tweet. Why should I tweet? Who am I talking to? On a professional front, I see sense in me tweeting about my latest blog posts to the world. Or a celebrity talking to millions of fans. Twitter still needs to answer, why should I use it (as a personal platform)?

All these are small changes and will not lead to any massive growth in Users. Here are some features that twitter should consider introducing.

The next big thing - 

1. Groups - Introducing groups can help twitter bring in the personalization aspect. Groups will have people with whom you are more connected and care about. One thing Twitter should be careful about is - "how members will join a group". Let me explain with an example - 

Imagine I create a group called college-buddies. I start adding people to the group. But, here it is better to restrict the members to maximum of 10 people. It should not become a Group of all the people in the college (that is what lists are for). So I have a personal group where I can share what I find interesting in twitter. I can tweet messages directed towards a specific group. Since, everyone in the group knows me, there will be higher engagement.

Now there are 2 question that arise

How is it different from Whatsapp or Facebook groups?

How far away from twitter's core philosophy is this? (There tweets may no longer be public tweets)

The above idea is not very different from the WhatsApp and Facebook Groups. More thought needs to go into the design. Here are my after thoughts.

Another option is to create groups with hashtags. People are invited to be part of these hashtag groups. These 'hashgroups' are not the same as the existing hashtags. You can only see them if you have an invite to the hashtag (Ahh sounds complicated)

Group handles - These are group messaging handles, where I can create a handle for a group and everyone can contribute and converse with the group by tagging the handle. Every member in the group is part owner of the handle, he gets notifications directed towards the handle and has the choice of making it either public or private groups. The key here is ownership.. no one owns a hashtag on Twitter. People understand owning a handle. Now the question arises - "what purpose does the group handle really solve?? And how are people supposed to use it??"

Repurpose the lists into something useful. Make it an engagement channel. Lists are as good as non-existant on twitter. I see no real value add due to lists. There are no famous lists people aspire to join, there is no engagement within lists. It is a simple grouping of users.

2. Fix Conversations - Conversations on Twitter are different from every other platform. On a higher level, they are extremely simple - "Every part of a conversation is a tweet. Everything you do on twitter is a tweet". But, I feel this over simplification is not good. Simplicity is good as long as the smallest piece of the machine is easy to understand. A tweet is very easy to make, but it still does not have a reason to be made. With so much information flowing there needs to be sense of grouping of conversations which is more readily understood. I suggest a very minor change - Add the count of replies to a tweet i.e. on every tweet show the number of replies it got (similar to no:of comments on other platforms)

3. Fix User Suggestions - Twitter does a decent job of suggesting users to follow. I would like to do an AB Test by adding the following information along with the suggestions.

Last Tweet along with profile details

Summary along with profile details (my bet is on this)

Last Active date along with profile details.

4. Get rid of the Top Navigation Tabs - The top navigation panel is a distraction. I only need the "Home" tab. The information in Discover needs to be merged into my Home tab information. I really don't need another dedicated tab where I have to go and see what's trending. Notifications and Me tabs can become icons so that they occupy less real estate. People will find it when they need it.

We had by far the best content in our category. We had the most visually aesthetic and informative content compared to our competition. But, our SEO rankings were abysmal. Initially, I was naive enough to assume that it was an architectural issue, and thought that Google was not able to find all the pages on our website. I went ahead and fixed the website - created a clear hierarchy, added breadcrumbs, added sitemaps, schema.org markup, etc. It did not help much in improving the rankings.

Google was still not able to figure out the quality of our content. At the end of the day, it is just a machine.

So what did we do?

We went for completely old school SEO tactics. There are few simple SEO rules everyone suggested. Initially, we ignored them, because they were not really adding value to our users. But, we could not ignore them forever. We agreed to play by Google's rules and decrease the quality of our site by adding elements that do not add value to our users. Here is what we did - 

Repeat the keyword at-least 3 times in the page. Initially we did not have any reason to repeat the same sentence again and again in the body. There was only one H1 tag with the keyword and even that was a shorter version of the page title. No one really cares about the obscure filters of "XL or L or M" or "the exact shade of the shirt" as long as the content matched the user's expectations. We had intelligent technology systems in place to ensure good suggestions were provided to the user. The title was kept short and crisp for easy understanding.

Improve Internal linking - We wanted our users to engage with the existing page and not move to other pages. But, as Google suggests, internal links are important (it did not make sense for us, because we wanted users to stay on a single page for more time)

Create meta content pages - Similar situation as above, where we had to aggregate content and create new pages. These pages were supposed to cover a specific topic or category. Even this did not fit into the product, as we wanted users to concentrate on individual items and not be overwhelmed with aggregated pages.

Add content to alt text - This really did not affect the user, but really did help improve our rankings in image search

Add proper file names to images - This too helped us get better rankings

Make content easy to discover for search engines - This forced us to add unnecessary links in the homepage leading to internal content pages. This led a lot of people to skip login and start browsing the site decreasing conversion ratios.

I have finally realized that Google has still not cracked the formula of valuing a website for its actual content. I have seen competitor sites that have zero content and ranked better than my site. All they had was a passage at the bottom of the page, which repeated the keyword in 3 different ways.

Lesson Learnt

I hated this Prof. He actually knew nothing about the subject he was teaching. It was a very interesting subject that was converted into a piece of shit and shoved down our throat.

But, in all the things bad. There was one thing that the Prof said that stuck. It was nothing to do with the subject, but more a life lesson. My Prof always used to say - "Ask Why 5 times - Why? Why? Why? Why? Why?".

Now I do the same in everything I do. If I have to build a feature the first question I ask is

1. Why do you want to build this feature? This generally gets the lamest answer. Never stop at this point.

Answer - To beat competition

2. Why do you think you can beat competition with this feature?

Answer - We will provide users better experience

3. Why do you think the users will have better experience because of this feature? - Here is where most discussions go into "MMMmm's and Maybe's" - Get your numbers right here. 

Answer - We have seen users request for these features

4. Why do you think users requested this feature? -

Answer - They wanted to perform a certain task.

5. Why do you think users want to complete this task?

Answer - They want to be able to share easily with friends.

Trying to benchmark yourself based on competition is the most natural way to present yourself to the world. You build a product X, and then you identify competitors Y & Z. You list off a bunch of features that makes you better than Y & Z. This is good for marketing your product.

But, don't take this to your heart while building a product. Your customers should be your benchmark, not your competitors. Imagine if your competitor (e-commerce) gave their customers a feature to filter by color. Now you want to outdo your competition and you decide to give your customers a feature to filter by a combination of colors (like a dress with red and blue). Now you are better than your competition. You have better features than them, and your marketing team decides to start promoting this new improvement. This is an easy decision to pass through, as everyone in your company wants to be better than competition. The clueless Manager agrees as he cannot risk sticking out his neck to say "NO", because everyone is excited about being "cooler than competition" and beating them to pulp. 

Here are the 3 scenarios that could happen. Scenario 1 - Your customers can be overjoyed and decide to flock to your website, just because you have better "color filters". Your traffic starts to pick up and you are on your growth trajectory again. But, you most probably will also be in Alice's Wonderland. So coming to the other Scenario 2 -  no one notices it. There is no change in the final conversion. Now you realize that people think only in a single color. No one thinks about a combination of "red and black". They only think about a "red dress" or a "black dress". The few who do think in combinations are too few and far in between. The final Scenario 3 could be even worse. Now since you started highlighting an unnecessary feature, people have started abandoning your site. 

All this because, you decided to beat your competition and not to improve your customer experience. You cannot be competitor-centric in your approach. You always have to back your decision from actual customers. Or at-least based on your understanding of your customers.

So firstly, let me clarify that I don't want you to be spending precious money on hiring a Product Manger from outside your company. Your best Product Manager is most likely part of the founding team.

Here are the questions you need to ask to identify the Product Manager in your founding team. It could be you or your partner or your first development hire. 

Question 1 - Is he open to taking opinions from people? 

Product Managers are supposed to absorb everything around them and come up with insight to be able to define the product. If he cannot handle the opinion of others well, he most likely is going to fail. He has to know how customers think and what they are using your product for. He needs to understand how marketing has understood the product. He needs to understand what the developers say when they are planning to build the next coolest feature.

Question 2. Is he thinking long term?

There are 2 sides of product management - tactical side and the strategic side. If a product manager is more tactical, he is solving short term problems (this is good enough, but a start-up cannot survive with "good enough"). The person has to be able to think 1-5 years ahead and work towards the long term goals. Failing to do so can show good short term results at the cost of long term sustainability.

Question 3. How frequently does he use numbers to justify his idea?

A Good Product Manger should always act as a gatekeeper for features. He will get requests for a feature from everyone - Users, Investors, Developers, Sales, CEO, everyone who has a mouth will have an opinion and idea. Actually, I would love this to happen. Who doesn't want to get people to think and come up with ideas to improve the product? 

But, if you start accepting every idea that comes your way. You will be filling up the product with more garbage than anything else. So ask yourself 3 questions before accepting a feature.

Question 1 - How will this feature improve the customer experience? You might have the coolest technology and a completely 3D modelled viewing experience and lots more. But, is it helping the user achieve his goal faster and with less pain? Or is the customer happy with what he had previously?

Question 2 - How frequently will the user use this feature? If the user has to go 10 steps deep, to use a feature that is going to impress him. You are better off not building it. Ensure everything you do adds significant value. In a start-up everything you do should have 50-100% impact. Else, you are wasting precious resources.

Question 3 - How much resources do you need to invest? Is it a 1 day task for one of your developer? Will this feature force you to make more changes across the system? Is it a standalone feature that can be outsourced? Will more support features follow once you actually implement it?

Money is the lifeline of startups. In the initial days, money is the only thing that keeps them alive by helping them hire talent, pay salaries and continue operations. There are 2 kinds of money sources in the ecosystem.

Dumb Money - The best money you can get in the first few months of your start-up, as long as you know what you are getting. This is when the investor is mostly ignorant about the market space and has no previous experience of investments (friends and family). This is easy to get and will help your start-up get started. These investors are easily impressed and are mostly investing as a hobby. Ensure that these investors know the risks involved and are completely aware of it before giving you their money.

Smart Money - This is more tougher to get. You need to really impress the investor with your knowledge of the market, with your MVP and some initial traction. Here the investor has a fair understanding of the market and has been involved with startups for some time. They know what parameters to judge you on and where you are stumbling. Being upfront and honest about your successes and failures is a good choice here.

An investment round is a "test of fire" for the start-up, where the start-up needs to prove the validity of its dream with meager resources. The start-up has to focus on operations, marketing, technology, hiring and scaling, and at the same time engage with investment firms. All this with a team size that can be counted on a single hand.

There are two kinds of investment firms - one the financials oriented investment firms, whose main goal is to see and understand start-ups in numbers. Everything about the startup is converted to numbers and decisions are taken based on that. The second kind of investment firm is the operations oriented firms, who want to understand everything about the business and then decide to invest in them. They know good numbers are a necessary condition, but go one step beyond numbers to truly understand the market space and factor in the qualitative aspects.

One thing that again differentiates these firms are the time frame they see themselves contributing to the start-up. Yes, most of the investment firms are well connected in the industry, and can help you get the best talent. But, if the investors themselves lack a clear understanding of the market, and their only goal is to see how the next investment round is going to happen, then, there is a disconnect between the founders and the investors. Let me explain this - one investment firm might ask - "What parameters do you need to get the next round of funding?" and another firm might ask you - "How are you growing?" ... both these questions will give me the same answer, but in the former question, the need to growth is to paint a "picture" of growth to investors and not actual growth. 

I have seen many firms, which are close to an investment round, that start using "not so moral" means of advertising, just to jack up the numbers. Since the goal is to paint a rosy picture, and are confident that they will get away with it. It almost makes me feel that "Due Diligence" by the investor is a joke.

I have been working on fixing SEO for my own website. This is what I have learnt in the past few months. What the Internet says about SEO is mostly bullshit. Some of them are still in the pre-panda era, others expect content to do everything for you. Here are few example sites I would like to talk about, that you should learn from.

1. eHow.com & wikihow.com

The most important learning from these two websites is keyword research. Keyword Research is more important than the content itself. Content on these pages can be at best rated 'average'. Yes, on few occasions, they are the only site that show up in search. But, mostly it leaves you as confused as you were, before visiting the site.

Coming back to keyword research - This is going to either make or break your SEO rankings. 

2. distancesbetween.com

This is a site that almost always shows up in the first page of google when someone searches for distance between 2 cities. They almost completely own any term that has the keywords "distance between" query.

So what do is the learning?

Keyword Research - Find out the supporting keywords for your service. You will never rank high for words like "new york", "apple", "iphone". These keywords are owned by the brands themselves. You have to research on what supporting terms are associated with you keyword. Here are a few examples

"buy iphone 4s"

travel to new york

tourist attractions in new york.

apple iphone charger

Once you have the keywords listed. You need to focus on internal linking and build a hierarchy inside your website. Create content for all the keywords that you have identified. Try to dedicate one page for one keyword with - good content will help ;). Now start linking pages in a hierarchical structure. The page with the top level keyword (that encompasses all the pages) should link to all lower level pages. This allows Google to understand your website better and pass link juice to each page.

Product Managers should always be careful on where they are investing the companies resources in. All decisions should be backed by data and not on any individual's perceptions. Just to be on the safe side I do a simple thing - Say "Fucking Never" to myself every time I have to do something. Only if I can really convince myself, do I invest time and resources on it.

Here is a conversation that will help you better understand this idea -

Random Guy: I have a great idea. We should do this and we will become the next big thing.

Me (Thinking): Fucking Never.

Me (Saying): "That really seems to have some potential. Let us discuss it a bit more."

Random Guy: We need to build feature X, Y and Z. Then the user will be able to do more things with our product. They will be more happy.

Me (Thinking): Fucking Never

Me (Saying): So this seems like a secondary feature on top of the primary feature. Don't you think, we need to focus on improving the primary feature and make people more happy there?

Random Guy: But our users have more things to do on the product. We should be able to give them everything they want to do. And let them choose what they want to do.

Me (Thinking): Fucking NEVER!!

Me (Saying): Do you want to make your new idea as the primary feature?

Random Guy: No. Our primary feature is what we had envisioned about. This feature adds more value for the product.

Me (Thinking): Fucking Never.

Me (Saying): I need to understand more about this feature. Is it really necessary for our users? Do they really want it? Will it really add any value? How much percentage of users will this feature help??

Random Guy: I think every user on our system will love to have a choice. He can choose to either use or not to use this feature. But we should provide it to them.

Me (Thinking): FUCKING NEVER!!

Me (Saying): I would never invest time on a feature that is just an option. At the current stage, we need to figure our our acquisition and retention rates. I need to figure out, if this feature will help me improve those parameters. Does any of the analytics help substantiate this investment of resources on a new feature? Are people trying to do this, with us actually providing the feature?

Random Guy: There is no way I can test this, without actually building the feature. If the users have not seen the feature, how will they know if they want it? We need to first build it and I believe they will like this feature.

Me (Thinking): FUCKING NEVER

Me (Saying): We cannot take a decision on building this feature, without data. I think we should run a smaller experiment or observe user behavior and see if they are trying to do something similar. We should also send out a survey and ask if users want such a feature. let us both go through the data and see if users are looking for a feature like that. I will draft a survey and send it to our beta users.

Random Guy: I will see how to build the feature and how much time it takes to have it production ready.

Me (Thinking): FUCKING NEVER

Me (Saying): Let us first concentrate on figuring out if our users want it. I would want your help in figuring out what our users want. If we do that, we can actually invest more resources along with you and build the feature.

Random Guy: Ok

Me (Thinking): OK

Here are the top 10 mistakes founders do that can slowly kill a startup. You will never know when you have burnt all the cash and still have nothing to show for. Here are the list of things founders think - 

We need to reach out to more people. If enough people know me, they will start using our product. Reach does not improve product usability. If you cannot be useful for a 1000 people, there is no way you can be useful for a million people. Any actions based on this thought is just equivalent to burning money.

We need that one big ticket event. - Startups are hard. You cannot rely on one big event to build sustainability. You need to keep trying and figuring out how to build a useful product. Once you have Product Market Fit(PMF), only then, can you bank on the big ticket event.

We need to do a lot of PR to get users. PR is mostly overrated. It gets you visibility, but does not guarantee any conversion. If your product doesnt add value, people wil not stick to your product.

We need to build more features to get people interested - The decision to build more features, means your product in its current state has failed. But, if your primary feature has failed, then, any secondary support features you build will not contribute to growth.

We need to support every platform that exists - In a resource constrained environment, you can only support few platforms. And only after you have PMF, should you consider supporting a new platform.

I need to scale to different categories and cities. - Before you do this, decide whether you have captured your current segments. Else, you are again burning money.

We need to outsource our work to cut costs - Outsourcing in an ever changing and dynamic environment is suicide. You will always be dependent on people who have other priorities and decreasing patience on your product.

We can get a VC to invest money, by showing some temporary growth. - VCs are the smartest people in this industry. They might be wrong a few times, but they are mostly right. They will ensure that they scrutinize everything you have done and are capable of doing before investing.

We need to tie up with a bigger player - Bigger players are not very interested in you. If they do tie up, they will have their own agenda that you need to align yourself. Unless you are closely aligned, you will have to compromise on your own vision.

All through my MBA, I had a dream. I will one day walk into a room full of people and present my "new idea". I will present the idea with a lot passion and vigor and at the end ... at the end everyone will rise from their chairs and give a standing ovation. 

But, I soon realized, within a few months into my own startup. Only results mattered, not the idea. You can share your idea in a presentation with everyone else, but if your idea doesn't match to their own idea for the product, you are not going to impress them. So how do you get every one to love your idea?

Inception - You have to take your idea, and make it their idea. You need to make them believe it was what they always wanted to do. An idea that they came up with. Presentation plays an important role.

Identify the problems - Make sure everyone is aware and accepts the problems your startup is facing. If everyone is not solving the same problem, there is little chance you will have the same/similar plan of action to solve it.

Identify the Goal - You have identified 10 problems. But, your goal is to decide the biggest problem you want to fix, not everything. Agree on what problems you want to fix. This will increase your chances of agreeing on a solution.