2/10/2014 Building Product, Talking to Users, and Growing — Adora Cheung, Founder, Homejoy

startupclass.co

Have a lot of time on your hand. When you start a startup you should have a lot of time. If not, it is better to work two days straight on your idea versus spending two hours here and there.

What are the things that most people do in startup that are wrong? The n00b approach:

1. Build product in secret

2. Exclusive press launch!

3. Wait for users

4. Buy users

5. Give up

Problem

You should think what problem you are really solving. You need to describe it in one sentence. What is it? How does it relate to you? Are you really passionate about it? Verify others have it. Let’s say you have a problem and you can state it, where to start?

Lean a lot, become and expert

People that have been working in the industry for a very long time, have used to the things as they are. That is an advantage a person new to the industry. You need to spend couple of months understanding bits and pieces of it and how it works, this is when you will start seeing things that are really inefficient. For example to go to a course or become employed in the company of the industry. Get into the shoes of the customers from all the angles that you are trying to build.

Be obsessed with your competitors. Go through all the pages in the search results.

Identify customer segments. Even though at the end you are building your product for all the customers in the world, first you have to focus on your main customers and their needs.

Storyboard idea user experience: before you even start coding or working on your products, you should write your ideal user experience of how you are going to solve the problem, starting with how your customers will find out about you, come to your site, what does the text there say, what do you communicating to them, how they sign up to your product or service, what they are actually getting and what happens after they finish using it, will you send them a survey. You need to visualize every single step of the perfect using experience, then put it into the paper, then into the code.

So you know all about your main customers and industry, time to build..

V1 – MVP

Many people skip this part and start building the whole perfect users experience. Minimum viable product means literary the smallest feature set that you need to build to solve the problem that you want to solve – users’ immediate needs. Need to talk to your customers!

Need to develop simple product positioning: this is XYZ in one sentence. One liner is very important, it describes the functional benefit of what you do. Later you can speak about emotional part of the brand or whatever, but in the beginning you need to communicate what users will get out of it. So, you have your MVP, how to get your first users start trying it?

How to get first users and who are they?

You! Your mom, friends +co-workers, online communities, HackerNews, reddit, local communities, mailing lists, nice influencers, mommy bloggers, cold calls + emails, press. Go places where your potential users are or where people are (fairs, etc). So you have users using you, but what do you actually do with them?

Customer feedback at this stage is vital!

1.Make sure there is a way for people to contact you, for example support@..com, leave your phone number, don’t neglect the voice mail, so you don’t have to pick it up all the time.

What you should be doing is going to users and talking to them, get out of the building and do the work! This will help you to understand which features you need to build and which to skip in the beginning.

2. Surveys are ok, but people will respond only when they are either very love you or very hate you, you never get in between.

3. Go to meet people using your products. Make it as an interview, make them feel comfortable, so that will be honest with you and help to improve the product. For ex, taking people to drinks.

4. Track how you are doing on general perspective - what is customers retention: how many of them come back, how often do they user your service, but it takes a huge amount of time to understand that.

A good leading indicator is collecting reviews – five, four etc star reviews. Try to calculate NPS (net promotion score): you are basically ask users — from 1 to 10, how likely they will recommend it to their friends. So, over time, if the score goes up, seems you doing great job, if it goes down – bad job, and if it is flat, you need to go and talk to your customers, understanding what new things you should be building.

5. The one thing you should be remembering about is the honesty curve: let’s say axis X is the Degree of Separation, axis Y is Respondent’s Honesty. In case of a free product, closest to you people won’t give you an honest feedback (cause your mom or friends love you). But the further it goes, the feedback becomes more honest. But at the same time totally random people don’t care about giving a proper feedback.

The opposite is happening with a paid version. Paying random customers will tell you straight if they don’t think that the product worth the money they are paying. So, if you are thinking about releasing a paid version, do it as soon as possible.

So, what to do before the launch.

Build fast but optimize for now, for this stage of growth, meaning if you have 10 users, do not create features or infrastructure for 1 000 000, think about the stage you are in, that is 10 – 100 users, what features you need for that.

Manual before automation, meaning before having software for every step of the process, manually do it yourself first to better understand the flow.

Temporary brokenness > permanent paralysis: do not worry about the edges and perfection, worry about who your core users are going to be.

Beware of Frankenstein: when talking to different customers, they might be suggesting different features. Do not build them straight away! Start asking ‘why’ they need them. Usually they are suggesting the best feature, but most important they are talking about their problem trying to understand a real reason and new problem behind that.

You have a product that you are ready to ship. SHIP IT! Someone will steal your idea anyway, so there is a first-mover advantage.

Ready for a lot of users?

Don’t try different growth strategies at once, take one channel at a time (FB, google, etc), understand what is working, iterate working channels, revisit failed channels later.

There are three types of growth:

1) Sticky growth: trying to get your existing users come back and use your product more. Users experience is the core!!! Analyse CLV and retention cohort (customers segments).

2) Viral growth: when people talk about you to their friends, etc. You have to deliver really good [WOW] experience and have good mechanics for referrals:

• Customer touch points (where people are learning that they can make referrals). You want to put opportunity for referrals where they [customers] are highly engaged and loving you.
• Program mechanics (invite your friend, receive 10 usd, ect)
• Referral conversion flow ( when customer’s friend clicks on the link, this flow should be very optimized)
• Know the key scenarios / time for switching from your competitors’ products.

3) Paid growth: CLV should be always bigger than CAC (customer acquisition cost). Good to look at paid growth by breaking it down by customers segments.

Sustainability is the key.

Art of pivoting. When to pivot?

If:

1) Bad growth

2) Bad retention

3) Bad economics

Growth is the trickiest part. Make an optimistic plan of growth. For ex, get 1 user in week #1, 2 users in week #2, 4 in week 4, etc. It is pretty easy if your product is wanted. But if 3 or 4 weeks you have zero growth, you are probably doing something fundamentally wrong. The growth line is curvy, but it always have to go up.