Thread by @NamanSr

I have taken 5 products from idea to the first 1000 users over the last 11yrs.

What have I learned?

You cannot get Engineering estimates right. When the team says it's done, it will take as much time after that, as the initial estimate.

Get UAT done as frequently as possible.
Give written instructions and ask the engineering team to explain them back to you.

Partnerships take time.

We launched a recharge site without Airtel. It matters. But does not matter.

Price comparison site without Amazon or ebay. That's OK.

Insurance site without ICICI Lombard. Just launch.

Get the product out of the door, partners will fall in line.

There is no big bang launch. Get a closed user set to keep using the iterations. (Closed Beta).

You cannot build a product if the first few users are not onboarded first.

It's nearly impossible to build a product flow if you are not a target user.

Stop mental masturbation in isolation.
Talk to a user. Observe them.

Differentiate between what is a product and what are bells and whistles. MVP.
Don't wait for the bells and whistles for the release.

You will incur a lot of tech-debt. That's OK. Don't re-write code unless you are very sure of revenue/PMF.
Tech debt generally makes the engineering team slow, rarely kills the utility of the product.
You will NOT get more users simply by clearing your tech-debt.

"One more feature" is tempting.
Identify if that will add more users at the top of the funnel (it rarely does). Acquisition, retention, monetization- what does the new feature do. Is that your focus now?

In build vs. buy, engineering will always select build.
A hardware company, I recently spoke to, chose to build their own invoicing system.

Always buy first. Figure out what is broken, give the vendor some feedback, then find another vendor, then build.

https://twitter.com/NamanSr/status/1251556037699801090

Keep talking to users and dropouts. You will get a lot of top-of-the-funnel ideas.

Use your product. Every week. Like a new user. Like a repeat user. Like an advanced user.

Individual Engineering team members always know lesser about the product than the customer support team.

Software products are part code, part copywriting.
Both are equally important. One cannot make up for another.
Bad code- unexpected behaviour.
Bad copy- comprehension error.

Excel sheet is your best friend. It's generally faster and more flexible than most analytics tools. Learn it.
Pivots, vlookup, charts, conditional formatting, basic string manipulation. Someone in your team should be a pro at it.

Share the progress with everyone. People love to see the impact of their work. Create that feedback loop for them. You need to be measuring correctly to even enable this.
The metrics is the dope they worked for.

This is equivalent to the like button on social media.

Things will break just when you thought you won it.

1st product - We sent wrong couriers to 5000 people.
2nd product - Flipkart had paid us and removed our code, in the same hour.
3rd Product - An engineer dropped a table from master DB.

Nothing is as bad as it looks at first.

Rely on 1-2 channels to get your first 1000 users. Generally, each channel will need 1 human resource. How many can you afford and manage?

It's tough work. Most products never get their 1000 users, ever. If you get them, preserve them.

I am building @RefrensApp for freelancers and agencies.
Follow me on @NamanSr for more on products and startups.