Live from Business of Software 2011 - a guest post by Engineer and Product Marketer Magnifique Justin Goeres, @justingoeres. In these notes, Dharmesh Shah, founder of Hubspot and author of the well-loved book, "Inbound Marketing,", shares his brilliant insights on building a super-successful company.

Disclaimer: Justin took these notes in real-time and they went up for your immediate enjoyment in their original raw form.

If you thought kitten photos were great on slides, BABIES ARE BETTER!

Not going to talk about what HubSpot DOES, but why HubSpot IS.

How does HubSpot pull "non-consumers" in?

HubSpot Customer Growth - over 5000 customers now. ~290 people. 2nd fastest growing software company in US.

Dharmesh's blog: onstartups.com

$30,000+ PERSONAL meal budget -- no meal expense ever questioned -- you can buy ANYONE a meal. This was an alternative to the "golf" budget of executives, where golf is expensive and frustrating.

Randomly sends out emails to people in a city to invite them to meals when he travels. Millions of dollars in revenue can be traced back to these meals. A lot of his learning has come from 'onstartups' meals.

"You're a precious little snowflake. Just like everyone else. Get over it."

We all make remarkably similar mistakes.

"Most of the time I have no idea what I'm doing."

Goal is not to improve your business, but to improve YOU.

Purpose of your business is to create happy customers.  (not to make customers happy!)

If you sign a suboptimal customer, you've already created a ceiling on how happy you can make them.

Cancellations kill SaaS companies.

Types of churn:

1 Customer - # of customers that cancel, on a % basis

2 Revenue - had 200k of revenue, lost customers representing 20k - 10% churn rate.

3 Discretionary - discounts for yearly subscription, etc., -- of the customers who have the option to cancel their subscription, some cancel - you shouldn't get credit for yearlong subscriptions who don't cancel THIS MONTH because they CAN'T cancel THIS MONTH. 

4 Involuntary - customers who cancel due to external factors (changed jobs, etc.)

Typical financial statements are Revenue Churn. That's not good or bad, but it's important to know.

CHI 2.0 - Customer Happiness Index

Single number that quantifies to what degree a customer is happy (had this since HubSpot's first year)

goal was a PREDICTOR of churn.  "Here are the customers who are most likely to cancel?"

New change for CHI 2.0: instead of solving for how likely the customer is to cancel, algorithm now solves for how likely the customer is to be SUCCESSFUL with HubSpot

- achieve marketing success.

- how much growth in customer blog traffic.

- etc.

Encouragement: AT LEAST do CHI 1.0 -- try to predict your churn.

New CHI 2.0 is now EXPOSED to customers.  Old CHI 1.0 was not.

Treats each month of new customers as a "class" (like in school). Show a leaderboard where they can compare their success with other people who joined at the same time.

Don't fall in love with your business model too early.

You will learn surprising things and you need to be able to change to account for them.

"Pricing is hard. It's very, very hard." - Simester

Have spent the most time debating PRICING on the HubSpot management team.

The price is always greener in your neighbor's company. You will drive yourself crazy if you try to look at them.

Cheapium - An alternative to Freemium:

It costs money to support free users.

Devalues the product: "Is the paid version THAT MUCH better than the free one?"

*Cheapium* model:

Suppose you charge your former "Free" users exactly what it costs you to provide the product.

Makes "cheap" users break-even.

Improves the quality of the customers you do get.

Pricing is hard, pricing changes can be even harder.

See ZenDesk.

See Netflix.

Assume customers are CONNECTED and UNITED.

If you go and calculate pricing changes, don't ignore that pissed-off customers have a microphone and will unite.

It's OK to change prices, but think through the implications if everyone on the internet who knows about your brand was responding.

Rule at HubSpot: Grandfather in your existing customers whenever you change prices. Don't screw existing customers.

Hubspot is maniacal about sales data. Every time they announce a price increase, their sales go through the roof (because it's their last chance to lock in the cheap number).

At what point do you need humans to sell software?

1 - when the product is complex

2 - when the market is new (people are not goggling for it) - evangelical sale

3 - when the price is somewhat high

You want the right features in your product and in your customers. Cherry pick.

Q: if you're in the enterprise business and your revenue depends on specific customers, how can you cherry-pick?

A: It depends, but it's tricky. Enterprise software sucks from a startup perspective because you're saddled with convincing LARGE customers/companies to make a binary decision.

If you're doing a startup, especially if you're self-funded, try to find a way to start smaller. Find a way to pick a smaller market rather than going after elephants.

"Strategic" is code for we don't have data but we think it's good.

The donuts are in the details.

Not every insight needs to be counterintuitive. Just because it's obvious to you doesn't mean it's not a great blog post.

MBAs are not intrinsically toxic.

"Smart and Gets Stuff Done." - that's how HubSpot hired their VP of Sales, VP of Customer Happiness, etc.

You have to be reasonably smart.

As code hackers, we need to appreciate the value of business hackers.

You may be asking yourself…

Self, what about work-life balance?

"Smart people will often get their asses kicked by other smart people that work harder." - Dharmesh

"We are not family. We are a team." - NetFlix

You get to pick your team. You don't get to pick your family.

Colleen Coyn - '98 USA Women's Olympic Hockey Team.

How important was it for you guys to just work real hard?

"In athletics you have offseasons, and you have physical injuries. That's the world's way of telling you to stop working. In business we don't have that."

When did you decide you wanted to be on the Olympic team?

She said it on video when she was 16, when there was no Women's Olympic Hockey Team yet.

Building a largish software company requires deliberate decisions based ON THE GOAL of building a largish software company. It doesn't happen by accident.

Generalist vs. Specialist

Generalists are awesome in the first days of a company. As you grow, specialization starts to matter.