Posts

There are three ways to create negative churn that I have observed in the market. First, usage expansion. Second, feature expansion. Third, product expansion.

Usage expansion is the most common way to create negative churn. Utility based pricing models like buying SMS credits on Twilio, or compute on Amazon or data processing on Mulesoft lend themselves to gradually increasing account sizes, as customers use more and more of the product. Per seat pricing also leads to usage expansion if product lends itself to growing within an customer. Slack and Expensify might win more seats as word of the products get out and the company grows.

Founded in 2006, Mulesoft is an 850 person company based in San Francisco that builds data integration tools. The company started originally as an open-source product and then focused on its paid offering. Today, the business generates nearly $200 million annually in revenue, and is growing at 70%. The business filed to go public last week, and the documents reveal a very impressive business operating at scale.

The chart above shows the subscription revenue is roughly at $150 million and professional services revenue is about $35 million this year.

We’ve taught computers to do many things. We’ve researched how to teach them to identify cats, spot fraudulent charges, even categorize cucumbers. But what can we apply in our daily lives that computers have taught us? That is the premise of the book called Algorithms to Live By. Which of the advances in computer science can be applied to laundry, choosing an executive assistant, picking the best strategic plan and optimizing your schedule?

From the millions of Amazon Alexas to the self-driving car, new products are coming to market infused with machine learning. The innovation offered by machine learning techniques are real, and they will changed the SaaS world. But how? How can startups use machine learning to their advantage?

There are four broad applications of machine learning:

1. Optimize - this morning, fastest way to travel from Sand Hill Road to South Park in San Francisco is highway 101. The job requisition for an account executive on our website uses too many clichés. To close more business, speak slower, talk about pricing later in the call, and use this case study.
2. Identify objects - the photograph you just took with your smartphone contains a cat. Find all red plaid woolen shorts in an ecommerce store. The CT scan shows high likelihood of Parkinson’s Disease.
3. Detect anomalies - your credit card shows a $10,000 charge for a piano from a store in Nairobi. Your server cluster is operating at historically high CPU usage. Customers are responding to this morning’s lead generation email at 25% greater rates than last week’s campaign.
4. Segment data - customers who come to our product through the mobile app store show 15% higher engagement.

These applications alone make for tremendous advances. But, combinations of these applications lead to incredible things. Object identification + anomaly detection + robotics = self driving cars. Or brick-laying robots that erect walls three times faster than humans.

One of the hardest thing to do in sales, especially for early stage SaaS companies, is to disqualify customers. When a startup disqualifies a customer, they turn away a revenue opportunity, a chance to add $1k of MRR or $3k of MRR, and meaningfully grow the top line. But if the customer isn’t the right customer, that incremental revenue bears a hidden cost.

In the earliest days of the business, those potential customers waving checks promise an attractive revenue boost. Imagine a startup at $15k in MRR. A customer prospect worth $2k in MRR is a chance of growing monthly revenue by 13% in one shot. In addition, a seed stage or Series A startup typically has a narrow customer pipeline. Because the number of interactions is “rare,” each interaction with prospect seems precious. The business feels an urgency to grow, so the temptation to accept an non-ideal customer increases.

Advances in machine learning are transforming the software world. Two of the most exciting applications of machine learning are speech recognition and natural language processing. After researching the space for more than a year, we are thrilled to announce our investment in and partnership with Chorus, a pioneer in speech analysis for sales.

Chorus has a unique technology that enables it to listen to inside sales phone calls and provide real-time feedback to salespeople while they are speaking on the phone. Chorus provides call transcription, keyword analysis for competitive tracking, among other features. Most important of all, Chorus understands what the best sales people do to close more business, and teaches everyone in a team to sell in that same way.

Win probability charts like the one above have become the icons of popular predictive data analysis. I love data, but let me whisper a heresy to you. I detest these charts. I Instead of provoking thought, insight and questions, they close minds. They support the ideas of inevitability, of odds too great to overcome.

Congratulations to the Patriots who faced a 99% probability of losing the game, but passed the ball down the field into the endzone twice, and rendered this fallacious prediction worthless. Another example for Nassim Taleb to use if ever he writes a sequel to the Black Swan.

I met a physicist this week who told me all the Nobel laureates he had met in his studies have been the most modest of physicists. “They realize how small they are in the world, after discovering something incredibly special and new.” Separately, I referenced an executive this week. A former colleague of this person told me," this is not a person who sees a model work once or twice, and instantly subscribes to the notion that it will work every time for every business."

Which is the more important priority? Growth or churn? Churn or growth? Early-stage companies have limited resources to focus their efforts. On one hand, growth is important in order to raise a venture capital round. Growth shows demand for a product. On the other hand, churn is a huge source of friction and raises questions of product market fit.

Especially in the early stage, churn is the more important of the two priorities, and when founders ask me which to emphasize that the seed and series A, I’ll always respond churn.

How much should a founder raise for their startup? I imagine almost every founder contemplating a fundraising round ponders this question. There are many different paths to developing an answer. The right answer that every startup founder has told me is as much capital as possible at the highest possible price. But what strategies exist to justify increasing the round size and consequently price?

These are the three most common I’ve observed.