How we grew Primary Metric in YCombinator — Start Up Note Part 4 (Translated)

Author: zmzlois


Primary Metric and Secondary Metric, Product Market Fit, Paid monthly user retention curve.

By Alan Chan

In my previous article, What we did before being accepted to YCombinator — Start Up Note Part 3, I talked about what we have done before we were accepted to YC and the application process. During our three months in YC, Heptabase iterated 60+ versions, recurring revenue increased 40 times and started to be profitable. In this article, I will share what I have learnt and what we did to accelerate the company’s growth during our three months in YC.

YCombinator Y Combinator

Startup = Growth

The best reference to understand what a company would experience after joining YC is the all-time classic, Startup=Growth, written by Paul Graham. I think this is an article that every start-up founder should read. Its notion prevails in the title: The essence of a startup is to grow, continuously.

To grow faster, and achieve a start-up’s vision on a larger scale, start-ups often fundraise. Likewise, an investor would decide to invest in a new company because the investor believes in its future value and scaling opportunity. The common goal of ‘Growth’ between investors and start-ups makes the deal happen.

If you are starting a normal company, namely a restaurant, or barbershop, you don’t need to deal with the capital market and pressure related to high growth as long as your profit is enough for you and your employee’s living. But if you are doing a startup, continuous growth is your innate mission.

As an accelerator, YC’s mission is to help Startups grow at the highest speed.

Primary Metrics and Secondary Metrics

Metrics Photo by Isaac Smith on Unsplash

For most software start-ups, the primary metric is revenue if you are making a for-profit product, or daily active users(DAU). If you are making a free product(community website like Facebook, etc). In YC, almost all companies focus on only one thing: do whatever it takes to increase the primary metric.

There are a lot of factors influencing a start-up's primary metrics, and YC called them secondary metrics. For example, if you are making a SaaS product, the Primary Metric is revenue, but secondary metrics might include paid user retention rate, conversion from landing page and traffic on the landing page, etc.

As a start-up, your resource and time are limited. You can make your landing page as pretty as possible, and change your wording to increase conversion; you can use advertisements to increase the traffic to the landing page; you can iterate your product, talk to users and increase paid user retention. All these gestures are good for your secondary metrics, and if any secondary metrics increase, it will lead to the growth of Primary Metrics.

And which metric is the most difficult to grow among all the secondary metrics? The answer from YC is paid user retention.

For most software start-ups, user retention is much more difficult than obtaining users. You will face a bottleneck if the retention is low and your leaving user number is higher than your onboarding user number, even when conversion or traffics are high. Products with a higher retention rate have higher growth potential, and their number of new users is higher than the number of leaving users. You can buy traffic from advertisements, but the retention rate completely depends on your product’s quality.

How to increase the retention rate? Three ways to go about it — talking to users, making decisions based on data and iterating your product based on user's requirements. And this is why YC has always told startups to do things that don’t scale — talk to users and understand pain points. It’s a complete waste of money to pay ads, expand the team, and increase server numbers when the retention rate is low because your users will leave due to faulty products.

Product Market Fit

We now know that we need to focus on doing things that don’t scale and iterate a product until it has a high retention rate to reach stable growth. Then when should you scale? The answer is after finding product market fit(PMF). The definition of PMF — is you would be making a product that users actually need in a market.

To help companies understand how far off they are from PMF, YC analysed 3000+ portfolio companies they have invested in over 20 years and defined a quantifiable metric for almost every different type of company:

Paid monthly user retention curve: how many paid users are still paying after several months? What’s the proportion?

To reach PMF for a product focusing on Consumer Subscription like Heptabase, the proportion of paid user monthly retention needs to reach more than 50%. If it is a free consumer product like Facebook, the daily active user monthly retention rate needs to reach more than 40%. For B2B SaaS products facing SMB, Mid-Market or Enterprise, the monthly retention needs to reach more than 60%, 70% and 80% respectively.

After a company finds its PMF, its daily work will be shifted — you don’t need to look for new users because there will be a ton of new users looking for you, and you won’t have enough time to make new features but focus on keeping the product up and running, hire more employees and purchasing service to cater large demand. You have to reverse what you have been doing during the stage of ‘Do Things That Don’t Scale’ but start to hire a lot of employees, customer support, sales team, and servers and do anything to reach a broader range of users. And it would be about time to raise Series A.

But of course, many companies being accepted to YC are still pre-PMF — have an MVP but haven’t found PMF. The founding team needs to keep the team lean, control the burn rate and focus on iterating products and talking to users to increase user retention rate and let primary metrics have stable and sustainable growth.

Insane target on Demo Day

Spotlight Photo by Alex Litvin on Unsplash

To conclude what I have learnt in YC about startups in a simple way — The priority of a startup is to ensure:

  • Primary Metric(Revenue, Active User) has stable growth by optimising secondary metrics.
  • For a consumer product, the retention rate is the most important secondary metric.
  • We can initially determine if a company found its PMF by its retention rate from each cohort/period.

After understanding these concepts, I will share more about Heptabase’s experience in YC.

At the beginning of each batch, an unreachable target for demo day will be set for every YC portfolio company by a YC Partner based on the company’s status. When we just got into YC, our Demo Day target was to reach 40 times of monthly recurring revenue compared to where we were, having stable growth and sustaining the team and operation based on our revenue without paying any advertisement.

After we knew our Demo Day target, the next step was to break them down into weekly targets. To grow MRR 40 times in three months means we need to grow 35% every week. In the Group Office Hour every week, we have to report to YC Partner with other portfolio companies whether we reached our weekly target or not and review why we reached it or why we didn’t. The peer pressure was mad. We didn’t want to see ourselves fail to reach our target while other companies did. It forced us to consider our weekly target and strategise it carefully.

Focusing on a weekly target might sound brutal and rough, but it helped us a lot because it made us more focused and decisive. Every decision we made about whether or not to build a certain feature, fix a bug, post any product-related information on a platform, change our pricing and hire a contractor would come back to the essence — does it help us to reach the weekly target? Many correct decisions are opposite to instinct, but we were able to make the right decision by focusing on the weekly targets, even though it seemed wrong to us at the beginning.

Subscription model

Of all the ‘Correct but sounds wrong’ decisions, the most notable one was our subscription model.

When I was just a note-taking user, I’d follow my instinct and consider a good product should be free for a long time in the beginning, so we can let more users try it and give us feedback. So before we entered YC, our subscription model was ‘ Free to use, pay to sponsor’.

But during my study at YC, I realised that when we are running a start-up, we need to filter out unnecessary information and differentiate core users that actually need our product and are willing to pay to satisfy their needs. A company is only sustainable and profitable when it can solve the core users’ needs.

When we release our free version, 9 out of 10 users I have served are free users. These free users have mixed backgrounds and needs with varying levels of intensity. Some people wanted to use Heptabase to sort documentation; some wanted to use it to write essays; some others wanted to use it for knowledge management, and some people just wanted to use it to draw flow charts. When our product was free, we couldn’t focus on who actually needed Heptabase to solve important problems; we had a lot of pressure on customer support, and a lot of the requests were exclusive to one another.

To maintain our quality in serving customers and focusing on the needs of core users, I decided to adjust our subscription model — we stopped free testing for new users and users can only use our product if they pay the annual subscription, with one week refund period. The new subscription model increased the entry barrier but yielded many unexpected and positive results.

Important Progress

After we adjusted our free subscription model to the annual subscription, we made three important progress.

Progress Photo by Jungwoo Hong on Unsplash

First of all, we had a better understanding of who has intense needs for Heptabase and what problems they needed to solve. The fee-paying users for Heptabase don’t limit to people who are enthusiastic about note-taking but also includes executives in companies, investors in venture capitals who need to allocate multi-million funds, start-up founders who need to set up product strategy, researchers who need to conduct research and write essays, online creators who are creating a lot of content and students who are preparing for legal, medical or accounting exams. They undertake a lot of responsibility in their career, thus they need to read and understand a lot of complex information, but Heptabase is the only tool that can help them on the market.

I have learnt that we don’t need to worry about charging fees early. The actual value of the product was determined by what problems it solves but not how many features it provides. When your house is on fire, the value of a fire extinguisher is determined by whether it can extinguish the fire or not, but not how many features it has. It is the same for products; the earlier you start to charge, the earlier you can find who actually needs the product, and start to understand the real landscape of a market.

Secondly, we had a lot of feedback from our paying users. The quality of feedback was a lot higher than when we were providing the product for free. And the feedbacks were not exclusive to one another anymore, so it helped us build a clear roadmap on what to build in each milestone. Some paying users wanted a refund after trying it because it wasn’t reaching their expectations. We were also able to analyse the biggest fault of our current product and iterate it as fast as possible by analysing the forms submitted by leaving users.

Thirdly, our first month’s revenue in YC was larger than our competitors, Logseq and Athen (Web-based Obsidian)combined in the last two years on Open Collective. We were able to break even in the second month; not only have we become sustainable, but also reached every weekly target in YC. During our time in YC, our product became much better, and we can finally focus on serving 1,000 users who actually need Heptabase and provide them with immediate, almost 24 hours customer support and tutorial within a three-person team, without facing 10,000 users who wanted to try but have no actual needs.

One simple change in the subscription model enabled us to iterate our product faster with better precision, increase our recurring revenue, and break even and all these results were far exceeding our expectations. It was a decision completely different from our instinct. If it wasn’t because the pressure from Demo Day, we wouldn’t have thought that we could make it.

Looking back, the benefit of revenue growth and breaking even is huge — I have more cards on the table during fundraising, and we were able to choose investors who understood us, created more synergy and aligned with our vision, but did not need to surrender to the ones unfit because I know that even without investment, we can keep operating the company and pay our team.


In the ninth week of our YC batch, we attended Fundraising Office Hour. Our YC Partner looked at our metrics and gave us a shocking comment,

“Your growth metrics are great! Fantastic! How did you do it?! I think you are in a very good state!”

In the tenth week, I went to the Pitch Practice in YC, and the YC Partner who listened to my pitch told me that we broke her impression of note-taking products. In all the note-taking companies YC has backed in previous years, the growth of Heptabase was the most outstanding and impressive.

To give it my reflection — we were set in an environment that forced us to focus on growth, only, and that’s the reason why we could reach an unimaginable growth rate which sounded clueless to us at the beginning. I have only realised the meaning behind the mindset of ‘Startup = Growth’ from Paul Graham in this environment.

‘Startup=Growth’ doesn’t mean that startups ‘should’ focus on growth, but should focus on growth ‘only’. Once you consider growth is your only target, your decision-making will become much clearer and aligned; your product will become better faster; you will start to understand the pain points clear; you will have more sustaining power in the commercial world; you will be able to survive in the high-risk startup world and have the opportunity to achieve your vision.

Ending notes by translator:

This article is translated by Zhao. To see the original article written in traditional Chinese by Alan Chan, you can see it here. Zhao skipped the part he introduced about Y Combinator as most of the information is available on the internet. Alan and Zhao have both agreed that an English version would be beneficial to a broader range of audiences. Alan founded Heptabase, and his company was accepted to YC in the 2022 Winter Batch. Although his articles inspired Zhao in deciding what to do in Zhao’s start-up, his opinion doesn’t represent Zhao or the related company’s opinion. The translation was done objectively.

This blog will also publish other translated articles in the Start-Up Note Series by Alan Chan.