How To Start A Dev Tools Company | Startup School

Intro 0s

• The speaker's name is Nicola D, and he is a group partner at YC, previously the co-founder and CEO of Algolia, a search API used by millions of developers to build a great search experience in their apps and websites 9s.
• The topic of discussion is about dev tools and how to start a dev tools company, covering funding, team, finding an idea, starting the company from prototype to MVP, and going to market including sales and marketing 34s.
• A dev tool is software used by developers to help them build products, including aspects such as coding, testing, debugging, documenting, deploying, and running 51s.
• Dev tools is a broad category that includes IDEs like VSCode, APIs like Stripe or Algolia, libraries and frameworks like React, and infrastructure or cloud services like AWS 1m12s.
• YC has supported many dev tools companies, including public companies like GitLab, PagerDuty, and others like Stripe, Docker, Hoku, Segment, and Apollo 1m37s.
• The discussion will cover the founding stage of a dev tools company, starting with the importance of a founding team 2m7s.
• The key aspect of a founding team will be discussed, although the details of this aspect are not provided in the intro 2m13s.

Founding Stage 2m15s

• Building a dev tool company requires creating a technical product for developers, and it's beneficial if the founding team consists of developers themselves, as they can relate to the needs of their target audience 2m18s.
• The founding team can identify good and bad dev tool ideas by considering the current market and the needs of developers, keeping in mind that the advent of LLMs and AI has changed the landscape of what can be successful 2m47s.
• Ideas that were previously considered bad can now be successful due to LLMs, but some ideas are more difficult than others, and developers often focus on solving problems they face daily, such as documentation, QA, and testing 3m3s.
• However, these ideas can be crowded, and it's essential to consider whether building another QA tool, for example, is a viable option 3m27s.
• A different approach is to focus on the difference between build-time and runtime ideas, with runtime ideas being more critical and must-have products, such as APIs, which can be more successful for dev tools 3m54s.
• Runtime ideas also offer better usage and alignment of incentives, as customers' growth leads to increased usage of the product, making it a more attractive option for dev tools 4m17s.
• Another category of dev tools is libraries and frameworks, which can be great but challenging to monetize, especially if they are open-source, and one possible way to monetize them is by offering a hosting service 4m37s.
• Lastly, the trend of LLMs and AI can be leveraged to create successful dev tools, but it's essential to consider the current market and the needs of developers when evaluating ideas 5m7s.

AI Trend 5m10s

• The current time is exciting for creating new dev tools, especially those that help companies building NLMs, but the market is also very crowded, with many companies having similar ideas, such as LLM observability 5m10s.
• Having a clear idea of how to differentiate oneself is crucial when starting in a crowded space, as dozens or hundreds of companies may be working on similar ideas 5m41s.
• Waiting for the perfect idea can lead to a long wait, and it's better to start working on an idea, even if it's not fully differentiated, and learn as you go 5m58s.
• Sticking with the wrong idea for too long is a common mistake, and it's worth noting that 50% of YC companies eventually pivot from their first idea 6m21s.
• Having a business founder in the team is not necessary, as technical co-founders can learn to sell their product, and 74% of YC dev tool companies had only technical co-founders 6m39s.
• The founders of Algolia, a dev tool company, started with two technical co-founders and learned to sell their product without needing a business co-founder 6m42s.

Where to start? 7m20s

• To start a dev tools company, two essential steps are building and talking with users, and there is no specific order for these steps, as talking to users can help determine if the initial idea is good, and building a prototype can facilitate easier feedback from users 7m20s.
• When building a prototype, it's crucial not to over-engineer it, and instead, focus on creating a quick and dirty version to iterate as fast as possible and gather feedback 7m37s.
• The primary goal at this stage is to identify the valuable 10% of the code being built and discard the rest, as most of the initial code will likely be thrown away 8m1s.
• It's essential to show users a prototype and gather feedback as early as possible, rather than waiting for a perfect product, as this will eventually lead to a minimal viable product (MVP) that provides value to customers 8m29s.
• A successful MVP should provide significant value to a specific niche or group of customers, and it's better to be 10x better at a tiny thing for someone who cares about it, rather than trying to cater to a broad audience 8m45s.
• Expanding from a niche is easier once real customers who love the product are acquired, as demonstrated by the example of Algolia, which started as a minimal autocomplete feature but eventually grew into a full-featured search engine 9m4s.
• Even a minimal product can be enough to attract customers, as seen in Algolia's early days, where a command-line demo and a simple web page were enough to close a $2,000 a month contract 9m34s.
• After building something to show users, the next step is to talk with users and gather feedback to further develop the product 9m55s.

Talking to users 10m0s

• As a developer building a dev tool, you have a huge advantage in understanding your audience since you are a developer yourself and speak the language of your customers, making it easier to talk to users or customers, despite being an introvert 10m2s.
• It's essential to learn from users as soon as possible, rather than waiting for the product to be ready, as solving a real problem will make people willing to pay for it 10m27s.
• To find users, two strategies can be employed: outreach and launches, with outreach being crucial in the early stages since nobody knows the product yet 10m36s.
• A bottom-up adoption model, where individual developers come inbound and try the product, is often planned, but outreach is necessary to get people to know about the product first 10m42s.
• When outreaching, start with your network, including previous colleagues or classmates, and expand from there, leveraging LinkedIn to find specific people, but personalize messages to avoid being seen as a marketer 11m13s.
• Personalized messages should be crafted by asking yourself if you would be excited to open the message, and iterating until it's effective, with feedback from co-founders and developer friends 11m36s.
• Launching is another strategy, and it's possible to launch multiple times to get more feedback and users, which is a recommended approach 11m54s.

Launches 12m0s

• The best place to launch dev tools is Hacker News, a community of intellectually curious people, many of whom are developers, where there is a section called "Show HN" that's perfect for launching and gathering feedback on what you're building 12m2s.
• When launching, don't do marketing or try to sell your product, just explain plainly and simply what's new and interesting in what you're building, which will be engaging for the community and encourage them to share comments 12m28s.
• Engage with comments, including those from haters, but don't try to convince them, and instead focus on convincing other readers who will read how you reply to these comments 12m52s.
• Examples of successful launches on Hacker News include Segment, which launched their next idea and got hundreds of votes and comments, and AMA Olama, which started as a comment on another post and then launched their product to great success 13m3s.
• Launching repeatedly, every few months, can continue to fuel growth and create excitement in the community, as long as you have new things to share 13m44s.
• Doing things that don't scale, such as Stripe's early team going to customers and helping them implement their product side by side, can be a great way to learn from users and gather feedback 14m3s.
• Mistakes to avoid include choosing a tech stack because it's cool, not talking with users, overbuilding before getting feedback, misunderstanding developer feedback, and hiring too early 14m53s.
• When getting feedback, be careful not to misunderstand it, and focus on whether users would use your product and why, rather than how they think you should build it 15m24s.
• Don't hire anyone until you have convinced yourself that you're working on the right idea and your product is providing value 15m58s.

GTM 16m15s

• Being open source has become a primary go-to-market strategy for dev tools and can be a successful business model, as seen in companies like DataBricks, Elastic, HashiCorp, and GitLab 16m16s.
• It is essential to be open source when working on libraries, frameworks, or dealing with sensitive data, as developers expect this and are unlikely to implement new frameworks or databases that are not open source 16m30s.
• Being open source may not be as crucial when working on APIs or applications, but it can still be a differentiator 17m1s.
• Open source has several benefits, including being preferred by developers, creating awareness through the community, and serving as a differentiator 17m12s.
• Open source projects can receive contributions from the community, but it's essential to be cautious and not rely on these contributions due to potential quality issues 17m37s.
• Good contributions from the community are rare, and dealing with contributors can be challenging 18m3s.
• Being open source can establish trust, which is particularly important for large enterprises, and can potentially shorten sales cycles 18m10s.
• When creating an open source project, it's crucial to consider monetization strategies, as a company cannot sustain itself without a viable monetization plan 18m38s.

Monetization 18m45s

• To monetize a dev tools company, it's essential to earn money, and there are various approaches to consider, including hosting, open core, and charging for support and services 18m46s.
• Hosting involves offering a cloud version of the product, which may include additional features like Team Management, and customers can pay for the convenience of not having to maintain it themselves 18m55s.
• The open core approach typically involves offering an Enterprise version with advanced features not available in the open source version, such as SSO, Disaster Recovery, and SLAs 19m12s.
• Charging for support and services is another approach, but it can create a bad incentive to build complex products, and customers may not renew their contracts if the product doesn't require much support 19m35s.
• For non-open source dev tools, common monetization approaches include usage-based pricing, like APIs, with possible volume discounts or Enterprise-specific options 20m2s.
• Another approach is offering specific plans, such as a good-better-best model, where the "good" plan is a self-serve option for Developers, the "better" plan is for engineering managers who care about collaboration, and the "best" plan is for CTOs who prioritize security, audit logs, and SLAs 20m21s.
• The "best" plan often requires a sales-led approach, and it's essential to not fear outreach, even if most dev tools rely heavily on inbound marketing 20m54s.

Selling 21m10s

• To start selling a dev tool company, begin with founder outreach, as it is the most effective way to get the company's name out there, and eventually, inbound leads will increase, allowing for the hiring of a sales team to sell to bigger enterprises 21m11s.
• It is recommended to wait as long as possible before hiring a sales team, with a suggested revenue milestone of about $1 million ARR, and to hire people who are technical or understand developers 21m55s.
• Technical salespeople are more effective in selling dev tools, as they can understand the product and communicate effectively with developers, and having a title like "product specialist" can be beneficial 22m17s.
• PostHog's sales leader is also their CTO, and they view sales as a nurturing problem, which is an effective approach for a dev tool company 22m46s.
• Selling a dev tool is different from selling other products, and it is essential to have a sales deck that shows, rather than tells, and to focus on demonstrations 23m8s.
• Leaning into the tech aspect of the product is crucial, as technical buyers are more likely to make a purchase, and finding the ideal customer profile (ICP) is essential for success 23m43s.
• Most dev tools are bottom-up, meaning that many users are self-serve, even within enterprises, and sales should focus on helping management understand the product's value and potential 24m22s.
• Developer marketing aims to create awareness for the product, leading to inbound leads, and is an essential aspect of selling a dev tool company 24m59s.

Marketing 25m10s

• To start marketing a dev tools company, find the community related to the product, such as Hacker News, subreddits, or Discord, and establish oneself as an expert by being helpful, without directly selling the product, to build trust and recognition for the brand and company 25m11s.
• Participating in AMAs (Ask Me Anything) sessions and launch weeks can be effective marketing strategies, as seen in the example of Super Bas, which has a quarterly launch week to make a significant impact 25m41s.
• Making documentation a first-class citizen is crucial, as it is often the first point of interaction with the product after the homepage and pricing page, and should be written by developers, with the goal of making it a valuable resource for customers 26m6s.
• At Algolia, the approach was to consider a feature complete only when the documentation was finished, and a documentation team was created to assist developers in writing better documentation, rather than writing it for them 26m44s.
• The documentation team at Algolia built tools to make the documentation more useful, such as automatically inserting API keys for logged-in users, to make the code work out of the box 27m5s.
• Providing high-quality support is essential, as developers hate speaking to non-technical support staff, and support should be a significant part of the marketing strategy, recognizing that users are developers 27m25s.

Support 27m35s

• Engineers should do support, as it leads to two great things: users are more satisfied when interacting with developers who speak their language, and it helps developers better understand customers' needs 27m38s.
• This approach can result in impressive customer experiences, such as fixing bugs and deploying in production quickly 27m56s.
• Developers doing support can gain a deeper understanding of customers' pain points, enabling them to create better products 28m18s.
• Aloya waited until they had hundreds of employees before building a dedicated support team, and they asked their best engineering manager to lead it 28m26s.
• Support should not be a second-class citizen, and developers should be involved in documentation and support 28m41s.
• Founders should lead marketing for a long time, as traditional marketers may not understand the developer persona 28m52s.
• Traditional marketers may not be effective in marketing to developers, as developers hate being marketed to and prefer authenticity 29m11s.
• At Aloya, every engineer was asked to do a "marketing hack" every month, such as writing a blog post or speaking at a meetup 29m28s.
• Developers should be involved in marketing teams, as they know how to speak to customers and create effective marketing strategies 29m50s.
• Dev advocates can be effective, but the role can be ill-defined, making it difficult to measure and build accountability 30m0s.
• It's recommended to wait and leverage existing engineering teams for marketing and Dev advocacy before hiring dedicated Dev Advocates 30m18s.
• When hiring Dev Advocates, consider hiring from within or from the community, such as open-source contributors or active Discord members 30m35s.

Outro 31m0s

• To start a successful dev tools company, it's essential to begin with an imperfect idea, learn from it, and iterate to reach the best idea, as not starting at all will lead to no progress 31m0s.
• Building quickly is crucial, and it's recommended not to over-engineer, instead focusing on a fast iteration pace to learn from users while building 31m8s.
• Having a "crappy, quick, and dirty" approach is acceptable in the early stages, and then refactoring what matters is key 31m19s.
• Spending time with users is vital to learn from them, and doing things that don't scale, such as visiting their offices, can be beneficial 31m27s.
• Launching early is recommended, as it allows for invaluable feedback, and people are unlikely to remember if it doesn't work out 31m37s.
• When building a dev tool, considering open source as a market is essential, and observing the market and other players can help determine if open source is the right approach 31m53s.
• As the founder of a dev tools company, you are the best salesperson for your product, and no one else knows your product or audience as well as you do 32m10s.
• Learning to sell is easier than learning to speak with a developer, and you are also the best marketing person for your product, knowing the language of developers and how to access your target audience 32m17s.
• If building a dev tool, it's possible to apply to YC (Y Combinator) anytime 32m42s.