⛏️SOLO淘金
SaaS/开发者工具B2B订阅+按量付费落地可行性

调低目标野心,4年做到2.5万美元+月经常性收入

Hitting $25k+ MRR by making his goals less ambitious

ScreenshotOne · 开发者截图API · 5个月上线 · 2.5万美元+月收入

收入规模
$25K+ / 月
团队规模
独立开发者
启动速度
5个月
复刻难度
★★★★☆
访谈亮点
  • 放弃大目标,从小步可实现的任务开始,快速获得正反馈
  • 🎯花2年时间明确真实目标用户,彻底改变全链路运营策略
  • 🚀持续迭代访客-注册-付费转化漏斗,靠复利效应实现收入稳步增长
  • 💡不盲目照搬他人战术,优先学习经济学、商业底层规律

10年经验工程师调低创业目标,从小处着手,4年将面向开发者的截图API做到2.5万+美元月收入,核心经验是找准目标用户、持续优化转化漏斗。

ScreenshotOne

更小的目标

Smaller goals

我在辞职前做了10年以上的软件工程师,拥有计算机科学学位,一直想成为创业者,追求工作的自主权。我之前不断尝试创业但次次失败,因为我的目标定得太大了。核心心态直到我降低野心,专注于用简单步骤完成小目标,才开始看到成果。现在我全身心投入ScreenshotOne,一款面向开发者和AI代理的截图API,我们拥有超过800名付费客户,月收入超过2.5万美收入里程碑元。

I had been a software engineer for 10+ years before I quit my job. I had a degree in computer science and all that. But I've always wanted to be an entrepreneur. I prefer to have autonomy and independence — especially about the areas I want to work in. So, I kept trying to start something, but failed each time. My goals were to big. I only started seeing results when I downgraded my ambition and focused on achieving small things in easy steps. Now, I am working on ScreenshotOne, a screenshot API for developers and AI agents. It is my passion and my only focus. We have over 800 paying customers, and we're making over $25k MRR.

?

Who is it for?

  • 有后端开发经验的独立开发者
  • 面向B端开发者工具创业者

Not for

  • 缺乏技术积累的非开发者
  • 追求短期暴富的投机者

从极简到复杂的技术栈

From simple to exploded stack

启动开发2022年1月到5月我花了5个月时间开发它,当时还没有AI编程助手,过程非常艰难,我中途还多次休息。最初的技术栈非常简单:全栈用Go开发,用JavaScript管理浏览器。后来技术栈逐步扩展:用Next.js做控制台,Astro做营销官网,技术栈组合Go做限流和API密钥管理,TypeScript管理浏览器,Cloudflare做缓存、API网关和存储,Kubernetes托管浏览器服务,还有很多其他组件。

I spent five months building it, from January to May 2022. There weren't any AI coding agents back then. Plus, I took breaks because it was really, really hard. The stack started as one simple server, Go for everything, and JavaScript for managing browsers in the beginning. But later, my stack exploded. Now it is: Next.js for our dashboard, Astro for my marketing website, Go for managing rate limits and API keys, TypeScript for managing browsers, Cloudflare for cache, API gateway, and storage, Kubernetes for hosting browsers, and plenty more...

⚙️

技术选型思路

初期不要追求大而全的技术栈,用最极简的组合快速上线验证需求,后续随着用户规模增长再逐步迭代扩展组件,避免前期过度开发浪费精力。

了解你的市场

Know your market

最开始我到处发帖推广产品,也投放了付费广告,真正起作用的渠道是Reddit,但我认为这个渠核心渠道道并不适合所有人。你需要大量做实验,把资源集中在真正有效的渠道上。除了找到合适的推广渠道,最能改变增长轨迹的事就是搞清楚你到底在把产品卖给谁,我花了两年时间才想明白这件事,它彻底改变了我的营销方式、内容方向、产品设计和定价策略。如果让我重新创业,我要市场认知么做自己就是目标用户的产品,要么投入大量时间搞清楚我的用户到底是谁。

In the beginning, I grew by posting about my product everywhere. I also did paid ads. The channel that really started working for me was Reddit. But I don't think that would work for everybody. You need to experiment a lot and double down on what works. Outside of finding the right channels, the thing that changed my growth trajectory most was realizing who I was selling to. I know it might sound dumb, but it took me two years to realize that. And it changed everything. How I market things, what I write, what I build, pricing — everything. If I started over, I would build for myself. Or I would really, really invest in knowing who I build my application for.

📈

增长策略

不要迷信单一爆款渠道,多尝试不同平台,找到你的目标用户天然聚集、主动讨论相关问题的场景,集中资源投入放大效果。

优化收入

Tweaking revenue

商业模式我采用订阅制模式,超出请求额度的部分按量计费。我建议大家部署PostHog这类产品分析工具,追踪访客到注册用户再到付费客户的全链路转化数据,以及各个转化流量的来源。之后针对性优化各个环节:没有访客就做引流,有访客不注册就优化注册流程,听起来简单做起来难,持续调优转化漏斗是我收入增长的核心原因。

I use a subscription model, and I charge for extra requests — think "pay as you go." I recommend setting up product analytics like PostHog, and tracking how many people convert from visitors to users to paying customers — and from what sources. Then, try to change things to grow these numbers. Don't have visitors? Work on attracting them. Have visitors, but they don't sign up? Improve your signup flow. It sounds simple, but it is not. Tuning the funnel consistently is what helped me grow my revenue.

💰

收入优化思路

混合订阅+按量付费模式,既保证稳定的基础收入,又能覆盖高用量客户的成本,通过全链路埋点追踪转化漏斗,小步迭代优化每个环节的转化率,复利效应远大于单次大的营销动作。

别听我的建议

Ignore my advice

避坑提醒不要盲目听从任何建议,包括我自己的,大多数人都没法从自己的经历里提炼出可复用的经验,就算提炼出来了通常也已经过时了。多去读经济学、商业、历史、生物学的经典著作,掌握底层规律。

Don't listen to any advice. Most people — myself included — don't know how to extract lessons and advice from their experiences. And even if they extract it, it is usually outdated. Read the classics on economics, business, history, and biology. Fundamentals.

🧠

创业心态

不要直接照搬他人的战术经验,大部分成功经验都是特定时间、特定场景下的产物,可迁移性极低,优先学习底层规律才能适配自己的实际情况。

下一步计划

What's next?

接下来我要把现有产品的自动化程度做高,这样我就可以去徒步一周全程不用碰网络。同时我的目标是把月收入做到5万美元。你可以去我的个人网站和 ne未来目标wsletter 跟进我的动态,我在X平台也很活跃,欢迎大家体验ScreenshotOne。

From here, I want to automate my current product so that I can go on a hiking trip for a week without access to the internet. I also want to grow to $50k MRR. You can follow along on my personal website and newsletter. I am active on X too. And check out ScreenshotOne!

📅

全周期里程碑

  • {"year":"2022年1月","event":"启动ScreenshotOne开发,初始技术栈仅用Go和JavaScript"}
  • {"year":"2022年5月","event":"耗时5个月完成产品初版开发并上线"}
  • {"year":"上线后2年","event":"明确目标用户画像,调整全链路运营策略,增长速度大幅提升"}
  • {"year":"上线后4年","event":"产品达到25000+ MRR,累计拥有800+付费客户"}
  • {"year":"未来规划","event":"目标实现50000 MRR,完成产品自动化运维,支持离线徒步一周无需值守"}

用户评论与创作者回复

已过滤 spam、低价值附和与重复内容 · 原始 83 条 · 展示 11 条 · 创作者回复 1

11
Rai Ansar
👍 2
经验复用

Awesome, I learned a lot from this post. Because my product literally uses what you are offering but I am doing that in-house not off-loading to third party... So technically what VisualSentinel does is takes screenshots and compare with the baseline to find out if the UI of the website changed or DNS updated or site is down. Also insight regarding reddit was quite helpful I may utilize that in the marketing part of my product.

这篇文章让我收获很多,我的产品现在自研了截图能力,没有接入第三方API,主打网站UI变化、DNS更新、站点可用性监控。Reddit的推广经验非常实用,我打算用到自己产品的营销环节里。

笔记:来自同类赛道创业者的真实反馈,验证了截图API是大量B端产品的通用刚需组件。

Rafa López
👍 2
追问渠道

The "smaller goals" approach is exactly what we're applying at InnerSpark. We're building a modular OS for founders (big vision) but launching tiny bets first - cashflow modules like DomainFlow and CashOS before the full Idea Engine. Question: Did Reddit work because you were solving a developer problem that developers congregate around? Curious if channel-market fit was obvious early or discovered through trial.

我们正在InnerSpark实践小目标策略,做面向创业者的模块化操作系统,先推出DomainFlow、CashOS这类能快速产生现金流的小模块,再做完整的Idea Engine。想问Reddit能起作用是不是因为开发者群体天然聚集在那里?渠道市场匹配是早期就明确还是试错出来的?

笔记:点出了独立创业者做大愿景产品的正确落地路径:先做小模块快速获得正向现金流,再逐步迭代完整产品。

introlo
👍 2
观点赞同

This is pretty underrated. Most founders try to build the billion-dollar thing first. Gotta keep in mind that small, boring wins can win.

这个思路被严重低估了,大部分创始人一开始就想做估值十亿的产品,要记住那些不起眼的小胜利积累起来最终能带来巨大成功。

笔记:戳中了很多新手创业者好高骛远的通病,小步迭代的复利效应远好于一开始就追求宏大叙事。

Benjamin Martin
👍 2
深度补充

The "smaller goals" reframe is the most honest advice in here, and also the hardest to internalize because it feels like giving up when you do it, and only looks like wisdom in hindsight. Most tactical advice is context-dependent and the context is never fully transferable. The classics on economics and incentives hold up because they're describing human behaviour at a level of abstraction that survives context change.

调低目标的思路是最实在的建议,但也最难内化,因为做的时候会感觉像是在放弃,事后回头看才发现是智慧。大部分战术建议都高度依赖特定场景,不可直接复制,经济学和激励相关的经典理论才是跨场景通用的底层规律。

笔记:进一步解释了创始人建议不可直接照搬的核心原因,帮读者区分可迁移的底层规律和不可复用的战术经验。

Sophia Devy
👍 2
经验共鸣

This really resonates with me. I’ve noticed that my biggest breakthroughs usually came after I lowered the ambition and focused on smaller, achievable steps I could ship consistently because momentum beats “perfect plans.” The way Dmytro narrowed in on a clear market, then kept tuning the funnel is the kind of boring, repeatable work that actually compounds.

我非常有共鸣,我自己最大的突破都是在调低野心、专注于可以持续交付的小目标之后出现的,动量永远比完美计划重要。Dmytro找准市场之后持续优化漏斗的工作看起来枯燥重复,但恰恰是这些工作带来了复利增长。

笔记:用自身经历验证了小步快跑、持续优化的价值,给其他独立创业者提供了可参考的行动路径。

HermesIntel
👍 1
干货补充

What's underappreciated: ICP clarity is primarily a research problem, not a discovery problem. Real ICP precision usually comes from triangulating multiple signals: which customers have the lowest churn, which search terms bring the highest-LTV signups, which competitor's customers convert best when they find you.

很多人没意识到,明确理想客户画像本质是研究问题,不是靠碰运气发现的。精准的ICP需要交叉验证多个信号:哪些客户留存率最高、哪些搜索词带来的用户生命周期价值最高、哪些竞品的客户最容易转化到你的产品。

笔记:给想缩短用户画像明确周期的创业者提供了可落地的多信号交叉验证方法,不用花两年时间慢慢摸索。

Helena Jackson
👍 1
真实经历

I'm in Week 1 of building a developer API tool right now — pure validation mode, no code written yet — and I'm already seeing this play out in real time. The people responding to my validation posts split almost immediately into two distinct groups: API producers who want to know how well their own docs perform, and API consumers who want to evaluate third-party APIs they depend on but can't control. Same tool, completely different use cases, different messaging, probably different pricing.

我现在正在做一款开发者API工具,还没写一行代码,处于纯验证阶段,已经发现我的潜在用户分成了完全不同的两类:API生产者想测试自己的文档表现,API消费者想评估自己依赖的第三方API的稳定性。同一个工具,两类用户的使用场景、沟通话术、定价策略完全不一样。

笔记:用自己的真实验证经历说明,在产品上线前就做用户分层,能避免后期走很多弯路。

miadevelops
👍 1
经验分享

The "ignore my advice" section is honestly the best part. Most founder stories end with a neat list of lessons that sound great but don't transfer to other contexts. The fact that Dmytro calls that out is refreshing. It's not about lowering your ceiling, it's about finding a starting point you can actually reach.

"别听我的建议"这部分是整篇文章最有价值的内容,大部分创始人故事最后都会列一堆听起来很美好但根本没法跨场景复用的经验,Dmytro能点出这一点非常实在。调低目标不是降低天花板,而是找到一个你真的能触达的起点。

笔记:纠正了很多人对"调低野心"的误解,不是放弃做大做强的目标,而是找到一个可落地的起步点。

voxboard
👍 1
实用框架

The funnel improvement framework is super practical too: - Not enough visitors → fix traffic - Visitors but no signups → fix the signup flow - Signups but no payments → fix onboarding. I just launched a SaaS on Product Hunt today and I'm bookmarking this as my playbook for the next 3 months. Focus on funnel optimization, not new features.

漏斗优化框架太实用了:访客少就做引流,有访客不注册就优化注册流程,注册了不付费就优化新手引导。我今天刚在Product Hunt上线了自己的SaaS产品,接下来三个月就照着这个框架执行,专注优化漏斗,不做新功能。

笔记:把创始人的经验直接落地成了可执行的3个月行动指南,给其他新上线的SaaS创业者提供了明确的优先级参考。

RovaAI
👍 1
赛道洞察

The "who you're selling to" realization is the most expensive lesson in API products — and two years is actually fast compared to most. What makes it harder for API tools than regular SaaS: your users rarely tell you what they're building with your product. A screenshot API gets used by agencies automating client reports, startups building monitoring dashboards, solo devs scraping comparison data, and enterprises pulling screenshots for compliance. They all look identical in your logs.

明确目标用户是API产品最昂贵的一课,创始人花了两年时间做到其实已经比大部分API创业者快了。API工具比普通SaaS更难明确用户画像,因为用户几乎不会告诉你他们用你的产品做什么,截图API的用户可能是自动化客户报告的代理、做监控仪表盘的创业公司、爬取对比数据的独立开发者、做合规审计的企业,这些人在日志里看起来完全没有区别。

笔记:点出了API类产品做用户画像的特殊难点,给同类赛道创业者提供了提前避坑的参考。

Erik Mostert
👍 1
深度共鸣

The ambition doesn't go away — it just gets channelled into something you can actually finish. Automating yourself out of the day-to-day so you can take that hiking trip without anxiety sounds like the real milestone to me.

野心并没有消失,只是被引导到了你真的能完成的事情上。把产品自动化到你可以安心出去徒步一周不用处理工作,这才是真正值得追求的里程碑。

笔记:点出了独立创业者最向往的终极状态:产品完全自动化运行,不用被日常运维绑定,获得真正的时间自主权。

Localization Notes

  • 国内截图API赛道已有成熟玩家,可差异化面向AI代理、自动化UI测试、电商页面监控等垂直细分场景切入,避开同质化竞争
  • 替换海外Cloudflare、Kubernetes等云组件为阿里云、腾讯云本土云服务,大幅降低合规成本和运维难度
  • 适配微信支付、支付宝等本土支付渠道,降低国内中小开发者的付费门槛,提升转化率
  • 掘金、V2EX、GitHub是替代Reddit的国内开发者核心聚集社区,是冷启动阶段获取精准种子用户的首选渠道
  • 可在基础截图能力之上增加OCR识别、AI内容提取、网页智能分析等增值功能,提升产品溢价空间和用户留存率