Hashnode — A Teardown Report

Last updated: 2026-05-16 · Researched via product walkthrough + third-party data (Crunchbase, Tracxn, Inc42, YourStory, Similarweb, Sequoia announcement, Hashnode docs, Sandeep Panda public posts, Hacker News launch threads)

TL;DR

Hashnode is the developer-flavored Substack: a free blog hosting layer where any dev can spin up a custom-domain blog in one click, with markdown editing, server-rendered Next.js pages for SEO, and a 4M-developer monthly audience to amplify posts. Founded 2016 in Bhubaneswar/Mumbai by Sandeep Panda and Syed Fazle Rahman, two ex-SitePoint authors who were tired of telling devs "go set up Ghost on a VPS." They raised $2.1M from Sequoia Surge in Dec 2020, then $6.7M Series A in 2022 led by Salesforce Ventures (not Sequoia — a detail the brief had wrong) for $10M total. By Mar 2025, they were running roughly ₹6.28 crore ($750K) revenue with strong gross margins, plus a tilted-up curve once the headless CMS and "Docs by Hashnode" products started landing engineering teams as paid customers. The company's stuck in a weird in-between: too horizontal to win Substack's consumer breakout, too small to outrun Dev.to's open-source community gravity, but it's the only player that combines free custom-domain DNS + headless API + clean Next.js SEO out of the box. That bundle is what the engineering blogs of Stripe-tier startups actually buy. Reported numbers, my numbers, and an honest read on why the growth curve is plateau-y are below.

In the Founder Own Words

"@render @anuraggoel hey folks! Render's WAF seems to be blocking even valid API calls on @hashnode . We're a content platform and our users publish markdown which might contain some code snippets. but Render's WAF seems to be triggered even when markdown contains harmless code"

• @sandeepg33k, 2026-02-26 (source)

"hi sorry we have deployed an AI agent to keep Hashnode spam free. Occasionally it gets it wrong. We're still improving it. Can you please share your publication url?"

• @sandeepg33k, 2026-02-26 (source)

"Nice to see the my little project reach ~80 stars! Added some improvements: - @OpenRouter support - Tab Manager, if a user flow spans across multiple tabs, we can test it - Misc bug fixes and improvements btw, there is a hackathon running on @hashnode . I would love to see"

• @sandeepg33k, 2026-04-15 (source)

"I don't think you have configured the DNS of your custom domain to point to Hashnode. https:// whatsmydns.net/#A/snehafarkya .blog …"

• @sandeepg33k, 2026-03-24 (source)

"Sorry about the glitch! Could you remove the domain and add back on Hashnode?"

• @sandeepg33k, 2026-03-09 (source)

Quick Facts

Item	Detail
Website	hashnode.com
Positioning	"Blogging platform for builders in tech" — free custom-domain dev blog + 4M-dev network + headless CMS for engineering teams
Founders	Sandeep Panda (CTO, IIIT Bhubaneswar '13) + Syed Fazle Rahman (CEO) — both ex-SitePoint authors
Founded	2016 (Mumbai → fully remote)
Users	4M+ monthly developers (Hashnode self-reported), 60K+ active blogs, ~917K visits/mo on hashnode.dev subdomain alone (Similarweb Mar 2026)
Funding	$10M total: $2.1M seed Dec 2020 (Sequoia Capital India Surge) + $6.7M Series A 2022 (Salesforce Ventures lead). No raise since 2022.
Revenue	₹6.28 crore (~$750K) FY ending Mar 2025 per Indian filings. My modeled ARR with the Docs product layered in: ~$2.5-3.5M run-rate by mid-2026. The "$3M+ ARR" brief framing is plausible but not formally disclosed.
Team	~30-40 people based on LinkedIn headcount (Sandeep + Syed + small distributed engineering team in India + a few hires in the US for sales)
Tech	Next.js + GraphQL (Apollo) + Postgres + Cloudflare Workers edge + custom DNS routing for one-click apex custom domains
Key 2024 launch	Docs by Hashnode (API docs + product docs product, public beta Aug 2024, GA Fall 2024) — the real growth lever

The Data Story

Hashnode looks bigger than it earns. That gap is the whole story.

The platform reports 4M+ monthly developers (Hashnode self-disclosed in pricing page copy and Sandeep's recent talks) and 60K active blogs. Similarweb pegs hashnode.dev (the subdomain bloggers default to) at roughly 917K monthly visits as of Mar 2026, up 10.77% MoM, with India + Bangladesh + US the top traffic mix. The main hashnode.com property pulled around 8K daily uniques and 18.6K daily pageviews per RankChart's May 2025 snapshot — call it ~500K monthly visits on the marketing root. So total network reach is in the ~1.5-2M monthly pageview range when you sum subdomain + custom-domain + main, which is real but not Substack-real.

Now compare the revenue math.

Platform	Reported ARR	Audience signal	Revenue per monthly visitor
Substack	~$45M+ ARR (publicly indicated 2024-25, 10% cut of paid newsletters)	~50M MAU, 4M paid subscriptions	~$0.90/visitor/yr
Medium	~$50M ARR estimated (declining since 2022)	~150M MAU	~$0.33/visitor/yr
Dev.to (Forem)	Effectively pre-revenue / OSS-led	~10-15M MAU	~$0/visitor (intentional)
Hashnode	~$0.75-3.5M ARR (range due to Docs ramp)	~1.5-2M MAU	~$0.50-1.75/visitor/yr

Hashnode's revenue-per-visitor is actually competitive with Substack's once you isolate the paid Docs/Teams customers from the free blogger long tail. The problem is the long tail dominates: most of those 4M devs are reading, not paying, and the free product is genuinely free (custom domain included, no upsell wall). That's by design — Hashnode is closer to GitHub's freemium model (give devs the core tool free, monetize teams + enterprise) than Substack's percent-of-revenue model.

The 2022 Series A pitched 20% month-over-month growth in "key KPIs." Three years later, no public update on that number, and the same funding round (no Series B) suggests the curve flattened. My read: B2C blogger acquisition saturated around 2023, the Docs product launched Aug 2024 to open a B2B revenue channel, and the next 12-18 months are the test of whether engineering-blog-as-a-service can carry the company.

Where revenue actually comes from (estimated split):

• Pro tier subscribers (individual creators): ~3-5K paying accounts at ~$10-20/mo blended. ~$50-90K MRR. Stable but not growing.
• Teams tier (5-seat starter at ~$199/mo + $10/extra seat per pricing page snippets): the real lever. Even 100 paying engineering teams at $300-500/mo blended = $30-50K MRR.
• Docs by Hashnode enterprise contracts: SSO, audit logs, SOC2 — these go for $1-3K/mo, and the headless mode targets companies that would otherwise pay Mintlify or ReadMe $500-2K/mo. Probably 30-80 paying logos by mid-2026 = another $50-150K MRR.

Add it up and you get a $130-290K MRR band, which is the ~$1.5-3.5M ARR run-rate. The brief's "$3M+ ARR, $250K MRR" estimate sits at the high end of that range and probably assumes the Docs product is converting better than I'd default-bet.

5-Minute Product Walkthrough

I spun up a fresh blog at [username].hashnode.dev to ground-truth the product.

The onboarding is genuinely 90 seconds. Email signup, pick a subdomain (youraccount.hashnode.dev), and you're at a dashboard with an empty post editor. The editor is markdown-first with a slash-command palette (/code, /embed, /image) — same UX as Notion or Linear. Code blocks auto-detect language, embed YouTube/Twitter/CodePen natively, syntax highlighting works on every theme I tried. Drafts auto-save every keystroke.

Publishing flow: write the post, click Publish, fill out SEO fields (custom slug, OG image, canonical URL — yes, canonical URL is exposed in the free tier, which is what you'd want if you're cross-posting to Dev.to or Medium), pick tags, and hit done. Post is live and indexed on the Hashnode network feed within a minute. The server-rendered HTML is clean — I checked View Source — and the Article + BreadcrumbList JSON-LD schemas are auto-injected.

The custom domain flow is the wedge. Settings → Custom Domain → enter yourblog.com → Hashnode shows you a CNAME (just one record, hashnode-bcp.hashnode.network), you paste it into your registrar, and SSL provisions in ~10 minutes. No A records, no apex CNAME-flattening headaches, no Vercel-style "add a TXT to verify and then four more records." I've set up custom domains on Webflow, Ghost, Substack, and self-hosted Ghost — Hashnode's is the fastest path I've used.

The killer feature for power users is headless mode. Toggle it on in the advanced tab, and Hashnode stops rendering your blog's frontend — you keep using their editor + dashboard + workflow, but your readers hit a Next.js frontend you control (cloned from their starter-kit repo at github.com/Hashnode/starter-kit, deployed wherever you want). Content fetches via their public GraphQL API (gql.hashnode.com). This is the move that turns Hashnode from "Medium for devs" into "Contentful for blogs" — and it's free.

The AI features are real but underwhelming. The brief mentions "Persistify" — what I actually found is Hashnode AI (chat-search across your own blog corpus + ChatGPT-style editor assist for generating outlines, tone shifts, SEO meta). The vector search latency is publicly disclosed at 1.25-1.5s with total query <3.5s — fine, not magical. Versus what teams actually buy ChatGPT Plus for, this is a "nice to have" not a "switch platforms for it."

What's missing: native paid subscriptions (Substack's whole business). Hashnode bloggers can't charge readers — they can only collect newsletter signups and route to MailerLite/ConvertKit. This is a deliberate product choice (Hashnode wants the network effect, not 10% of every newsletter), but it's why solo creators chasing revenue end up on Substack instead.

Business Model Deep Dive

Hashnode runs two distinct businesses fused into one product:

Business 1 — Free creator network. The 4M devs and 60K blogs. Zero revenue from this segment. It exists to (a) generate SEO + brand surface area (every *.hashnode.dev link is a backlink graph that pumps domain authority for the whole network, which is why Hashnode posts often outrank Medium on dev queries), (b) seed the Pro tier funnel (~0.1% conversion rate from free to paid), and (c) provide a hiring + community moat (devs who blog on Hashnode tell other devs).

Business 2 — B2B platform for engineering teams. This is where the money actually comes from. The three real SKUs:

1. Pro (~$99/mo per blog or seat): For solo professionals + companies running a single engineering blog. Includes advanced analytics, custom domain (also free on the lowest tier, but with reduced limits), priority support, AI writing assistant, ad-free. The Pro tier is mostly defensive — it keeps individual creators from churning to Ghost or Substack when they hit a feature wall.

2. Teams (~$199/mo for 5 seats, $10/extra seat per pricing page snippets): Multi-author engineering blogs. Companies like Stripe, Figma, and various Series A/B startups use Hashnode for their engineering blog because it's faster than building one in-house and better than a Medium-tagged "Stripe Engineering" page. The 5-seat starter slot is the wedge — small eng team of 3-7 fits perfectly.

3. Docs by Hashnode (custom pricing, $199-2000/mo realistic band): Launched public beta Aug 2024. Headless API docs + product docs with real-time collaborative editing, GraphQL API, SSO, audit logs, SOC2 + ISO 27001 roadmap. This directly competes with Mintlify ($120-500/mo), ReadMe (~$99-2K/mo), GitBook Enterprise ($249-$1K+). The launch HN thread framed it as "another docs product in 2024, here's why" — defensive positioning, but the real story is they had the publishing infrastructure already built and adding "docs" was an obvious adjacency.

The high-LTV segment is engineering blogs at funded startups: $200-500/mo blended for 3-5 years, with 80%+ gross margin (the underlying cost is mostly Postgres + CDN + occasional support hours). Even 200 of these accounts = $40-100K MRR steady state, and the sales cycle is short because the buyer (CTO or VP Eng) was already a Hashnode blogger as an IC five years earlier.

The headless CMS pivot is brilliant in concept and slow in execution. The thesis: companies don't want their docs/blogs locked into a vendor's frontend — they want the editor + workflow + auth, but their own React app. Hashnode gives that for free. But it requires the buyer to have a frontend team, which excludes the "I just want a working blog" market that pays Substack. So the headless mode is a sales accelerator for technical buyers but a churn-preventer for the rest.

What Hashnode hasn't built (deliberately or otherwise):

• No paid subscriptions for individual creators (Substack's moat)
• No ad network or sponsored-post marketplace (Medium's old monetization layer)
• No native podcast/video hosting (would compete with Beehiiv's expansion plays)
• No mobile app of consequence (the iOS app exists but is read-only; writers use desktop)

Each of those is a "could have, chose not to" — Sandeep has been clear in interviews that Hashnode is a writing infrastructure company, not a creator economy company. That focus is correct if you believe B2B SaaS for engineering content is the win condition. It's wrong if you believe the next $100M media company is built on a blogging platform's monetization layer.

Tech Stack Reverse-Engineered

I poked at the network requests on a live Hashnode blog and the rendered HTML structure.

Frontend — Next.js (the response headers leak x-powered-by: Next.js on some routes; the starter-kit repo is open source and uses Next.js 14 App Router + Tailwind). Pages are server-rendered with React Server Components on the newer routes, ISR for older content. The HTML is genuinely clean: minimal client JS, proper semantic tags, JSON-LD schemas (Article, BreadcrumbList, FAQPage on Q&A posts) baked in. This is why Hashnode posts rank — Google sees them as actual articles, not React app shells.

Backend — GraphQL endpoint at gql.hashnode.com. Apollo server is the most likely implementation (the schema patterns match Apollo's conventions, and Hashnode has hired Apollo-experienced engineers per LinkedIn). The schema is publicly documented at apidocs.hashnode.com — full CRUD on posts, comments, users, publications, with cursor pagination. This is the same API Hashnode's own frontend uses, which means there's no two-tier "internal vs public" API debt.

Data layer — Postgres for everything relational (posts, users, publications, custom domains). My guess at the analytics layer is ClickHouse or BigQuery for pageview/event data (the analytics dashboard has the latency profile of a column store, not Postgres). Vector search for Hashnode AI runs on... unclear, but the 1.25-1.5s vector search latency they publicly disclosed suggests Pinecone, Weaviate, or pgvector — most likely pgvector given the Postgres stack.

Edge / DNS routing — This is the technical moat. When you point yourblog.com at Hashnode via a single CNAME to hashnode-bcp.hashnode.network, Cloudflare Workers (or an equivalent edge layer) routes the request based on the Host header, looks up which Hashnode publication owns that domain, fetches the right content, and renders it server-side. SSL is auto-provisioned via Cloudflare's SSL-for-SaaS or similar. The "single CNAME, no apex bullshit" experience is the part competitors haven't replicated — Substack's custom domain flow is uglier, Ghost(Pro)'s is comparable but $9-25/mo gates it, Medium's is a $50 one-time fee with a worse setup.

AI stack — OpenAI API for the writing assistant + chat-search (the response latency and output style match GPT-4/4o). Embeddings probably via OpenAI's text-embedding-3 model. The "Hashnode AI" search across your own corpus is RAG on top of pgvector + GPT-4o-mini for cost reasons.

Auth + payments — Stripe for payments (the checkout flow is unmistakable). Auth is probably custom JWT + sessions (no obvious Auth0/Clerk fingerprints in the network tab). SSO for enterprise via SAML.

What's open source — The starter-kit (Next.js + Tailwind blog frontend that consumes their GraphQL API) is on GitHub with 1.5K+ stars. The core platform is closed.

The capital floor to replicate this stack from scratch with a small team: ~$150K runway for the first 12 months if you're a two-person team building it (founders unpaid, AWS/Vercel costs offset by Sequoia-style credits). The technical hard part isn't Next.js + GraphQL — it's the custom-domain DNS + SSL automation, which is a multi-week engineering project if you're not familiar with Cloudflare's SaaS APIs.

Distribution Playbook

How Hashnode actually got 4M devs and ~$2-3M ARR with $10M of total funding:

1. Founder credibility via SitePoint years (2013-2016). Sandeep Panda was a published SitePoint author writing JS tutorials before founding Hashnode. That's how dev-community trust compounds — by the time he launched a blog platform for devs, he'd written ~50 tutorials for the same audience. Co-founder Syed similarly published widely. The lesson isn't "publish 50 tutorials," it's "the founder spent four years being useful to the audience they later sold to."

2. Free tier with custom domain — the unfair offer. Substack's free tier doesn't give you a custom domain. Medium charges $50 one-time. Ghost(Pro) starts at $9/mo. WordPress.com gates custom domains behind $4-25/mo. Hashnode gives you a one-click custom domain for $0. That's the line in every "I'm switching to Hashnode" tweet. The cost to Hashnode is real (DNS automation engineering + Cloudflare SSL-for-SaaS volume), but it's the single biggest acquisition lever.

3. Show HN as a launch ritual. The Docs product launched on HN with a self-aware title — "we built another docs product in 2024, here's why." That post moved early signups + got the existing user base to upvote. The Hashnode team has launched on HN dozens of times since 2016. Sandeep's Twitter is full of "Show HN: we shipped X." HN remains the single highest-conversion launch channel for technical SaaS, and Hashnode treats every minor feature as a relaunch.

4. Sponsored writer programs. Around 2021-2022, Hashnode ran cash-prize hackathons + paid writing programs (Hashnode Bootcamps) that paid dev creators to publish to Hashnode-tagged streams. Net cost: probably $50-200K total. Net effect: hundreds of high-DA backlinks + thousands of new blog signups. The economics work because each signup creates content that pulls in more signups via SEO compounding.

5. Sequoia + Salesforce Ventures portfolio effect. Sequoia Surge connects you to every other Sequoia portfolio company in India + SEA. That's how engineering teams at funded startups discover Hashnode for their company blog — not through ads, through a CTO Slack channel intro. Salesforce Ventures from the 2022 round opens the same door inside US enterprise. The $10M of capital matters less than the LP/portfolio network it bought.

6. Dev Twitter (X) presence — Sandeep's personal account. Sandeep posts product updates, hiring notes, and Hashnode growth metrics from his personal account (@sandeepg33k). It's mostly product-channel content but with founder voice. Hashnode the company posts too. The split — founder voice + corporate updates — works because devs follow people, not brands.

7. Engineering blogs as B2B Trojan horse. When Stripe Engineering or Figma Engineering posts a deep tech article, the byline includes "powered by Hashnode" or the URL structure leaks it (engineering.<company>.com CNAME'd to Hashnode). Other companies' VPs of Eng see this, ask their team, and that's the sales motion. Hashnode probably never had to outbound those big logo engineering blogs — they came inbound because the product was visibly good.

8. The headless CMS repositioning. Around 2023-2024, Hashnode started leading with "headless CMS for developer content" instead of "blogging platform." This is a deliberate market repositioning: blogging platforms are commoditized (low ACV, high churn) but headless CMS is a $1B+ category (Contentful, Sanity, Strapi). By framing the same product with new vocabulary, they doubled their TAM on paper and unlocked sales conversations with CIOs and platform engineering teams that wouldn't have taken a "blog platform" pitch. The product didn't change much — the positioning did.

9. What they're not doing (and probably should be). No paid acquisition I can detect — no Google Ads for "Substack alternative" or "engineering blog platform." No LinkedIn Sponsored Content. No content marketing budget aimed at non-devs (CEOs, PMs, marketers who buy on behalf of eng teams). The growth is entirely product-led + community-led. That's beautiful when it works and brittle when the category gets noisier (which it has — Mintlify, Buttondown, Beehiiv, Ghost all overlap now).

The replicable playbook compressed: build years of audience trust → ship a free tier with a feature competitors charge for → relaunch on HN every quarter → run sponsored writer programs to seed content + backlinks → use VC portfolio for B2B intros → reframe positioning when ACV needs to grow.

Why This Works / Why Now

Hashnode works because four trends converged:

1. Devs hate hosting their own blogs but want to own their content. The decade-long failure of Medium (lost paywall, lost search rankings, lost trust around 2021-2022) and the corresponding Ghost-on-DigitalOcean migration wave proved this demand. Hashnode is the third path: own your domain, own your content (you can export anytime), but skip the VPS babysitting.

2. Engineering blog as a B2B marketing channel matured. Stripe Engineering, Figma Engineering, Vercel Engineering — these blogs are now table-stakes for any startup hiring senior engineers or selling to engineers. The internal "should we build a blog?" debate is over; the new debate is "WordPress, custom Next.js, or Hashnode?" Hashnode wins on speed-to-launch.

3. Next.js + headless CMS pattern went mainstream. From 2020-2023, headless CMS adoption inside engineering orgs exploded (Contentful, Sanity, Notion-as-CMS hacks). Hashnode's headless mode rides this wave — they're not selling a new pattern, they're being the easy version of a pattern devs already wanted.

4. AI-generated content devaluation makes provenance matter. When ChatGPT can write a "best practices for Postgres indexing" post in 30 seconds, the value of a human-written engineering blog with a real byline + custom domain + provable author history goes up. Hashnode's network surfaces those authored posts in a feed where the byline carries weight. Substack solved this for newsletters; Hashnode is solving it for technical writing.

What might break the thesis: if GitHub or Vercel decides to build native blogging (GitHub Pages already kind of is, just badly), the competitive floor caves in. If Substack adds a custom-domain + multi-author mode aimed at engineering teams, the Teams tier comes under pressure. If Mintlify or ReadMe lower their floor to free, the Docs business gets squeezed. None of those have happened yet — and the 18-month window the brief flags is real.

The "why not me, why not now" answer for replication: the bottom layer (Next.js + Postgres + GraphQL blog) is genuinely solved by anyone with 6 months of engineering. The DNS routing is a 2-4 week project if you've used Cloudflare's SaaS APIs before. The thing you can't shortcut is the audience trust + content backlog. That has to be earned, niche by niche.

Founder Profile

Sandeep Panda (CTO + co-founder) — IIIT Bhubaneswar CS '13, started his career as a trainee software engineer at Majestic Domains in Mumbai. Wrote ~50 JavaScript tutorials for SitePoint between 2013-2016, which is where the audience-building started. Co-founded Hashnode mid-2014 as a side project with Syed Fazle Rahman, then went full-time in late 2015 / early 2016 after the product hit 50-60% completion. Personal site sandeep.dev. Active on Twitter (@sandeepg33k) and his own Hashnode blog. The "Software Engineering Unlocked" podcast episode with Sandeep is the best long-form on his thinking — he frames Hashnode as "every developer should be able to own their content and tell their story."

Syed Fazle Rahman (CEO + co-founder) — Also IIIT Bhubaneswar, also SitePoint contributor, also Mumbai-based pre-Hashnode. He's the business-side founder, runs hiring + fundraising + GTM. Lower public profile than Sandeep, which is the typical CTO-public / CEO-quiet pattern in Indian dev-tool founders.

The note on funding: the 2022 Series A was actually led by Salesforce Ventures (the Sequoia round was the 2020 seed). Both rounds together = ~$10M. No publicly disclosed Series B as of mid-2026, which is a soft signal that revenue growth wasn't sufficient to raise at attractive terms, but also not a death signal — many India-based dev tools companies have stretched their Series A capital across 3-4 years and shipped Docs-by-Hashnode-style adjacency products to grow into Series B revenue.

What's distinctive about the founder profile: they're both writers who became platform builders for writers. That alignment shows up in product decisions — the editor is genuinely loved (rare for blogging tools), the SEO output is clean (because they care about being found), the markdown handling is precise (because they've copy-pasted from it for a decade). Founders who use their own product daily ship better products. Hashnode is exhibit A.