Let me start with the number that stings: zero. After three months, 155 published posts, 15 pages, and 59 tags on devangshumitra.com, I have zero email subscribers. Not ten. Not five. Zero.
That number is important because it's real. It's easy to celebrate "155 posts in 3 months" as a productivity achievement. It's harder to sit with the fact that content volume alone built an audience of exactly nobody. This post is about what those 155 posts actually taught me - including the lessons I didn't expect and the ones that cost me something to learn.
The 5-Pillar Architecture
I didn't start with 155 posts and a clear architecture. I started with about 30 posts and a vague sense of what I wanted to write about. The architecture emerged from looking at what I'd actually written and noticing the clusters.
The five pillars I landed on:
- AI Product Management: frameworks, career path, skills development, the PM role in AI organizations
- AI Applications in Industry: how specific sectors (healthcare, finance, manufacturing, retail) are actually deploying AI right now
- AI Research and Trends: technical developments, model releases, benchmark results - written for practitioners, not researchers
- Agentic AI Systems: architectures, use cases, the practical state of autonomous AI workflows
- Enterprise Innovation: how large organizations build AI capabilities, the organizational dynamics, the failure modes
Healthcare appears across all five pillars - it's my deepest domain knowledge - but it's not a container. That was a deliberate choice, and I think it was the right one. A healthcare-only brand would have been narrower but more coherent for a specific audience. A multi-pillar brand is broader but serves the repositioning I'm doing toward a generalist PM role.
The lesson: your content architecture is a positioning decision. It tells your audience who you are and who you're for before they read a single word.
Why Content Balance Matters More Than Volume
In the first 60 days I published heavily in two pillars and lightly in the others. I was writing about what I found interesting, not what the architecture needed. The result: the site read as two separate blogs awkwardly coexisting.
I started tracking pillar distribution explicitly - not to enforce artificial balance, but to ensure I wasn't accidentally building a single-topic site when I'd committed to a multi-pillar strategy. The right balance for me ended up being roughly 25-30% AI PM content, 20-25% industry applications, 15-20% agentic AI, and the rest split between research/trends and enterprise innovation.
This matters for SEO as much as positioning. A site with strong topical authority in multiple clusters ranks better than a site that goes deep in one area and has a few scattered posts in others. Google rewards comprehensive coverage of related topics, not just depth in a single subtopic.
The SEO Strategy: What I Got Right
Comparison posts are my best-performing content by a significant margin. "[Tool A] vs [Tool B]: Which is Right for Enterprise AI?" format consistently outperforms think-pieces, tutorials, and trend analyses. Why? Intent alignment. Someone searching "GPT-4 vs Claude for clinical applications" has a decision to make. If my post helps them make it, they read the whole thing, they leave satisfied, and they sometimes share it. That's a complete reader-writer transaction.
Long-tail keywords with low competition are where a new domain competes. I can't rank for "AI product management" - that's dominated by established content. I can rank for "AI product manager vs data scientist: role boundaries in healthcare tech" - that's specific enough that there's limited competition and clear enough that it's what someone is actually searching for.
I found the right keyword pattern by looking at the questions I was personally asking when I started in AI product roles. If I was Googling it, someone else was too. This is a simplistic heuristic but it's been consistently useful.
What I Got Wrong: Trends Content
I published a significant number of posts analyzing specific AI developments in specific weeks. "This Week in AI: What Matters for Enterprise" style content. It generated reasonable early traffic but ages terribly. Three months later, those posts are stale, they're not earning ongoing search traffic, and they're diluting the site's overall authority because they're low-value pages that have accumulated.
The lesson: if you're building a content library rather than a news operation, evergreen content is a fundamentally better investment. The hours I spent on trends posts would have generated more long-term value as close look evergreen pieces.
I'm not deleting the trends posts - that would hurt the SEO I've built - but I've stopped producing new ones. I'm updating the better ones to be more timeless and letting the weaker ones age out naturally.
Gated vs. Public Content: The Fence I'm Still Sitting On
I have an Interactive Tools section with an AI Maturity Assessment and an ROI Calculator that I've partially built. The question I haven't resolved: should these be gated (email required) to start building the list I don't have, or should they be free to maximize reach and sharing?
The standard advice is gate them - use the tools to build the list. But I've seen too many tools that get buried behind gates and die there, generating neither list subscribers nor organic reach. The counter-argument is that free tools get shared, linked to, and indexed - and that SEO value potentially exceeds the list-building value at my current stage.
I'm leaning toward launching the tools free, measuring usage for 60 days, and then adding a "save your results" email gate if organic uptake is strong. That feels like a sequencing that respects my actual scale rather than assuming I have the authority to charge attention for access to my tools.
The Zero Subscribers Problem
Let's go back to zero. How does a site with 155 posts have no email subscribers?
The simple answer: I haven't promoted it. I have a minimal LinkedIn presence, no Twitter/X strategy, no newsletter promotion, no community participation in places where my target audience hangs out. The newsletter CTA on the site is a Ghost default - I haven't even customized the subscribe page copy.
The more complex answer: I built the content foundation before the distribution strategy, which is the right order - but then I kept building the content foundation without ever building the distribution. I kept saying "I'll focus on distribution once the content library is solid," and the content library kept not being solid enough.
This is a trap that content creators fall into constantly, and I walked into it despite knowing exactly what it was. The content is never solid enough to start distributing. Distribution is the work. You start it alongside the content, not after it.
What's Next: Month 4 and Beyond
The pivot I'm executing in month four:
Stop publishing new posts for two weeks. I'm auditing the existing 155, fixing internal links, updating stale content, and adding schema markup. One post with proper SEO infrastructure outperforms five posts without it.
Start the LinkedIn-to-blog flywheel. Three posts per week on LinkedIn, each pointing back to a relevant close look on the blog. LinkedIn is where my professional network already is. The blog is where I want them to go.
Build one genuinely useful interactive tool. The AI Maturity Assessment, fully functional, launched free, and promoted properly. One tool, done well, beats five half-built tools sitting in draft.
Write one post per week that I'm genuinely proud of. Not to add to the count. Because the 155 posts I have include a lot of adequate content and not enough exceptional content. I know the difference when I'm writing. The exceptional posts take twice as long and generate three times the engagement. I need to shift the ratio.
The thing this three-month experiment has clarified for me: content strategy is not about having a lot of content. It's about having the right content in front of the right people at the right moment. I have more of the first than I've earned yet of the second and third. That's the work ahead.
If you've been reading along and you have opinions about where I should focus - the tools, the distribution, the content quality, all of it - I'd genuinely value your perspective. This is building in public, which means the feedback is part of the process.
