380 pieces of content later - what actually worked
· Avery NXR
We've published 380 blog posts and Twitter threads over the last 18 months.
Some hit. Most didn't. The ones that hit weren't the ones we expected.
Here's what we've learned about content that compounds.
What we assumed would work (and didn't)
"Big idea" thought leadership pieces.
We wrote several 2,500-word essays about the future of AI agents, category dynamics, and where the industry is going. We were proud of these.
Reality: they underperformed by 3-5x versus tactical content. Big-idea content flatters the writer more than it serves the reader.
Product announcement posts.
"Introducing X." "We're excited to launch Y." Standard SaaS content marketing.
Reality: only converted for existing followers. Cold traffic bounced. Announcement posts assume the reader cares about the announcer. Cold readers don't.
Interview-format content.
We planned a "founder interview" series featuring other builders in our space. Only shipped two before abandoning.
Reality: interviews about other founders drive traffic to those founders, not to us. Unless the interviewee has a truly aligned audience, this format is a subsidy to someone else's brand.
What actually worked
Specific tactical posts with numbers.
"How we structured our AppSumo LTD" outperformed "The future of lifetime deals" by 8x.
Tactical posts get bookmarked. Bookmarked posts get shared. Shared posts drive sustained traffic for months.
Honest post-launch reflections.
Our "what we got wrong about PH" post drove more inbound conversations than our launch announcement itself.
Honesty about failure attracts operators who've had similar experiences. Those are the people who become customers.
Competitor comparison posts.
Our "Avery vs n8n" and "Avery vs Zapier" posts became the highest-converting pieces in our library.
Reason: buyers searching those queries have commercial intent. They're deciding between tools right now.
"How we decided" posts.
Any post explaining a specific decision (why we priced flat, why we're not raising, why we chose PH, why we said no to feature X) outperformed generic "here's what we do" content.
Decision posts show your thinking. Generic posts show your marketing.
Framework posts.
"The 5 gates before applying to AppSumo" — tactical, memorable, applicable. Format that works: numbered framework + specific criteria + honest tradeoffs.
The pattern behind what works
Every high-performing piece has three things in common:
1. Specificity.
Not "we optimize for growth." Instead "we're targeting 400 buyers in Tier 1 for AppSumo, projected $16K net after refund adjustment."
Specificity signals real work. Vagueness signals marketing.
2. Honesty about tradeoffs.
Not "our approach is best." Instead "we chose local-first, which means we lose customers who need cloud collaboration."
Honesty compounds. Marketing depreciates.
3. Usability.
The reader should be able to copy something specific from every post. A framework. A number. A decision criterion. A comparison table.
If your reader closes the tab and has nothing to do with what they just read, the post failed regardless of engagement metrics.
What "compounding" actually means
Content compounds when:
- Old posts continue generating organic search traffic months later
- Prospects reference specific old posts in sales calls
- Journalists find you via old posts and reach out
- Competitor employees quote your positioning back to you
- Community members share your posts in Slack channels you're not part of
Content decays when:
- The only readers are your existing followers
- Engagement drops immediately after posting
- No one references it 60 days later
- Search traffic never picks up
Track the compounding, not the initial engagement.
What I'd do differently
If I could restart our content strategy 18 months ago:
Ship less, publish deeper. 380 pieces was too many. 100 excellent pieces would have driven more compounding than 380 uneven pieces.
Prioritize competitor comparisons. These convert. Should have started them in Month 2 instead of Month 8.
Ship one framework per month. Frameworks (numbered lists, decision trees, criteria checklists) share better than essays.
Publish honest numbers regularly. Our metrics posts drove more inbound than our thought leadership. Should have committed to monthly public metrics from Day 1.
Skip the announcements. Product launches deserve announcements. Nothing else does. Everything else should be tactical or honest reflection.
For anyone building a content library
Publish for the reader who will find you in Google in 6 months, not the reader who sees the tweet today.
Tweet readers scroll away in 3 seconds. Google searchers spend 4 minutes reading if the answer is specific enough.
Optimize for the searcher. The Twitter engagement is a byproduct.
Compound over quarters, not weeks.
→ avery.software — our writing continues.