Every two weeks I'm going to give you a look at the actual numbers. The real ones, not the ones that make us look better than we are. I've written enough press releases in my legal career to have a deep distaste for metrics reporting that buries the inconvenient truths three paragraphs in.
Today is the Week 2 recap. Here's where things stand.
The newsletter.
Subscriber count, open rate, click rate — these are the foundational metrics for the content operation. They tell you whether the audience is growing and whether people are actually reading what you're sending.
[NUMBER BLOCK — REVIEW BEFORE PUBLISH: insert actual figures, directional language, or remove section entirely depending on where revenue stands at time of release]
What I'll say about the newsletter metrics directionally: the quality of engagement matters more than the size of the list at this stage. An open rate that suggests people are actually reading — rather than the list inflating with people who clicked a download link once — is the signal I'm watching most closely right now.
The directory experiments (LOZ and PCB).
The relevant metrics here are organic impressions in Google Search Console, rankings for target keywords, and (eventually, not yet) inbound lead inquiries or listing upgrade conversions.
[NUMBER BLOCK — REVIEW BEFORE PUBLISH: insert actual figures, directional language, or remove section entirely depending on where revenue stands at time of release]
The direction here is what I expected: slow early traction followed by a more meaningful inflection once enough content mass exists for Google to understand what the sites are about. I don't expect revenue from the directories in the first month. The question is whether the trajectory is consistent with what a revenue-producing asset looks like at month three or four.
The POD stores (KHD and EC).
Traffic, conversion rate, and revenue. All three are early-stage.
[NUMBER BLOCK — REVIEW BEFORE PUBLISH: insert actual figures, directional language, or remove section entirely depending on where revenue stands at time of release]
The POD stores have the most honest variance between expectations and reality so far. I wrote about this in the Day 6 issue. The design pipeline issues are resolved, the product catalog is expanding, and traffic is the current constraint. The numbers will look different when there's a systematic acquisition strategy attached to them, which is ahead of us.
What I'm not reporting yet.
I'm not reporting app revenue because the apps aren't live yet. I'm not reporting investment returns because the paper trading experiment is earlier than a few weeks old and reporting anything this early would be noise dressed up as signal. I'm not reporting community revenue because The Upstream is still in its early formation.
When there's something meaningful to report in those areas, I'll report it. Empty line items dressed up as transparency are not transparency — they're the appearance of transparency without the actual work.
The meta-observation about metrics dashboards.
I want to say something about metrics reporting that I think gets underappreciated in the "build in public" genre.
Reporting numbers is easy when the numbers are good. The test of a metrics dashboard is whether you report the same numbers when they're disappointing. The pressure to curate — to find the one metric that's trending the right direction and feature that one — is real, and it's a pressure I've noticed in myself over the past two weeks.
The commitment I'm making to you is: the same level of detail in good weeks and bad weeks. If the directory traffic drops in month two because of something I can't explain, that goes in the report. If the newsletter open rate declines, that goes in the report. If revenue in a stream is flat or declining, that goes in the report.
The reason this matters is structural: the documentation is only valuable if it's honest. An honest failure post is more valuable than a curated success story, not because failure is more interesting, but because honest documentation of failure contains information that curated success stories don't. This newsletter exists to be useful, and useful means accurate.
Looking ahead.
Week 3 is going to be more educational — practical AI tools, how to use them, what the limitations are. I've gotten a lot of questions from readers who are earlier in the AI learning curve and want something more tactical. Week 3 is for them.
Week 4 goes deeper into specific stream experiments. By then, there should be more to report.
This Week in AI: Google's continued update cycle for search ranking has kept the SEO community in a near-constant state of analysis and revision over the past year. For content operations like ours, the consistent signal in all of it is: depth and genuine usefulness outperform optimization tricks. Write for people, not algorithms, with enough technical attention to be findable. The fundamentals haven't changed; the execution tolerance has gotten narrower.
All of the frameworks we're using to track these streams are available in the free toolkit at start.tenstreamslab.com.