I lost 82% of my traffic while building my product.
I lost 82% of my traffic while building my product.
4.84k monthly users → 846.
At first I assumed I had a traffic problem.
So my brain immediately suggested:
More SEO
More content
More distribution
More promotion
But after looking deeper, I realized something uncomfortable.
Traffic wasn't the biggest problem.
It was easier to think about traffic than to think about execution.
While traffic was falling, I spent months rebuilding TruthLoop's investigation engine, debugging AI workflows, rewriting evidence collection systems, and constantly changing the product.
The visible problem was traffic.
The invisible problem was where my attention was going.
I think founders do this more often than we admit.
We focus on the metric that's easiest to measure while avoiding the question that's hardest to answer.
A question I keep coming back to:
Am I solving the problem that hurts the most?
Or the problem that's easiest to think about?
Curious if anyone else has discovered that the problem they were trying to solve wasn't actually the real problem.
The lesson is probably true, and I would still spend ten minutes ruling out the boring explanation, because the two are not mutually exclusive.
Sites do not usually lose 82% to neglect. Neglect produces a slow slide. Going from 4.84k to 846 during the months you were rebuilding the product is much more often the shape of something you shipped: URLs changed without redirects, a noindex left on from staging, a robots.txt edit, or a framework migration that changed how the pages render.
Search Console settles it quickly. Look at whether the fall was a cliff on one date or a gradual slope, and whether impressions dropped alongside clicks or only clicks did. A cliff means you caused it, which also means it is recoverable. A slope means it is genuine competition and your reflective read is the right one.
Worth checking, because those two conclusions lead to completely different next months, and the more comfortable one is the one that does not ask you to go and look.
An 82% drop is brutal, but at least it exposed the tradeoff clearly. I’d give the channel that used to work a fixed weekly maintenance floor, then allow only one build bet at a time. What caused most of the decline—publishing stopped, rankings slipped, or the product changed?
The 82% drop makes the remaining 846 users more interesting than the traffic loss itself. Do you see any meaningful difference in what the users who stayed are doing with TruthLoop versus the users who stopped coming back?
That's a great observation. I focused on the people who disappeared. The people who stayed may reveal a much more useful pattern. I haven't analyzed that group deeply yet, but I suspect there's a lesson there.
That retained-user segment could be more informative than the traffic loss itself. If you’re open to it, what’s the best email to reach you on?