Public number: 67 weekly users. So I built a tool for myself instead
Doing the Levels thing: one real number, in public.
My resume tool ResuHive had 67 active users in the last 7 days (GA4, as of Sep 26). Search Console, last 28 days: 354 impressions, 10 clicks. Small. Honest.
Yesterday I shipped something different: a Chrome extension called Triage for TrustMRR. While browsing TrustMRR listings you can mark each product Pass / Later / Deep dive, add a note, and export everything as Markdown. It also restores your filters when you go back to the list. TrustMRR itself forgets them, and that bugged me every single day.
It was built on my own SaaS extension template, so it took one day. Now it's waiting for Chrome Web Store review.
Reality check: the closest competitor, Markers, has 266 users after 4 years. So I'm not expecting a business. It's a tool I use myself, plus a way to get some exposure from the TrustMRR ecosystem.
Next number: install count after review.
Before treating 67 WAU as traction, I’d break source: GA4 sessions vs Search Console clicks (you mentioned ~2–3/week) vs direct/referral. That tells you whether the public number is product use or just noise. For Triage, optimize week-2 return and one successful export over install count. Time-box any “tool for myself” side bet to 1–2 days with a kill rule, or it quietly eats the main loop.
Ran it. 57 sessions over 28 days, minus 7 vercel.com referrals that are my own preview deploys, so ~50 real ones. Direct 34, referral 12 (5 from here), AI assistant 2, organic search 0.
Two things I didn't expect. First, organic is a flat zero in sessions even though Search Console reports 10 clicks for the window — so my SEO isn't "small", it's not landing at all. Second, the AI assistant bucket exists at all: 2 sessions from chatgpt.com, which is a channel I've never done anything deliberate about.
Content is still the answer, but the split says the thing to fix first is that nothing I've published is being chosen in search, not that I haven't published enough.
Checking Markers' 266 users before expecting anything is the most underrated move in this post. Most people build first and look at the competitor's numbers after launch.
I'd run the same reality check on ResuHive, because it's the one with a real market behind it. Resume tools is a category where plenty of companies clearly pay to acquire users: you can see their active Facebook/Instagram ads for free in the Meta Ad Library, and their pricing pages tell you what a user is worth to them. If the big players are spending on ads for keywords close to yours, your 354 impressions aren't a demand problem, they're a distribution problem, and that changes what you work on next.
Good luck with the Web Store review.
The uncomfortable part is that I only did it because the tool was for me anyway,
so the number couldn't talk me out of it. If I'd been deciding whether to build it
at all, I'm not sure a 266-user competitor would have stopped me — it's easy to
read that as "the market is wide open" instead of "nobody wants this."
What I'd keep from it: the number is only useful if you decide in advance what it
would take to say no. Checking afterwards and rationalizing is the default.
Something in your numbers doesn't line up, and I think it's the most useful thing to look at. 10 Search Console clicks in 28 days is roughly 2-3 a week, but GA4 shows 67 active users in 7 days. So most of your traffic isn't coming from Google.
Have you checked the GA4 traffic source breakdown? It could be direct visits or a referral you didn't know about, or it could be bots or your own visits inflating the count. If it's a real channel, that's the thing to double down on before Triage even clears review.
You were right, and I went and did the split I should have done before writing the post.
28 days: 57 sessions total. 7 of those are vercel.com referrals, which is my own preview deploys, so call it 50. Organic search sessions: 0. Google Search Console says 10 clicks in the same window, so those clicks aren't even showing up as sessions. IndieHackers sent 5. ChatGPT sent 2. The rest, 34, is direct — and at this size direct mostly means me and people I sent the link to personally.
So the honest version of my own post is: it isn't "mostly my distribution posts" either. My distribution posts brought five. Thanks for making me look.
That's a real answer, and it's rare that "I went and checked" ends with the story looking worse, not better, than the version that would've been comfortable to leave alone. Most people who go check come back with the number they hoped for or just don't post the result.
The 10 GSC clicks not even showing up as GA sessions is worth sitting with too. That's not "traffic from a different source," that's your own two tools disagreeing about whether 10 real people even landed on the page at all — which is its own small version of the thing I keep running into: two systems that are supposed to be measuring the same event and aren't.
34 direct, 5 IH, 2 ChatGPT, 0 organic. So the actual next move isn't "double down on whatever's already working," it's "find something that works at all." Is Triage still on track to clear review, or did this change what you're doing next?
It changed what's next. Triage is still in review and I've stopped treating it as the main thing — "find something that works at all" is the honest framing, and for me that means fixing the paid path and then going and getting strangers to it, rather than shipping a second tool. The two-tools-disagreeing problem I'm resolving by trusting Search Console on clicks and my own deduped user count on everything else. I'll post the number again in four weeks either way.
The asset here isn't Triage, it's that your extension template turns a one-day build into a repeatable way to show up inside someone else's ecosystem — Markers' 266-in-4-years ceiling matters a lot less when the cost of each attempt is a day.
Your 354 impressions / 10 clicks resonates. We run an SEO tool — 4 visits a month right now, DA 3. Sharing real numbers is uncomfortable but it's the only way to build in public honestly.
The "bugged me every single day" line is the most reliable product signal I've seen. Market research, surveys, competitor analysis — none of them match the certainty of scratching your own itch daily. If you're the first power user and you can't stop using it, the product-market question becomes distribution, not validation.
The 266 users after 4 years for Markers is actually useful data. When you can see the ceiling clearly, you stop optimizing for growth anxiety and start shipping for utility. That's a better headspace for building. One-day build on a SaaS template plus real competitor benchmarking — that's a more honest launch process than most Series A decks describe.
DA 3 and four visits a month is the stage nobody writes posts about, so thanks for putting it here.
The uncomfortable part for me this week was splitting sessions by source: organic search came out at 0 over 28 days, while Search Console claims 10 clicks. 436 impressions means Google knows the pages exist and people are choosing not to click — that's a title and intent problem, not an indexing one. Separately I found I need roughly ten real backlinks before any of it ranks, and I have one.
If you ever write up what moved your first real click, I'd read it.
The one-day build is the right cost for this kind of bet — the danger with "built it for myself" tools is they quietly eat weeks. My rule: side bets get time-boxed to a day or two; anything longer has to earn a slot on the main roadmap. And the metric that matters isn't install count, it's whether Triage feeds ResuHive: if one Triage user a month turns into a real conversation about your main product, the extension pays for itself even at 50 installs. At 67 weekly users you can still talk to everyone, so pair the public number with one user conversation a week — the scoreboard is good content, but the interviews are the strategy.
thats intresting. i love building my own tools also to solve my problems. i remember when i built an automation tool because doing it with python became a long task i started with minor task and bam it became a huge project and a serious automation tool
"This is such a relatable journey! Building a tool out of personal need is often how the best solutions are born. Moving past the initial review hurdles and focusing on actual user engagement makes all the difference. Wishing you great success with it!"
Install count may be the wrong next number. For a triage tool, I’d watch how many people classify a few products, export once, then come back the next week. One install proves curiosity. A second session proves the workflow found a place.
Second-session return is a much harder and much more honest metric than installs, agreed. Installs I can move with a launch post; week-two return I can't fake.
I'm defining activation as "classified something and exported once", then tracking only how many of those come back seven days later. Right now the equivalent numbers on the resume side are 6 people created something and 4 exported, out of 255 first visits — so I already have the shape of the problem, just not the return leg of it.
If that seven-day number comes back near zero, the tool is a nice afternoon and not a product, and I'd rather find that out in two weeks than six months.
I actually like the way you're tracking the numbers honestly. With 67 active users, I'd be more interested in understanding where those users came from and what made them stick than chasing bigger traffic immediately. That could reveal the most repeatable acquisition channel before you scale anything.
Once Triage is live, what result would make the ecosystem exposure worth maintaining—new ResuHive visits, signups, or something beyond installs for the extension itself?
Respect for the honest number. With 67 weekly users you can still ask every one of them something. A single question after they finish a resume ("what almost made you leave?") would tell you more than the Search Console numbers right now.
respect for posting the real number. 10 clicks off 354 impressions is about 2.8%, which isn't bad at all for a brand new site, so the bigger lever is probably getting more impressions in the first place. building something you use every day is a good call too, you'll hit every annoying edge before a user does