Lesson 8 of 87 min read
Measure, listen, and find your first users
Read your own numbers, tag your links, and find the conversations where people already ask for what you built. Your first users are in other threads.
In this lesson you will
- Set a simple baseline of your own numbers and read it monthly
- Tag launch links so you know which posts bring signups
- Find and join conversations where people describe the problem you solve
After a few weeks of posting you will want to know if it is working. The trap here is the opposite of the one in lesson 1. Instead of ignoring distribution, you stare at it: refreshing like counts, comparing yourself to accounts that have been posting for years, and changing your plan every time one post does badly.
This lesson gives you a calmer way to measure. You will set a small baseline of your own numbers, read it on a schedule, tag your links so you can see which posts actually bring people to your product, and then do the thing that finds early users faster than any post: go to the conversations where people are already asking for what you built.
Measure against yourself, not others
Other people's numbers are noise for you. Their audience, niche, history and luck are different. The only comparison that tells you anything is you against your own past.
Pick a small set of numbers and track them monthly. For a developer building in public, these are enough:
| Number | Why it matters |
|---|---|
| Posts published per week | Did you keep your rhythm? Everything else depends on this |
| Replies and reposts per post | People engaging, not just scrolling past |
| Profile visits or follows | People curious enough to look at you |
| Clicks on tagged links | People who went to your site, docs or repo |
| Signups, stars or installs from tagged links | The outcome you actually care about |
Ignore impressions for decisions. They swing with things you do not control. Use them only as context.
Read your numbers once a month, not daily. A month of posts is enough to see a pattern; a day is not. Write down three things each month: what you posted, what moved, and one thing you will change next month. That small log becomes your playbook.
Posting frequency: more is not always better
One question comes up quickly: should I post more? The honest answer is "it depends on what each post earns". Doubling your posts is only worth it if the extra posts earn something. If each post earns half as much at the higher pace, you doubled your work for nothing.
Here is a simple way to check, using your own history:
- Group your past weeks by how many times you posted (for example, weeks with 2 posts, 4 posts, 6 posts).
- For each group, calculate the average replies and clicks per post.
- Compare. If per-post results hold steady as frequency goes up, posting more is paying off. If they drop sharply, you are past your useful pace.
You need several weeks at each pace before this means anything. Until then, keep the rhythm from lesson 4 and let the data collect.
Content decay: how long a post keeps earning
Posts do not all fade at the same speed. A short X post may earn nearly everything it will ever earn in its first day. A dev.to article or a YouTube demo can keep bringing readers for months. That difference should shape where you spend your effort.
To see it, look at a few of your posts more than once: a day after posting, a week after, a month after. Note how much of the total arrived after the first day.
- If most of a network's results arrive in the first day, plan for frequency and timing there; each post is a short moment.
- If results keep coming for weeks, invest in quality there; each piece is a long-term asset, and it is worth updating old posts and linking to them from new ones.
This is also how you check whether your depth network from lesson 2 is doing its job. Depth pieces should decay slowly. If they fade as fast as short posts, they may not be finding their audience.
Tag your links
Platform analytics tell you what happened on the platform. They do not tell you what happened after someone clicked. For that, add UTM parameters to every link you post.
UTM parameters are small tags on a URL that your website analytics read. A tagged link looks like this:
https://yourproduct.dev/docs/quickstart?utm_source=x&utm_medium=social&utm_campaign=v1-4-release
Keep a simple naming scheme and stick to it:
- utm_source: the network, in lowercase (x, linkedin, bluesky, reddit, devto).
- utm_medium: always
socialfor posts. Use something else for newsletters or ads. - utm_campaign: the release or topic, in lowercase with dashes (v1-4-release, launch, rollback-guide).
Then once a month, check your site analytics for visits and signups grouped by campaign and source. You will learn which networks send people who actually try your product, which is often different from which networks send the most likes.
Listen where your users already talk
Here is the most useful idea in this course for finding early users: they are probably not following you yet. They are in someone else's thread, asking a question your tool answers.
People describe their problems in public all the time: "Is there a tool that does X?", "How do you handle Y without Z?", "Our deploys keep failing because…". If you find those conversations while they are fresh and reply with a genuinely helpful answer, you reach a person at the exact moment they need you.
How to do it by hand:
- Write your problem words. Not your product name; the words people use when they have the problem. "Migration rollback", "schema change broke prod", "webhook retries", "release notes take forever".
- Search them weekly on the networks where your audience talks: X, Bluesky, Reddit, Mastodon, the forums from lesson 2.
- Reply to help, not to sell. Answer the question first. Mention your tool only if it truly fits, and say it is yours.
- Keep a list of the people and threads you replied to, and what happened.
Watch for the community rules from lesson 2. A helpful reply that mentions your tool once is welcome in most places; the same reply pasted into twenty threads is spam.
Budget twenty minutes, twice a week. In the early days this often finds more real users than all your own posts put together, and every conversation teaches you the words your audience uses, which makes your posts better.
Turn the numbers into one change a month
Numbers are only useful if they change what you do. At the end of each month, look at your log and make exactly one change, using these rules:
- If you missed slots, cut the rhythm down before changing anything else. Fewer posts you keep beat more posts you skip.
- If replies are low but clicks are fine, your posts inform but do not invite. Try ending more of them with a real question.
- If likes are high but tagged clicks are near zero, your posts entertain but do not lead anywhere. Link to something specific: a guide, a demo, a changelog entry.
- If one network sends almost all your signups, shift effort toward it, even if another network gets more likes.
- If nothing moves after two months of steady posting, revisit your positioning line from lesson 1. The audience may be wrong, not the posts.
One change at a time keeps the cause clear.
Worked example: how we listen for our own product
We do this for our own product too, and our problem words are a good example of the difference between product words and problem words. Our product words would be things like "social media scheduler" or "MCP server for social posting". Our problem words are what people actually type when they are stuck:
- "post to every network at once"
- "schedule posts from my coding agent"
- "announce releases automatically"
- "cross-post to dev.to with canonical"
- "alternative to" + the tools we have compare pages for
That last one matters. When someone asks for an alternative to a tool we have written an honest comparison of (our compare pages cover several), the most helpful reply is often the comparison itself: an honest side-by-side is more useful than a pitch.
A reply in the shape we aim for, to a question like "Any way to post release notes to X and LinkedIn without doing it by hand?":
A few ways, depending on how much you want to automate:
- Simplest: a CI step that creates a draft from the release
and you approve it.
- If you use GitHub releases, any tool that watches the
releases.atom feed can do the drafting.
Disclosure: we build one of those tools, happy to share
how we set it up if useful.
It answers first, gives options that do not require us, and discloses. That is the pattern.
Recap
- Compare yourself to your own past, monthly, with a handful of numbers.
- Check posting frequency by results per post, and decay by how long posts keep earning.
- Tag every link with UTM parameters so you can see which posts bring real users.
- Your first users are asking questions in other people's threads. Find them and help.
Exercise
Set your baseline and listening list
- Make a simple spreadsheet with one row per week for your last four weeks (or start this week): posts published, replies, reposts, follows, link clicks.
- Choose your UTM naming scheme for source, medium and campaign. Use the free UTM generator to tag the links in your next four scheduled posts.
- Write ten problem words or phrases your audience uses, in their words, not your product's.
- Search five of them on two networks where your audience talks. Save the three most recent threads where someone asks for help you can genuinely give.
- Draft a helpful reply for each, answering first and disclosing that the tool is yours if you mention it.
- Done looks like: a baseline sheet, four tagged links, a list of ten problem words, and three ready replies.
Go further
Radar leads and analytics in FoPost
Radar in FoPost runs your problem words as trackers on a schedule across supported networks, sorts each mention by type, and turns posts that show buying intent into scored leads with a drafted first message you send yourself. Add a keyword tracker, then work the results in Radar leads; read what each posting pace earned per post in posting frequency. See social listening for the trackers and networks covered.
Check yourself
1Why read your numbers monthly instead of daily?Show answer
A day of results is noise; a month shows patterns you can act on without changing your plan after every post.
2How do you check whether posting more often is worth it?Show answer
Compare results per post across weeks with different posting counts. If per-post results hold steady as you post more, the extra posts pay off.
3What makes a good reply in someone else's thread?Show answer
It answers the question first, offers options that fit, mentions your tool only if it truly helps, and says plainly that it is yours.
Keep reading, free
Tell us a little about you and the rest of this lesson, and every lesson in every course, opens on this device. No account and no card.