datahogProduct analytics,
run by your agent.
datahog is the first analytics platform built agent-first. You paste one snippet, connect your agent, and tell it to set things up. It names your events, builds your dashboards, and watches your app to catch issues before your users do.
placeholder
Free tier at launch: 1M events a month. See a real digest
That Jul 3 spike is your Reddit launch thread — 1,204 visits, 89 trials, $445 MRR so far.
See the cohortTop sources
visits · conversion
Top pages
visits/2,933/pricing1,401/blog/launch1,188/docs512/signup486Countries
shareSignup funnel
14 daysThe card form loses 42% of mobile users. About $6,400 a month.
Watch replaysUser journeys
top paths68% of TikTok visitors bounce on Landing. Reddit users read Pricing 3× more.
Retention
weekly cohortsUsers who invite a teammate in week one retain 3.2× longer.
Revenue attribution
first click → Stripe, 14 daysReddit converts at 5.3× ROAS. TikTok converts at 0.7×. Moving that budget is worth about $1,900 a month.
See the math
Stripe
Next.js
Vercel
ClaudeYou set up analytics once. Then it went stale.
Be honest. You added a tracking snippet the week you launched. You built a dashboard, checked it every morning for two weeks, and then never opened it again. The events you named back then no longer match your product. Half the buttons you ship today are not tracked at all. The one person who understood the setup got busy, or left, or was always just you. So now, when a number feels off, you dig through stale charts and guess. The answers are sitting in your data. Nobody is asking the right questions.
So we hired a hog. It digs while you build, and reports what it found.
Three steps. That’s the whole setup.
Paste one snippet
Drop a single line into your app’s head tag. Autocapture starts recording every click, pageview, and form on its own. There is nothing to name and nothing to configure.
Install the MCP
One command connects your agent to datahog. Claude, Cursor, or anything that speaks MCP. Every dashboard, funnel, and query becomes a tool your agent can call.
Tell your agent to set it up
Say “set up my analytics.” Your agent already knows your product, so it names the events that matter, builds the dashboards, and wires the funnels. Ten minutes later you have analytics a data team would bill a retainer for.
Insights with dollar signs, not charts to interpret.
The hog ties every visit to what it turned into. Not “traffic is up.” You get findings like: your Reddit launch thread converted $2,350 at 5.3× ROAS, four times what TikTok returned on the same spend. Every finding names the channel, the money, and the fix. The users and replays behind it are one click away, so you can act on it in the same sitting.
“Upgrade” precedes every plan change. Track it as upgrade_intent?Zero setup. Your agent names what matters.
Autocapture grabs every interaction from the moment the snippet lands, so nothing is lost while you decide what matters. Your agent watches real usage and proposes the events worth naming. A button that fires before every plan change becomes upgrade_intent. The form that gates your signup becomes signup_started. You confirm with one click, and when your app changes next month, the tracking updates with it instead of rotting.
See the exact moment. Skip the scrubbing.
When a number looks wrong, watch what actually happened. datahog records real sessions and jumps you straight to the friction: a rage click, a dead button, a form abandoned at the card field. You will not scrub through hours of footage. The hog queues the ten sessions worth your time and pins the exact second things went sideways.
The hog hunts at night.
You close the laptop. The hog keeps digging. All night it sifts your clicks, replays, funnels, and revenue, looking for the things that shifted while you were not watching. By Monday morning the findings are in your inbox, sorted by dollars, each one with the evidence attached. You read it with your coffee and know exactly what to fix this week.
Your Reddit launch thread converted $2,350 at 5.3× ROAS. That is 4× better than your TikTok spend.
Shift the budget? →
The card form loses 42% of mobile checkouts. 214 users, about $6,400 a month. Replays attached.
Watch the replays →
@tokgrowth’s video drove 1,204 visits → 89 trials → $445 MRR. Worth re-booking.
See the cohort →
Thursday’s deploy cut pricing-page signups 38%. The diff is attached.
View the diff →
Every button is a tool. If a human can click it, your agent can call it.
Everything datahog does is reachable over MCP, REST, or the CLI. Your agent can pull a funnel, list the users about to churn, mark a deploy on every chart, and read the digest as data. It already has your context. It knows what you shipped and why. Give it the numbers and it closes the loop.
run_funnelbuild a funnel, get conversion at each stepfind_at_risk_userslist accounts trending toward churn this weekquery_in_englishask a question in words, get back a chartannotate_deploypin a release marker on every relevant graphget_digestpull the latest weekly findings as datawatch_sessionsfetch the replays worth watching for an eventPick a plan. Start free.
No credit card. Change your mind whenever.
Free
For the solo founder who wants answers, not setup.
- 1,000,000 events per month
- 1 site
- All analytics: trends, funnels, journeys, attribution
- Weekly digest of what you missed
- 2,500 session replays (coming soon)
- Community support
Pro
For the growing team that wants more room and faster help.
- Everything in Free
- 2M events/month, then $5 per extra million
- 10 sites
- 25,000 session replays (coming soon)
- Unlimited seats
- Priority support
- 2 years of data kept
Questions. Straight answers.
What is datahog?
Analytics for websites and apps, built for agents to set up and run. Agent-first means every button here is also a tool your agent can call over MCP, REST, or the CLI. You get the dashboards; your agent does the work.
How is it different from the analytics I already set up?
Be honest: when did you last open it? Traditional analytics goes stale because a human has to maintain it. Events rot, dashboards drift, and one person on the team holds it all. With datahog your agent sets it up on day one and keeps it current as your app changes.
Why does an agent get better answers?
Context. Your agent already knows your product, your funnel, and what you shipped this week. datahog hands it the data and the tools, so it asks the questions you would not think to ask, and updates the tracking when your app changes.
Do I need an agent to use it?
No. Everything works by hand too. The dashboards, funnels, and replays are all yours to click around. The agent is just the fastest operator.
What’s the catch with the free tier?
There isn’t one. Most small sites fit in 1M events and stay free for good. When you grow past it, Pro is there. We only make money when you’re already winning.
Ready to see what you’re missing?
Paste one snippet into your site. That is the whole setup. The hog starts watching right away, your agent builds the dashboards, and your first digest lands Monday morning. If it finds nothing, it says so. If it finds money, you will know exactly where.
placeholder
Free tier at launch: 1M events a month. No spam, one email when your spot opens.