Feature Adoption Tracking Agent

The short version

The Feature Adoption agent tracks adoption curves daily for every feature shipped in the last 3 months. It pulls from Amplitude or Mixpanel, breaks the data into three signals (trial rate, retention rate, cohort breakdown), and flags lagging segments before the trend hardens. When SMB adoption sits at 12% while enterprise is at 40%, you get a root-cause hypothesis and a recommended intervention (in-app tour, email campaign, onboarding tweak). Daily snapshot at 4 PM, deep dive Monday at 9 AM. Connect feature flags so the agent knows when each user got access, then run it on your three most recent launches.

You shipped a feature on Monday. By Thursday, you're wondering: is it getting adopted?

The problem is you won't know for 2-4 weeks. By then you've already moved on to the next feature. And if adoption is slow, you have no idea why - was it hidden in the product? Did the onboarding not explain it? Is it genuinely not valuable?

The Feature Adoption agent tracks adoption curves daily. For any feature released in the past 3 months, you see: what % of your user base has tried it, what % is using it regularly, and which cohorts (segments, tiers, geographies) are using it most. It also identifies: which cohorts are lagging, and what intervention might help them adopt.

How It Works

The agent connects to your analytics and tracks adoption in three ways:

Trial tracking: How many users have tried the feature (performed the core action at least once)? What % of your total user base is this? How fast did trial growth happen?

Retention tracking: Of those who tried it, how many came back to use it again? This is the real adoption signal - first use doesn't count if they never return.

Cohort analysis: Breaking adoption down by segment (SMB/enterprise), company tier (free/paid), geography, or use case. "Enterprise is adopting at 40%, SMB at 12%. Why?"

Intervention recommendation: If a cohort is lagging, the agent suggests why and what might help. "SMB adoption is low, but those who did try it are using it regularly. They might just need better onboarding. Try a Slack notification or in-app tour."

Data Sources and Setup

Prerequisites: You'll need:

• Analytics platform: Amplitude or Mixpanel with feature event tracking
• Feature flags: Rolled out features so you know when each user got access
• CRM / Segment data: Customer attributes (company size, industry, tier)
• Support data: Support tickets related to new features
• Interview/feedback data: Why customers do or don't use new features
• Email/in-app messaging data: What onboarding did they receive?

Schedule: Daily at 4 PM (adoption snapshot). Weekly Monday at 9 AM (deep dive and intervention recommendations).

The Claude Prompt

You are tracking feature adoption and identifying intervention opportunities.

Here are the features we shipped recently:
[FEATURES: release date, description, target audience]

Here's the adoption data from the past 4 weeks:
[ADOPTION DATA:
- Feature: feature name
- Launch date
- Total users with access
- Users who tried it (% and trend)
- Users retaining it (using it 2+ times in past week)
- Adoption by segment (SMB, enterprise, free tier, paid tier, etc.)]

Here's cohort/segment details:
[SEGMENTS: company size, industry, tenure, use case, geography]

Here's support data related to these features:
[SUPPORT: tickets mentioning new features, topics, volume]

Here's feedback we've gathered:
[FEEDBACK: why users like the feature, why some avoid it, what's confusing]

Please analyze and report:

1. **Feature Adoption Summary**
   For each feature:
   - Trial rate (% of user base who tried it)
   - Trial velocity (how fast is trial growing? flat? accelerating?)
   - Retention rate (of those who tried, % using it regularly)
   - Current trend: accelerating / stable / declining adoption?

2. **Cohort Breakdown** (IMPORTANT)
   - Which segments are adopting most? least?
   - Is adoption pattern expected? (does it match target segment?)
   - Any surprises? (a segment you didn't target is adopting heavily? or lagging?)

3. **Retention Deep Dive**
   - Of users who tried the feature, what % never came back?
   - How long did they stay engaged before dropping off?
   - Which cohorts have best retention? worst?

4. **Lagging Cohorts** (PRIORITY)
   For any cohort with under 25% trial rate:
   - Why might they be lagging? (feature doesn't fit their use case? hidden in product? poor onboarding? competitive alternative?)
   - What's the consequence of low adoption? (revenue impact? churn risk?)
   - How do we know? (support tickets? survey? research?)

5. **Intervention Recommendations**
   For each lagging cohort, suggest:
   - Root cause hypothesis
   - Test: how would we validate the root cause?
   - Intervention: what would help them adopt? (better onboarding? email campaign? in-app tour? product changes?)
   - Effort to implement
   - Expected impact (how much would adoption improve?)

6. **Success Features**
   - Which features are breaking through adoption-wise?
   - What are they doing right?
   - Can we apply those lessons to slower features?

Format for quick scanning: I want to see adoption metrics, flagged lagging cohorts, and recommended actions.

What You Get

Instead of wondering if features are being used:

• Daily adoption visibility: You see adoption curves in real-time
• Cohort signals: You know which segments love the feature and which ignore it
• Early intervention: You catch low adoption before it's too late to do something about it
• Validated hypotheses: Not guessing why adoption is low - backed by data
• Actionable recommendations: "SMB adoption is low, but they need better onboarding. Try in-app tour." Not "increase adoption" (unhelpful).

Real outcomes:

• You catch adoption stalls within days, not weeks
• You can run quick interventions (emails, in-app tours) to unblock adoption
• You understand feature-segment fit (some features just aren't for SMB - that's okay to know)
• You build momentum: early wins get attention and more feature launches

For the full agent fleet and scheduling details, see Your AI Agent Fleet. For the OKR tracking layer that ties feature adoption to outcome metrics, see the OKR Progress and Prediction Agent.