Getting Started with LENS

Get LENS configured and receiving insights in under 5 minutes.

Prerequisites

• A Sealmetrics account with an active plan
• At least 7 days of tracking data (so LENS has enough history to compare against)
• Email notifications enabled in your account settings

Step 1: Access LENS

1. Log in to your Sealmetrics dashboard
2. Click LENS in the main navigation
3. You'll land on the LENS overview, where you can open Chat or Reports

Step 2: Pick a provider

The LENS assistant supports two options:

• Seal AI Private — the managed platform LLM (EU-hosted, no prompt retention). No API key needed; usage counts against your organization's quota of 5M tokens per month, plus any non-expiring token packs. Owners/admins can check consumption anytime in Organization → Seal AI Usage.
• Bring-your-own-key (BYOK) — connect your own provider API key. No quota imposed by Sealmetrics; you pay your provider directly.

If your organization is entitled to Seal AI Private (included in Scale/Enterprise, add-on on Growth), it's active by default. To use a BYOK provider, go to My Account → LLM Providers and add an API key.

Supported providers (LENS picks the model automatically based on task complexity):

Provider	Simple tasks	Complex / critical tasks
Seal AI Private (platform)	Managed for you	Managed for you
Anthropic (BYOK)	claude-3-5-haiku	claude-sonnet-4 / claude-opus-4
OpenAI (BYOK)	gpt-4o-mini	gpt-4o
DeepSeek (BYOK)	deepseek-chat	deepseek-chat
Google Gemini (BYOK)	gemini-2.5-flash	gemini-2.5-pro

Mark one provider as your default and save.

Seal AI Private quota

Seal AI Private includes 5M tokens per calendar month for the whole organization, plus any token packs you've purchased (packs never expire). The org gets an email at 80% and 100% consumption; once quota and packs are exhausted, Seal AI stops with a "token limit reached" message until the monthly reset — the owner can buy a pack from the Seal AI Usage screen, and any user can switch the chat to a BYOK provider for immediate headroom.

Step 3: Ask Your First Question

Try the AI assistant:

1. Go to LENS → Chat
2. Type a question like:
   • "What were my top traffic sources last week?"
   • "How is my conversion rate trending?"
   • "Are there any issues I should know about?"
3. LENS will analyze your data and respond

Tips for Better Answers

Do	Don't
Be specific about time periods	Ask vague questions
Mention specific metrics	Assume LENS knows context
Ask follow-up questions	Ask multiple questions at once

Good: "Why did organic traffic from Google drop last Tuesday?"

Less helpful: "Why is traffic down?"

Step 4: Review Your LENS Reports

LENS produces weekly and monthly reports — an executive summary plus key metrics, channel performance, top insights and action items. Reports are generated for a completed period (there are no custom templates or per-report recipient lists).

1. Go to LENS → Reports
2. Open the latest weekly or monthly report
3. To receive a recurring summary by email, enable email notifications and set the analysis frequency in Settings → LENS (daily, weekly, or manual)

See LENS Reports for what each report contains.

Automated insight detection (roadmap)

Not live yet

The rule-based detection library is built but not active. No detection rule currently runs on customer accounts, and no automated insight or anomaly alert is generated today. This section documents the planned behaviour — see the LENS overview for what is live. What works today is LENS chat (Step 3) and LENS reports (Step 4).

The planned library produces five types of insight, so it won't be only about anomalies:

Type	What it means
Anomaly	An unexpected change in a metric
Opportunity	A potential improvement worth acting on
Trend	A significant directional change over time
Alert	A warning that needs attention
Health	A tracking or data-quality issue

The planned launch set ("LENS Basic") covers traffic drops and spikes, conversion drops, tracking health, device and landing-page performance, channel efficiency, and source-concentration risk — none of which will require e-commerce or microconversion setup.

Once detection ships, an insight will look something like this (here, an anomaly):

🔴 HIGH PRIORITY
Conversion Rate Drop Detected

Your conversion rate dropped 35% compared to the
previous 7-day average.

Current: 1.2%
Previous: 1.85%
Change: -35.1%

Detected: 2026-08-10 14:32 UTC
Affected segment: All traffic

[View Details] [Investigate] [Dismiss]

Insight Severity Levels

Level	Icon	Meaning	Action
Critical	🔴	Immediate attention needed	Investigate now
Warning	🟡	Significant change detected	Review within 24h
Info	🔵	Notable but expected	Acknowledge when convenient

Next Steps

Explore All Rules

See the planned LENS Basic rule library across all 11 categories.

AI Best Practices

Get better answers from the AI assistant.

Troubleshooting

"No insights or anomalies shown"

Expected: automated detection is not active yet (see the roadmap note above), so no account currently generates insights. Use Chat to ask about your data, and Reports for the weekly and monthly summaries.

"Assistant not responding"

The answer depends on which provider you're on — check My Account → LLM Providers to see which one is active.

On Seal AI Private (no API key involved):

1. Your organization may have exhausted its 5M monthly tokens plus any packs — the chat says "token limit reached". Check Organization → Seal AI Usage; an owner can buy a token pack, or any user can switch to a BYOK provider for immediate headroom.
2. Confirm your plan still includes Seal AI (add-on on Growth, included in Scale and Enterprise).

On bring-your-own-key:

1. You've added an API key in My Account → LLM Providers
2. A default provider is selected and enabled
3. The key is still valid and hasn't expired or hit your provider's own rate limit

Email summaries not arriving

Check:

1. Email notifications are enabled in account settings
2. Messages aren't going to spam