Schedule Your Free Demo

See how Shilte AI can automate your enterprise operations.

Operational Intelligence Tutorial | Understand Business Analytics with Shilte AI
Shilte AI
Learning Centre / Operational Intelligence
Shilte AI Learning Centre · Business Insights

Operational Intelligence

Understanding your business analytics, sales and performance. Learn how to read the activity and metrics available in your business workspace, recognize patterns, ask better questions, and use information to guide your next action.

UnderstandKnow what a metric is telling you—and what it isn't.
CompareLook for changes across meaningful time periods.
ActTurn observations into practical follow-up.

1. Start here

Operational intelligence is a practical habit of using business information to understand what's happening and decide what to do next. A dashboard can bring activity into one view, but useful insights come from asking focused questions and interpreting the data in context.

Before beginning, sign in to your Shilte AI business workspace and familiarize yourself with the dashboard areas available to your account. Your view may vary depending on enabled features, integrations, activity and workspace setup.

Learning outcome: By the end, you'll be able to describe a metric, compare it thoughtfully, identify a question worth investigating, and choose an appropriate next step.

2. What is operational intelligence?

Operational intelligence is the process of examining everyday business activity—such as inquiries, leads, sales, orders, customer interactions and workflow progress—to gain a clearer picture of business operations.

Business activity

What has happened? Review records, transactions, customer interactions or tasks.

Performance signals

What is changing? Compare counts, rates, volumes or time periods where available.

Questions

Why might the change have occurred? Consider timing, campaign activity, product mix or process changes.

Action

What should you check, improve, follow up or test next?

Observe→Compare→Ask why→Take action→Review

Analytics describe the activity recorded by your systems. They are useful signals, not a complete explanation of your business or a guarantee of future outcomes.

3. Getting oriented in your dashboard

Start with the dashboard overview. Take a moment to identify the sections, date ranges, summaries, charts, lists and detail views that are available to your account.

  1. Sign in to your Shilte AI workspace.
  2. Open the main dashboard or analytics area.
  3. Identify which numbers are summaries and which sections open detailed records.
  4. Check whether date filters or other filters are available, and note the selected period.
  5. Open a relevant record to understand what data supports a summary, where that capability is available.
Tip: Don't assume every dashboard card uses the same date range or calculation. When available, check the metric's description, filter state or underlying records before comparing values.

4. Understand key business metrics

Metric names can differ between systems and workspaces. Use the descriptions below as a guide to common business measures, not a claim that every metric is displayed in every Shilte dashboard.

01

Revenue or sales value

The value of sales recorded during a period. Confirm whether the figure reflects completed, paid, refunded, cancelled or otherwise adjusted transactions before treating it as revenue.

02

Orders or transactions

A count of order records or purchases. Clarify whether the count includes pending, failed, cancelled or duplicate records.

03

Leads captured

People or potential customers whose information has been recorded. A lead count is not the same as qualified leads, active opportunities or sales.

04

Customer engagement

Interactions such as conversations, messages or inquiry activity. Volume alone does not show whether customers were satisfied or whether a conversation resulted in a sale.

05

Conversion rate

A proportion of a defined group that completes a desired action. You must know the numerator, denominator and time window to interpret it correctly.

06

Average order value (AOV)

Often calculated as sales value divided by the number of included orders, using consistent definitions and periods. Verify which transactions are included.

Metrics are only comparable when their definitions are consistent. If one period includes pending orders and another excludes them, the comparison may be misleading.

5. Reading sales and order activity

Use sales and order views to explore transaction activity. Rather than looking only at a total, examine the records and ask what may explain the result.

Questions to ask

  • How many orders or transactions were recorded in the selected period?
  • Which items, services or categories appear most often, if product detail is available?
  • Are there pending, failed, cancelled or refunded transactions that affect the total?
  • Are changes concentrated on certain days or around promotions?
  • Does the order count tell the same story as sales value?

Example: sales value increased

Check whether more orders were placed, customers purchased higher-value items, or a few large transactions influenced the total. Don't assume the increase is caused by a campaign without evidence.

Example: orders increased but value did not

Explore average order value, product mix, discounts and order status. More transactions do not necessarily mean higher revenue.

6. Understanding lead performance

Lead data helps you see whether customer interest is being captured and followed up. A useful lead review considers quality, response and progression—not only the number of records.

  1. Review the new leads recorded in your selected period.
  2. Check available contact information and source details.
  3. Identify leads that need a response or have not been updated recently.
  4. Review status or follow-up notes where available.
  5. Look for recurring inquiries that could inform your marketing, product pages or knowledge base.
A high lead count can be encouraging, but incomplete details, duplicate records, low intent or delayed follow-up may reduce the practical value of those leads.

7. Learn from customer conversations

Customer conversations can reveal needs that don't appear in sales totals alone. When conversation records or analytics are available and your access permits, review recurring questions, points of confusion, product requests and issues that require human help.

Questions customers repeatedly ask

Could product details, pricing, delivery information or business policies be easier to find?

Where customers get stuck

Are there recurring misunderstandings or steps in the buying journey that need clarification?

Human hand-offs

What types of requests need staff attention, and is the follow-up process clear?

Feedback and requests

Are customers asking for products, features or services you may want to evaluate?

Use these themes to improve your business information, website content, support process or customer experience. Protect customer privacy when reviewing or sharing conversational data.

8. Identifying patterns and trends

One data point is a snapshot. A pattern is a repeated or sustained observation. If your dashboard supports comparisons, use the same metric definition across similar periods and look for changes that deserve further investigation.

Simple comparison method

  1. Choose one metric and define what it measures.
  2. Choose two comparable periods—for example, this week and the previous week, if complete periods are available.
  3. Compare the values and calculate the change if useful.
  4. Consider contextual factors such as seasonality, campaigns, stock availability, holidays, staff changes or tracking changes.
  5. Record a plausible explanation as a hypothesis, then check for evidence.

Illustrative example

If inquiries rose from 40 to 60, that is a 50% increase in inquiry volume. It does not automatically mean sales increased by 50%. Investigate the lead quality, follow-up and eventual conversion.

Be careful with small sample sizes. A change from 2 to 4 transactions is a 100% increase, but the absolute number is small and may not indicate a sustained trend.

9. Turning insights into business action

Analytics become useful when they inform a specific, manageable next step. Avoid changing many things at once if you want to learn which change had an effect.

ObservationQuestion to investigatePossible next action
Many inquiries, few purchasesAre customers missing important information or encountering friction?Review FAQs, pricing clarity, checkout experience and follow-up.
Leads not being contactedIs ownership or follow-up timing unclear?Assign responsibility and set a repeatable follow-up routine.
Repeated questionsIs business information incomplete or difficult to find?Improve the knowledge base or website content, then test the chatbot.
Sales value dropsDid order volume, product mix, availability or transaction status change?Review order records and verify the cause before acting.
Customer complaints repeatIs there a process issue or recurring expectation gap?Investigate the issue, update guidance and monitor whether it recurs.
Try a small improvement. Make a clearly defined change, note when you made it, and review the relevant metric over an appropriate period. This makes it easier to evaluate whether the change may have helped.

10. Create a simple analytics review routine

You do not need to study every metric every day. A consistent review cadence can help you notice issues and avoid reacting to random fluctuations.

Daily · 5–10 minutes

  • Check new inquiries and leads.
  • Review urgent customer follow-ups.
  • Look for unusual order or payment activity.

Weekly · 20–30 minutes

  • Review sales and order activity.
  • Compare leads and follow-ups.
  • Identify recurring customer questions.

Monthly · 30–60 minutes

  • Compare broader trends.
  • Review process improvements made.
  • Choose priorities for the next month.

Record decisions

  • Write down the observation.
  • Note the action taken and date.
  • Set a time to evaluate results.
My business review checklist
  • □ I checked the time period and filters.
  • □ I understand how the metric is defined.
  • □ I examined supporting records when needed.
  • □ I identified one actionable observation.
  • □ I assigned a follow-up or recorded a decision.
  • □ I will revisit the result later.

11. Interpret analytics responsibly

Dashboards are only as reliable as the data collected, how it is processed and how the metric is defined. Missing events, duplicate records, delayed updates, integration gaps or status differences may affect what you see.

  • Don't confuse correlation with cause: two things changing together doesn't prove one caused the other.
  • Check completeness: activity may not be recorded if a system or integration was unavailable.
  • Understand exclusions: cancelled, pending, refunded or failed transactions may be treated differently across reports.
  • Respect privacy: share reports only with people who should see the data, and avoid unnecessary exposure of personal information.
  • Validate important decisions: use source records and business context before making financial or operational decisions.
Important: Analytics are decision-support information—not a substitute for financial reconciliation, professional accounting, or careful business judgment.

12. Troubleshooting and common questions

My dashboard shows zero or no data

Confirm you are in the correct business workspace, check whether activity has been recorded, review any available filters or date ranges, and verify that relevant integrations or features are configured.

The figures don't match my records

Check what each report includes, its time range, transaction statuses, refunds, duplicates and any reporting delay. Reconcile against the source system before relying on a total.

Metrics changed unexpectedly

Review the comparison periods and filters, then investigate recent promotions, product availability, tracking or integration changes, and unusual activity.

I can't find a particular analytics feature

Dashboard sections can vary by workspace, plan, release and enabled capability. Consult the current dashboard or contact Shilte support for confirmation.

How can I make analytics more useful?

Keep records accurate, review information consistently, define the business question before looking at charts, and follow up on insights with small measurable improvements.

13. Frequently asked questions

What is operational intelligence in Shilte AI?

It is the practice of using recorded business activity and available analytics to understand operations, notice trends, and guide follow-up or improvements.

Do dashboard metrics update in real time?

Update timing depends on the data source, feature and processing behavior. Don't assume every chart or integration refreshes instantly; check the current dashboard behavior.

Does more chatbot conversations mean more sales?

Not necessarily. Conversations can represent support requests, casual inquiries, repeated messages or purchase intent. Review the quality and outcomes of conversations in context.

Which metric should I focus on first?

Start with a business goal. For example, if your priority is customer follow-up, begin with leads and follow-up status. If your priority is orders, review transaction counts and sales value with consistent definitions.

Can Shilte AI tell me exactly why sales changed?

Metrics can highlight what changed, but the cause usually requires investigation and context. Treat explanations as hypotheses until supported by evidence.

How often should I review analytics?

Use a cadence that fits your activity. A short daily check for immediate follow-up, a weekly review of performance, and a monthly assessment of trends is a practical starting point.

Is analytics data a replacement for accounting?

No. Use operational dashboards to understand activity, but rely on appropriate financial records and reconciliation for accounting, tax and reporting.

Video walkthrough

Use this section for a future screen-recorded walkthrough showing where to find the dashboard metrics and how to perform a weekly review.

▶
Operational Intelligence video tutorial

Video walkthrough coming soon

Replace this placeholder with your published video embed when ready.