Tines Academy › Lesson
Start with the decision
Help Maya in customer support turn her team’s questions into a clear metric, then build the first chart.
Objective: Identify who a dashboard is for and what they need to decide, define a metric that answers their question, and choose only the data and access it needs.
Where you’re starting
Maya leads a customer support team that uses a ticket system like Plain or Zendesk. After a ticket is closed, customers receive a short survey. They rate the help they received as positive, neutral, or negative, and can leave a comment.
Every Monday, Maya and her two team leads decide which problems need attention and what to raise with the product teams. To prepare, she sorts the previous week’s ratings by hand in a spreadsheet. Her team leads also use the feedback during the week to follow up with customers who left negative ratings.
In this learning path, you’ll build a dashboard that helps them spot trends, see which product areas need attention, and find the feedback behind the numbers.
Where you’re headed
With a full brief, Tines 3B could build Maya’s dashboard in one go. We’ll take it one stage at a time, so you can practice describing what you need and checking that the numbers answer your questions.
As you build, you’ll control who can see the data, check the numbers, and decide who keeps it current. These habits help people use the dashboard without exposing information or relying on misleading results. Each lesson ends with a Build it safely note about a habit you’ve practiced.
First, you’ll work out what Maya’s team needs to know, then build the first chart. By the end of the path, you’ll have a dashboard that looks something like this.
Key terms
Common questions a dashboard can answer
Turn each question into a metric
Ask two people for a “customer satisfaction” number, and one might give you last month’s percentage of positive ratings while the other counts this year’s negative ones. Both use the same data, but they answer different questions.
To turn a question into a metric, you need to decide what to measure and which period it covers.
Maya wants to know whether negative ratings are becoming more common. She’ll measure the percentage of survey responses rated negative each week, counting each response in the week the customer submitted the rating. Comparing those percentages shows how the feedback changes over time.
Before you start building:
Open your Tines 3B tenant. If you don’t have one yet, sign up for one.
Choose a space where you can build. If you’ll use a ticket system connector, check that it’s available there and that the space’s members are allowed to see the saved data.
Use approved survey responses or made-up data. For made-up data, choose Without a connector wherever it’s offered.
Keep the workflow as a draft while you work through the path. You don’t need to publish it to complete the learning activities.
Give the connector only what it needs
A connector stores the credentials for your ticket system. Tines 3B adds them to outgoing requests, so the workflow’s code doesn’t need to handle the password or token. Never paste passwords or tokens into chat; they’re saved with the conversation and could end up in generated code.
Before using a connector, check what it can read and who looks after it. Ask whether its permissions can be limited to the information the dashboard needs. Maya’s dashboard reads survey responses; it doesn’t reply to customers or change tickets.
Be selective about what you load and save. Anyone who can open the workflow can also look through its saved data. Maya’s team follows up in the ticket system, so the dashboard doesn’t need customer name or email fields. Check the comments too, and remove any personal details before using the data in your practice dashboard.
Before loading real responses, check that you’re allowed to use them here. If you’re unsure, use made-up data.
Connect your ticket system
Complete this setup if you’re using real survey responses.
Open Connectors and find a connector for your ticket system. Check that you can use it in your chosen space.
If there isn’t a suitable connector, create one if you can, using only the read access the dashboard needs. Otherwise, ask whoever manages connectors to help. You can continue with made-up data while access is being arranged.
Check the connection status. If it shows “Error,” read the details and ask the connector’s owner for help. A successful connection test doesn’t prove it can read every field you need.
Don’t remove people from a space or widen connector access just to finish the exercise.
Load the data and build the first chart
From Home, open the space where you’ll build.
Choose the prompt for your setup. For real data, tell Tines 3B whether comments are approved to load into this space. You can give it the connector name now or when it asks.
Paste the prompt into the chat box and send it to create the workflow. If you’d like to review a plan first, turn on Plan mode before sending it.
Build a support feedback dashboard for our support team. Start with the data and one chart, and keep it as a draft.
Use [your ticket system connector] to load survey responses from the last 12 full weeks and the current week so far. Keep each response's source ID if available, ticket ID and real link, rating date, rating, product area, and channel. Include comments only if I confirm they're approved for this space. Leave out customer names and email addresses. Only read from the ticket system; don't change tickets or message customers.
Confirm which ratings count as negative and what identifies one response. If a ticket can have several legitimate responses, ask me how they should be counted. Use the source report's time zone and week boundaries, and ask me if those aren't clear.
For each of the last 12 full weeks, chart negative responses as a percentage of all responses with a valid rating. Show both counts and label the reporting dates and time zone. Group by rating date, not ticket creation date. Save the current week's responses for later views, but leave them out of this chart. If a week has no responses with a valid rating, show the percentage as unavailable rather than 0%.
Keep the page and any routes supplying its data private to this space. Report how many responses you loaded, the fields kept, how responses were counted, and who can open the page and inspect the saved data.
Check that the chart includes the last 12 full weeks and excludes the current week. For the last full week, show the negative count, total valid-rating count, and calculated percentage, with the source records or report used to check them. State the dates, time zone, filters, and counting rule used.
Report actual test results and anything you couldn't verify. Record unfinished checks in the workflow's Readme.
Build a support feedback dashboard for our support team. Start with test data and one chart, and keep it as a draft.
Make up 300 survey responses across the last 12 full weeks and the current week so far. Give each response a unique fictional ticket ID, rating date, positive, neutral, or negative rating, comment, product area, and channel. Use one response per fictional ticket and no real personal details.
Use "Billing," "Account access," "Shipping," and "Mobile app" as product areas, and "email," "chat," and "phone" as channels. Include every area and channel, including some "Billing" responses from chat. Include a few responses with no product area and a few with two product areas. Label the page "Test data," and don't invent external ticket links.
Use Monday-to-Sunday weeks in UTC. For each of the last 12 full weeks, chart negative responses as a percentage of all responses with a valid rating. Show both counts and label the reporting dates and time zone. Group by rating date, not ticket creation date. Save the current week's responses for later views, but leave them out of this chart. If a week has no responses with a valid rating, show the percentage as unavailable rather than 0%.
Keep the page and any routes supplying its data private to this space. Report how many responses you generated, the fields kept, and who can open the page and inspect the saved data.
Check that the chart includes the last 12 full weeks and excludes the current week. For the last full week, show the saved test responses, negative count, total valid-rating count, and calculated percentage.
Report actual test results and anything you couldn't verify. Record unfinished checks in the workflow's Readme.
Check the result
Read the build summary in workflow chat. Find the number of responses loaded or generated, the fields kept, how responses were counted, and who can access the page and saved data.
Notice which checks have supporting results and which remain unfinished.
Check the reporting period
Open the page in the Browser tab.
The chart should:
Show the last 12 full weeks, excluding the current week.
Label the reporting dates and time zone.
Show each week’s negative count, total valid-rating count, and percentage.
A week with no valid ratings should show an unavailable percentage, not 0%.
Check one week’s numbers
Use the last full week.
With a connector, compare the chart with a source report or source responses. Make sure both use the same rating dates, time zone, filters, and counting rule.
Without a connector, compare the chart with that week’s saved test responses shown in the build report.
Check the negative count and total valid-rating count, then calculate the percentage of negative ratings.
Only compare ticket counts with response counts when there is one counted response per ticket.
Read the chart as Maya would
Can you tell what the percentage measures and which dates it covers without reading the build summary?
If not, ask Tines 3B to clarify the chart’s labels.
If something doesn’t match
Share what you expected and what you saw in workflow chat. Ask Tines 3B to investigate, then repeat the affected check after a correction.
Ask it to record unresolved checks in Readme.
You’ll keep building Maya’s dashboard in the next lesson. Alongside it, you’ll answer a few questions about a dashboard you’d like to build for your own work.
At the end of the path, we’ll use your answers to create a brief you can review and give Tines 3B. Use the same idea throughout, and leave out passwords, tokens, and personal or confidential details.
What to do if something still feels fuzzy
You worked out who Maya’s dashboard is for, turned her first question into a metric, and built a first chart using your team’s survey responses or test data. Next, you’ll look for missing or inconsistent details and decide how to handle them before they affect the dashboard’s numbers.
Build it safely: You learned why the dashboard only needs read access to the ticket system. If you used a connector, you checked that its permissions match that job.