Building charts in plain English

The fastest way to build a chart in Dotwave is to describe what you want to see. Ask-your-data takes a plain-English question and turns it into a chart spec — you don't pick axes, aggregations, and groupings by hand unless you want to. This article explains how the question-to-chart flow works, which questions tend to produce good results, how to refine a chart after it is generated, how to add it to your dashboard, and how the chart coach and annotations help you communicate more clearly.

How ask-your-data works

Inside a dashboard, you type a question about the cleaned dataset the dashboard is built on. Dotwave reads your question against the dataset's columns and generates a chart spec: it chooses a chart type, the fields for each axis, and any aggregation needed, then renders the chart. Because the dashboard draws from cleaned data, the answer reflects the data you have actually vetted rather than the raw upload.

You are not locked into whatever it produces first. The generated chart is a starting point you can accept, adjust, or discard. Ask another way and you'll often get a cleaner result — the phrasing of your question genuinely shapes the output.

Questions that work well

Clear, specific questions produce the best charts. Some reliable patterns:

• "revenue by month" — a measure broken down over a time dimension, which becomes a trend chart.
• "top 10 customers by total spend" — a ranked comparison, which becomes a sorted bar chart limited to ten rows.
• "orders per region" — a count of records grouped by a category.

The common thread is that each question names a measure (revenue, spend, order count) and a dimension to break it down by (month, customer, region). When you give Dotwave both halves, it has everything it needs to build the right chart.

Adjusting and adding the chart

Once a chart is generated, you can refine it before committing. Change the chart type, swap or add fields, adjust how a value is aggregated, or fix the sort and limit until the chart says exactly what you mean. When it looks right, add it to the dashboard so it takes a permanent place on the board alongside your other charts. From there it behaves like any other dashboard chart — it recomputes when the underlying cleaned data changes.

The chart coach

Sometimes the chart type you first reach for isn't the clearest way to show your data. The chart coach watches for these cases and suggests a better chart type when appropriate — for instance, nudging you away from a pie chart with a dozen slices toward a bar chart that ranks them legibly. The suggestion is advice, not an override: you decide whether to take it. Over a full dashboard, following the coach's prompts tends to produce a set of charts that are consistent and easy to read at a glance.

Annotations

Charts rarely speak entirely for themselves. Chart annotations let you add context directly onto a chart — calling out a spike, marking when a campaign launched, or explaining an anomaly so a client doesn't misread it. Annotations turn a bare chart into a narrated one, which is often the difference between a stakeholder understanding your point immediately and emailing you to ask what they're looking at.