The one rule before choosing a visual
Before selecting a visual type, answer one question: what decision or understanding should this visual produce in the reader? Every visual choice follows from the answer. A visual that looks impressive but does not produce a clear decision or understanding is decoration, not analysis.
The six chart types covered here each answer a different analytical question. Bar charts answer "how do these things compare?" Combo charts answer "how do two related metrics move together?" Donut charts answer "what is the proportional composition?" Funnel charts answer "where does volume drop off in a sequential process?" Waterfall charts answer "what drove the change between two values?" Area charts answer "how has a metric trended over time, and by how much?"
The Visualizations pane in Power BI Desktop — every chart type is one click away once your data is loaded.
Bar and column charts
Bar charts (horizontal) and column charts (vertical) are the workhorse of business reporting. They are best for comparing discrete categories — branches, agents, suppliers, destinations, hotel names — on a single numeric measure. The eye can compare bar lengths accurately and quickly, which makes ranking immediately legible.
Use a column chart when your categories are time periods (months, quarters, years) — time reads naturally left to right. Use a bar chart when your categories are names or labels, especially when there are many of them — names read more easily in horizontal layout, and bars extend to the right without running out of vertical space.
In a travel platform dashboard, bar charts appear on the branch-wise page (YTD bookings by branch), the agent-wise page (top agents by business), the supplier page (supplier group bookings), and the destination page (top countries and top cities by performance). In each case the question is the same: which entity ranks highest, and by how much?
Avoid putting more than fifteen to twenty categories on a bar chart. Beyond that, the bars become too small to read and too numerous to compare. Use a TopN filter in the visual to show the top ten or top fifteen, and note the filter in the visual title so readers know they are not seeing all values.
Combo charts — bar with line
A combo chart overlays a line on a bar chart, allowing two related metrics with different scales to share a single visual. The bars use the primary Y axis on the left; the line uses the secondary Y axis on the right. This is the correct visual when you want to show volume alongside growth rate, or bookings alongside conversion ratio.
In a travel dashboard, the combo chart appears on the bookings overview page: bars show the absolute number of bookings per month (YTD, LYTD, LLYTD as grouped bars), while two lines overlay the booking growth percentage (YTD vs LYTD and YTD vs LLYTD). A reader sees in a single glance both the absolute volume trend and the growth acceleration or deceleration.
The combo chart requires careful axis labelling. When two Y axes are present, both must be clearly labelled with their units. A visual with an unlabelled secondary axis forces the reader to guess whether the line represents a percentage, a ratio, a currency amount, or something else entirely.
Avoid using a combo chart when the two metrics are not meaningfully related, or when both are best read as bars. The combo chart implies a relationship between the bar and the line — if that relationship is not real or relevant, use two separate visuals instead.
Donut and pie charts
Donut and pie charts show proportional composition — what percentage of a whole does each category represent? They work when you have a small number of meaningful categories (three to five is ideal) and when the proportional split is the genuinely important insight rather than the absolute values.
In a travel platform context, donut charts appear on the agent behaviour page to show the split between never-booked agents (70.98%) and historically booked agents (29.02%). The proportion is the point — the reader needs to understand that approximately 71% of registered agents have never made a booking, which has direct implications for the platform\'s agent activation strategy.
The donut chart is also used on the payment gateway page to show PG business split across LLYTD, LYTD, and YTD — three time periods, each showing the proportion of business going through payment gateways versus direct. The three donuts placed side by side allow a trend in proportional split to be read across time periods.
Avoid pie and donut charts when you have more than five categories — humans cannot accurately compare slice sizes beyond that number, especially for similar-sized slices. Avoid them entirely when the absolute values matter as much as the proportions — a bar chart communicates both; a pie chart communicates neither.
Funnel charts
Funnel charts show sequential volume reduction through a process — each stage of the funnel is narrower than the one before it, visualising where volume drops off. They are designed specifically for pipeline and conversion analysis.
The classic use case is a sales pipeline: number of leads at the top, qualified leads, proposals sent, negotiations, closed deals at the bottom. In a travel platform context, a funnel chart might show the agent onboarding pipeline: registered agents at the top, agents who have attempted a booking, agents who completed a booking, agents who have booked in the current year.
The funnel chart makes drop-off immediately visible and quantifiable. A stage where 60% of volume drops off is instantly recognisable as the bottleneck in the process — far more immediately than a table of numbers showing the same data.
Funnel charts only work for genuinely sequential processes where the stages have a natural order and where volume always decreases from one stage to the next. Do not use a funnel chart for non-sequential categories or for data where some stages might be larger than preceding ones — the visual convention will mislead.
Waterfall charts
Waterfall charts decompose the change between two values into its contributing factors — showing what drove an increase or decrease. Bars going up represent positive contributors; bars going down represent negative contributors. The starting and ending values are shown as full-height anchor bars.
In a currency analysis dashboard, a waterfall chart shows the top five growing selling currencies — starting from the total baseline, each currency adds or subtracts from the running total, and the final bar shows the result. A reader can immediately see which currencies drove growth and which were drags on the overall total.
Waterfall charts are also powerful for variance analysis: why did business revenue this year differ from last year? A waterfall can decompose that variance into contributions from new agents, lost agents, price changes, volume changes, and currency effects — making the narrative of performance change immediately legible.
Area charts
Area charts are line charts with the area below the line filled in. They emphasise the magnitude of a trend over time — the filled area gives a stronger visual impression of volume than a line alone. They work well for showing cumulative metrics (total bookings year-to-date over time) or for comparing two overlapping time series where the area between the lines communicates the gap.
In a travel platform dashboard, area charts appear on the bookings comparison page showing this year\'s sales versus last year\'s sales by fiscal month. The two filled areas — one for current year, one for last year — overlap, and the gap between them at any point in time represents the year-over-year difference. When the current year area is larger, the business is ahead; when it falls below, it is behind.
Avoid area charts for more than two or three series — overlapping filled areas quickly become visually confusing. When you need to compare four or more time series, use a line chart instead, where the lines remain visually distinct without the overlap problem.
The visual decision framework
Start with the question your visual needs to answer, then work forward to the chart type. Comparing categories of different sizes? Bar chart. Showing volume and growth rate together? Combo chart. Showing composition as a percentage of whole? Donut chart. Showing where a process loses volume? Funnel chart. Explaining what drove a change? Waterfall chart. Showing how a metric has trended over time? Area or line chart.
When in doubt between two chart types, build both and ask which one a colleague who has not seen the data before can interpret correctly within five seconds. The one they interpret correctly is the right visual for your audience.
`