Video |

Tableau | Data visualisation | Data prep | Analytics

Analysing Fuel Economy with Tableau, Workflow & Google sheets

My fuel data showed 12 pence per mile on average, but a spike to 20p told the story of an oil leak.

Tim Ngwena •4 November 2016 •1 min read •3 min watch •1.1K

Watch on YouTube

Connecting Tableau to Google Sheets lets you build a database-style analytics setup for very little cost
Hide unneeded fields and use the custom split function to break a date out of a single column during quick data prep
A true/false premium tag should be converted to a dimension before analysis
Dividing cost by miles driven and adding a formatted currency reference line reveals an average cost per mile (12p) and outliers (20p)
Cost-per-mile spikes can surface real-world issues, like an oil leak that dropped engine performance, while peaks and troughs otherwise reflect city versus motorway driving

0:00Okay. So I'm going to start

0:02this blog post from a slightly different

0:03place.

0:04I'm going to start by showing you the

0:05output.

0:06So I'm going to connect to Google Sheets,

0:08and I'm going to connect to a dataset I've

0:10just been looking at with

0:12my fuel consumption and mileage on it.

0:16So the instant I connect,

0:18I know I need to do some data prep straight

0:20away.

0:20I don't need some of these fields.

0:23These two fields here are created in Google

0:26Sheets.

0:26So I'm actually going to hide them because

0:28we're going to do some of

0:29our own calculations and I'm going to use

0:31the custom split functions just to split

0:33out

0:33one column which has the date in it.

0:36So very, very quickly,

0:37I've done some data prep.

0:38I'm now ready to do some analysis in Table

0:41au.

0:41So the other thing I can do while I'm here

0:44is this premium tag

0:46is like a true or false.

0:49So I'm just going to put it here

0:50because it really should be a dimension.

0:52I'm going to drag out the dates and I want

0:54the exact date so I can

0:56see each day that I actually went to the

0:58fuel pump,

0:58and that's the cost.

1:00So here is a chart showing me the cost,

1:03how much I've been spending at the fuel

1:05tank on each visit over time.

1:07I started doing this in July.

1:09One of the really interesting things is I

1:12've started analyzing

1:13my fuel economy and therefore the

1:16performance of my car,

1:18and it's been really interesting.

1:20So let me just give you a very, very quick

1:22example.

1:22So I'm just going to divide the cost by the

1:26number of miles driven.

1:28What you can see here is if I just drop a

1:32reference line into the chart,

1:34with the value, and we actually format this

1:39so that this is a currency.

1:41You can see that the cost on average per

1:45mile that I drive is 12 pence.

1:49So that's 12 pence per mile.

1:50But up here in the 5th of September is

1:52actually 20p per mile.

1:54So what happened?

1:56Essentially, I had an issue with my car.

2:00If I just bring out the miles driven

2:03underneath that,

2:04you can see the actual day where I had the

2:06most miles driven was late September.

2:09I actually had to drive to collect my dog

2:13from Scotland.

2:14So here we have a very,

2:18very standout piece of data.

2:21What actually happened here is I had an oil

2:22leak.

2:23So the engine performance dropped basically

2:26through the floor.

2:27I was able to rescue that,

2:30but it was costing me eight pence per mile

2:32more than normal compared to my average.

2:35You can see it tends to stay roughly the

2:37same.

2:38These peaks and troughs are to do with the

2:41type of traffic I'm driving in.

2:42So if I do a lot of city driving, it's up

2:44high.

2:45If I do a lot of motorway driving,

2:46it tends to be just below 12 pence.

2:48So it's very easy for me to analyze this.

2:51This was all done because Workflow,

2:54Google Sheets, and Tableau were able to

2:57work together.

2:58For very, very minimal cost,

3:00I have the kind of setup most companies

3:03have in a database or data warehouse.

3:06Of course, with the new functions in Table

3:08au 10,

3:09you can do lots, lots more with that.

3:11So that's a very quick insight.

3:13Read the rest of the blog posts to find out

3:15more about how I've actually done this,

3:17and I'm able to do this kind of analysis.

3:19I'll see you in the next one

Rollup

Tim shows how connecting Tableau to a Google Sheet of everyday data (his car's fuel purchases) can surface real-world problems, like an oil leak, purely by tracking cost per mile over time.

He's logging fuel cost, miles driven and a few other fields in Google Sheets each time he fills up. The goal is to turn that log into a running view of fuel economy without building a proper database.

The Breakdown

Connect Tableau to Google Sheets 0:00

You can connect Tableau directly to a Google Sheet and treat it like a live data source, giving you a database-style analytics setup for very little cost.
Hide fields you don't need 0:20

As soon as you connect, do a quick pass of data prep: hide any fields that were only used for calculations inside the sheet itself, since you'll rebuild what you need in Tableau.
Convert true/false tags to dimensions 0:42

A field that behaves like a true/false flag (his was a 'premium fuel' tag) should be converted to a dimension before you start analysing it, otherwise it won't behave the way you expect in the view.
Calculate cost per mile and add a reference line 1:20

Divide cost by miles driven, then drop in a reference line formatted as currency to show the average — this is what turns raw spend into a comparable, standardised metric.
Spot and investigate outliers 1:42

With the average line in place, spikes above it stand out immediately; cross-referencing the date and miles driven for that spike let Tim work out what actually happened on that day.
Interpret the pattern behind the numbers 2:21

The average cost per mile stayed fairly stable, with normal peaks and troughs explained by city versus motorway driving — but one sustained spike (from 12p to 20p per mile) pointed to a real mechanical fault, an oil leak, rather than just driving conditions.

Worth Knowing

The oil leak wasn't found by inspecting the car first — it was flagged by the data, then confirmed afterwards.
Ordinary variation in cost per mile is expected and tied to driving type, so you need a stable average and reference line before you can tell a genuine anomaly from normal noise.

Use It When

Reach for this setup when you want lightweight, ongoing analysis of personal or small-scale operational data (fuel, expenses, usage logs) without setting up a full database — a spreadsheet plus Tableau is enough to catch real problems early.

How this Rollup was made provenance & method

A Rollup is drafted by AI from the video's transcript, then reviewed and edited by Tim. Everything used to produce this one is listed below — the model, the exact prompt, and the source video — so the process is transparent and reproducible.

Transcription: On-device — NVIDIA Parakeet v3 for recent videos, OpenAI Whisper large-v3 for earlier ones. The transcript never leaves the machine or gets published.
Drafting: Claude Sonnet 5 in the cloud, from that transcript.
Prompt: The exact Rollup prompt (v2) — the full system prompt, unedited.
Source video: Watch on YouTube
Drafted: 5 July 2026 at 09:38
Reviewed & edited: 5 July 2026 at 09:41 · by Tim Ngwena

Model + prompt + video is everything you'd need to recreate a Rollup like this yourself. The one thing we don't share is the transcript.

Rights. The video and its transcript are the property of TN Media Ltd. Unauthorised use or download is prohibited. © TN Media Ltd.