See it in action

He did the maths: six hours in the chair plus twelve hours filming isn't a schedule, it's a shutdown, and the film would suffer.

So he didn't demand "get it to an hour." He asked for three. Then two, and by the end of the shoot they were down to one hour.

Same team. Same skill. Someone just kept asking "what's stopping us going faster?"

That's margin in a nutshell. Even the best in the world are sitting on slack they haven't found yet. You just have to keep asking, and have someone who knows what data is needed to improve.

One of them was the hairnet check. Every hour, the team downed tools to check their hairnets in the one mirror in the factory. That lost time was costing £12,000 a year.

The simple change was a small mirror at each table, for about £10.

Nobody signed off on "let's lose 12 grand checking hairnets." It just accumulated, process by process, until it was normal. This happens a lot when compliance is layered onto production and nobody checks.

This is why I time every part of production when I analyse margin: to give a true cost of every action, and then ask, can we do it a different way?

One supermarket contract was over 80% of our sales. One supplier was the only one in the country who could supply the exact meat that contract needed. Two calls, in either direction, and we were finished.

So we pivoted into a chain of shops, not just for growth but for insulation. More customers, more suppliers, more phone calls standing between us and disaster.

Margin analysis isn't just where the money's made. It's also risk analysis: which products, which suppliers and which single customers could end you if they walked, and what that would actually do to the business.

Those (amazing) systems are built to run the business, not to let you play with the data how you want. So the data sits there, technically "available", practically untouchable.

I use Mammoth to get past that. Pull it out, reshape it, actually interrogate it, without touching the source systems or waiting on IT.

So, are you getting the best out of your data, or is it stuck behind walls you've just stopped noticing?

Group by invoice size and you'll spot which order sizes quietly bleed margin. Group by stock age and you'll see who's costing you storage. Group by lines ordered and you'll see who could be sold more. Each cut shows a different margin story that the raw numbers hide.

The skill isn't finding the one perfect group. It's designing several, running your data through each, and letting the patterns show you where the wins and losses actually are.

Then automate it, so you build it once and benefit for years. That's how you grow more profitably.

The bins tell you which products are causing waste, how wasteful the team is, and what's being hidden from you.

What doesn't sell? Check the bins. Which product has production errors? Check the bins. Hidden mistakes, overused lines? Check the bins.

When I do a margin review I use a waste costing system to bring this out, and to show clients their true margin, not the margin they wish they had. So yes, in short: I open up a load of bins and weigh the waste.

You take a low-margin, high-hassle product (food) that's a nightmare to deliver, gets complaints and has waste, and instead of trying to squeeze more out of it, you lower the price.

That creates goodwill. Expectations drop, and you get thanked for selling a basic product.

But when customers come to an event they have a budget in mind. Save them $10 on food and it doesn't go back in their wallet. They head to the merchandise stand, where the margins are better, the waste is lower, the stock is easier and the emotional value is higher.

Margin per person goes up, and you get great press on top. A margin masterclass.

In the lead-up to the 2019 Ashes, Broad wasn't in great form. He sat down with his coach and analyst to dig into the numbers, and their theory was quickly confirmed: he was letting batsmen leave too many deliveries. At his best, only 15% of his balls went untouched. When that crept above 30%, his impact faded.

So he focused on making batsmen play more, forcing errors. Left-handers had long been his weakness, so he studied the data and switched to bowling around the wicket. By 2019 he was doing it 93% of the time, and dismissed David Warner seven times in a single Ashes series.

He stayed true to himself, bail-switching antics and all, but listened when the numbers told him to change or crash. That's how to use data in a business too.

Retail is all about getting the most out of your shelves (stock turnover, margin and sales), because you only have a finite amount of selling space and it has to pay for the business.

Benchmark the shelves and you'll find which are the most profitable, and can make changes accordingly.

You may also find products that will sell anywhere in the shop. For us it was milk and eggs, so you don't want them in the best positions.

They can cut software costs, let you push your data harder, give you control of the problems unique to your company, and they're built for you. What's not to love?

Yet I rarely see anyone using them or talking about them. Is it a dull subject? Are we all using them and just not saying so? Or am I in love with the MP3 player when there's an iPhone option?

It was delicious. It was big in America, and the UK follows US trends. It came with a skewer so it cooked in under 18 minutes. It was a cheaper option than a roast, it had no competition so we could set a price with a good margin, and it was vac-packed with an 11-day shelf life for more stores and cheaper distribution. It could be made in half the time of our burgers and sold at double the price, its tall, bold pack took up less shelf space but stood out, and it fitted our production process, so it made the burger lines more profitable too.

It was a hit in Waitrose, and after three months of trials we went up to 180 stores, with other supermarkets keen. To make it more profitable we needed to switch the skewers from UK-made (40p each) to China-made (8p each), taking the delivered margin to 44%. The catch: we had to order 100,000 at a time. So we took the risk, with a two-month lead time.

Then, with our skewers still at sea, the horse meat scandal broke. Suddenly a product called ‘meatloaf’ seemed a lot less appealing, our buyer's enthusiasm (quite rightly) waned, and we had cash tied up in 100,000 skewers.

You can't predict a horse meat scandal, but you can question it when you increase your risk to save margin. We should have looked harder at whether we needed the skewers, or tried alternatives, but we chased the cost cut as the answer.

We did sell all the skewers, over five years. Two lessons. First, margin isn't just about cutting costs. It's also why your ERP shouldn't be your margin control: it can report what you made, not the margin you didn't. Second, skewers can be used for so much. We were sent photos of them as tent pegs, gate pins, note holders and, eleven years on, my nephew's art project.