Periodically the 1% gains theory gets mentioned by some business guru or other. The story they usually tell to evidence the concept is the tale of Dave Brailsford, the sports coach who took British cycling from being about as much of a joke in the cycling world as British cooking is to a French person, to a slew of Olympic gold medals and Tour de France wins.
Cutting a long story short, he did this over a period of time by making tiny improvements in hundreds of different aspects of how cyclists prepared for, and rode in, their events. Shoes that were one ounce lighter. Helmets that were 1% more aerodynamic. Bikes with only one coat of paint instead of two. That sort of thing.
Over time, all those 1% gains added to other 1% gains and the compounding effect took the UK to the very top ranks of the world of cycling from pretty much ground zero a few years earlier.
Now, Dave Brailsford clearly did a fantastic job and has rightly been lauded ever since for his approach.
So much so that all manner of business gurus have picked up on the concept and incorporated it into their own teaching.
Like a lot of things in business, it’s entirely possible to have something which – at its core – is valuable insight but which, applied unthinkingly at scale, can have the exact opposite effect and end up making everything worse.
Some of the more zealous “1% gains” crowd fall into this trap.
Your business is not an enclosed system
The biggest difference between improving the performance of cyclists and running a business is that in the former, you are in control of pretty much all the elements you are playing with. In a business you’re in control of only a fraction of all the elements you’re playing with…and not even a terribly large fraction, at that.
A cycling coach is trying to get a single rider (usually) from Point A to Point B in the shortest amount of time possible. They know, or can easily find out, key aspects like the surface my rider will be cycling on, whether there are any hills and the precise gradient of each of them, the points where a cyclist is likely to be cycling into the prevailing wind or has the prevailing wind at their back, and so on.
Now, there are some variables at play here – the weather on the day of the race, the talents of the rider (some are better at climbing hills, others are better on the flat), the pace of development in the field of carbon fibre technology for making bike frames, and so on.
But a lot of those can be selected for – if you’re choosing a team to compete in the UK in October, select your best wet weather riders.
If the race has a lot of hill climbs in it, choose people who are good at hill climbs. And so on.
And, unless you’re very lucky, the pace of developments in carbon fibre technology impact all the teams equally as scientific developments tend to become widely known pretty quickly. So you need to keep up with those…and you might occasionally be slightly ahead of another team, or another team might occasionally be slightly ahead of you…but as long as you keep up with industry developments, that’s unlikely to be a major source of over- or under-performance.
So, once you’ve aimed off for all those “big picture” issues, you’re left with the aspects which are essentially in that 1% gains territory, and which are all under your control.
You have a closed system to work with. Selecting the fabric for riders’ jerseys is entirely up to you. Likewise choosing their shoes. And subtler aspects such as coaching them to hunch their bodies over the handlebars in the optimum aerodynamic shape.
And it’s fairly easy to change any or all of those aspects – and the hundreds of other similar factors in a cyclist’s performance – to see where it gets you. If it makes a difference, keep doing it. If it makes no difference at all, stop doing it.
Since a cyclist’s performance is, in large part, down to how effectively they use their muscle power to propel a bike of a known weight as quickly as possible from A to B along a pre-determined route, it’s very clear that anything you can do to improve their muscle power or reduce the weight they have to propel along the road…even by a tiny amount…is worth doing.
In cycling races the margin of error is tiny. In the Tour de France – a race of over 2,000 miles – the smallest winning margin between first and second place was just 8 seconds.
When gold and silver medal are separated by small a difference as that, even taking a coat of paint off a bike, over the course of 2,000 miles or so, really can make the difference between a win and a second place.
What about your business?
Contrast the job of a cycling coach with someone running a business – even a very small business.
Despite what you might think, you control very little in your business environment.
Governments implement ideas that increase your costs whether you like it or not. Customers decide to buy from someone else, no matter how long they’ve bought from you. Your staff – or the better ones, at least – can get another job at any time for more money than you’re paying them…no matter how much you pay them.
And that’s before we factor in all the companies in your supply chain, including companies you don’t know anything about because they’re suppliers for one of your suppliers. Or interest rates which you can’t control. Or changes in accounting policies about how you report your profits.
Not to mention changes in the external business environment, like new competitors discovering a different technology that does what your business does at a fraction of the cost. Or new competitors from overseas entering your market. Or one of your biggest customers going bankrupt.
Compared to a cycling coach, the number of aspects of their business a CEO can truly control – no matter how intelligent and hard-working they are – is tiny.
And that’s why, like a lot of ideas implemented into businesses from a sports or science/engineering background just don’t work as well in practice as the theory suggests.
Even a theory as superficially attractive as the 1% gains approach pushed by those business gurus.
The laws of statistics are against you
While it’s far from the only reason, one reason the 1% gains theory doesn’t work that well in most businesses is that the laws of statistics are against you.
When you’re coaching cyclists, you can attach all sorts of monitoring equipment to them while they ride their bike on a simulator on a lab somewhere. If you’re trying to help them get off the starting line faster, you will probably practice that manoeuvre hundreds of times and you’ll have thousands of data points about those crucial first couple of seconds of a bike race to work with.
In that environment, statistics can be helpful because that’s a science that only works with large numbers of data points.
In many businesses, you might not make hundreds of sales in a month or even in a year, especially if you’re in B2B. With cycling training, you could practice the first two seconds of 100 starts in a morning.
Sure, you can try out packing box B instead of packing box A and see if that slightly cheaper option is strong enough to protect your products all the way to your customer to the same standard as your original packing materials.
But if box B survives the first journey, what do you learn from that? Comparatively little – you might have just been lucky, maybe it wasn’t raining that day, or perhaps your driver took a little extra care because they could see there was something different about that box, even if they weren’t exactly sure what the difference was.
Now, at this point, people I talk to like this often leap to the conclusion that I’m going to tell them to collect lots more data points so they can use statistics to draw statistically-valid conclusions.
That’s almost never the right course of action.
It either means taking a massive risk by going “Big Bang” on a new, untried, untested theory, in the hope of collecting the amount of data required to assess whether that was a good idea or not. This is rarely a sensible way to manage business risk.
Or it means spending a small fortune on either people or systems to collect, track, and report on data which will be mostly a waste of time. Through an unscientific process of trial and error, odds are you’ve iterated your way to a pretty reasonable way of working by now, at least most of the time. So the number of times you’ll stumble across a huge, previously undiscovered, benefit to your bottom line isn’t zero, but it’s not a huge number.
So many of the data collection and analysis programmes I see are of very dubious benefit to the business overall, at least on a net basis after factoring in the costs.
If it was a piece of machinery in your factory which cost £100,000 a year to run, and produced occasional £20,000 upsides, but most of the time brought no benefit at all to the business, you wouldn’t buy it.
Yet plenty of businesses spend that £100,000 a year without a second thought because they believe there’s a previously undiscovered secret hiding inside all the data they weren’t collecting previously. Dear reader, I have to say there almost never is, in my experience.
In particular, the data is unlikely to give you information that someone who really knew what they were doing couldn’t have told you already. But instead of listening to those people, many businesses would rather spend £100,000 a year and still not find the answer.
The lack of a system
From what I’ve been able to tell, one of the main reasons corporate data collection efforts are so often a waste of time is that they lack the clarity about the systems they operate in the first place. And often, they don’t really have systems at all.
Oh sure, they’ll have some procedure manuals around the place, or a “systems bible” (as I heard it described once), which tell people what to do.
Lots of people think if they talk about systems and processes often enough, it’ll make them sound super smart and they’ll get a promotion. But what they’re usually talking about is a procedure manual of some sort, which is a million miles away from having a system in place.
This comes out in lots of different ways in practice, but I’ll just home in on a small number of them here. I should say at this point that the work of my long-time business hero W. Edwards Deming is especially relevant to this section, so spend some time with his writing if you want to take this seriously.
There are four key elements of a well-running system, according to Deming. And the number of times I have seen anything close to this in a real business are vanishingly small, so the odds are your business has some work to do in most or all of these areas:
1 – Systems are not silos
I often hear people talk about their marketing system or their HR system. But however good those might be within a single area of operation, that’s not the same as your business having a system in place.
To operate in any meaningful sense, a system has to be an end-to-end piece of work which encompasses your business’s entire range of operations.
As an example, you can have the world’s best marketing system, but if your sales team takes the leads your marketing team produce and make a complete hash of converting them into customers, all the effort your business has put into creating the world’s best marketing system has been a waste of time when it comes to the bottom line impact.
2 – Is it a blip or a trend?
For the purposes of illustration, let’s assume you have a system in place. And that your business has achieved 80% of what would be expected today. Should you care?
Well, that’s a trick question. From that information you can’t possibly know.
But in most businesses, there would be people dashing around all over the place, collecting data, printing reports, having meetings and crisis talks with one another.
Which all looks a bit foolish when the following day clocks in a 120%, meaning the average across both days is the 100% of target you expected all along.
Understanding what a variance really means is critical – and that’s something too few organisations really understand.
Within every system, however well-run, there will always be a degree of natural variation.
But organisations have a tendency to either dramatically over-react to differences which are entirely within the expectations of a statistical model, or dramatically under-react and do the corporate equivalent of Emperor Nero fiddling while Rome burned.
3 – The right data
While I was uncomplimentary about most organisations’ data collection efforts above, I acknowledge that some data is required to make decisions about your business. The trick is working out which data is the right data to collect.
There’s a temptation, especially for people who like to live their lives through the lens of an Excel spreadsheet, to believe that all data is equally valuable, so they set out to collect it all.
The reality is, in most organisations, a relatively small subset of all the data which might potentially be collected is truly mission-critical. The vast majority of it conveys no particular benefit or insight at all.
A common example of this is the NPS stats larger organisations like to obsess over. I’m always puzzled by the way in which those organisations can convince themselves they’re doing a great job based on their NPS scores, while still being widely hated by a large proportion of their customers.
I can tell how happy your customers are by spending 10 minutes in your call centre at the busiest time of day. I don’t need NPS scores to help me.
However, in a previous life when on-time delivery was a critical factor for the business I ran, one of our key metrics – and one I genuinely obsessed over – was the on-time delivery of the materials we needed for client orders.
Because pretty much every job for this business was bespoke, we couldn’t buy ahead of customer orders. But if an incoming delivery was late getting to us, odds are we would be late delivering to our customers.
Even there, although we collected the stats, the fact that I could see our suppliers’ trucks pulling into our loading dock from my office was a benefit I often used to save me having to pour through the stats. I knew that, whatever else was going on, on-time delivery from our suppliers wasn’t an issue because I saw the trucks passing by my office window long before the on-time delivery stats came out.
4 – People impact
Science and engineering types think that writing down a supposedly water-tight process is all that’s required, and people will just follow that process like robots.
That’s the complete opposite of my experience. At both ends of the spectrum.
There is almost no formal procedure a human being, deliberately or otherwise, can’t interpret in a completely different way to the intention of whoever wrote down the procedures in the first place.
And some of your best employees won’t follow that process either, but on the upside.
Perhaps your process instructs a supermarket cashier to wish customers a nice day and move on to the next customer in line. But perhaps today they see something in your customer’s eyes that makes them think they might be experiencing some personal or emotional difficulty, so they take another minute to ask the customer “Are you OK?”.
At that moment, they are being inefficient and going against everything the system some alleged genius wrote down at head office says they should do.
But in that same moment, at a cost of virtually zero, they are helping another human being out the kindness of their hearts and, possibly, making it more likely that this particular customer and everyone within earshot will think “What a caring place this supermarket is. I must shop here more often.”
In organisations where the impact of people on an organisation is not fully understood, all the data collection in the world won’t help you. Your decisions will still be sub-optimal.
So what do you do?
Much of the time, collecting data and tinkering with things on the 1% gains principle is somewhere on a spectrum between being a waste of time and being positively detrimental to your bottom line.
So, as Adam Ant used to say, what do you do?
I’ve always found “what’s the upside?” is a good question to start with.
If you spend £5,000 a year on something, and you can save £50 a year (which is 1%, after all), how much effort is it worth?
Pretty much zero, in this case. Although this sort of thinking does explain some of the terrible-quality paper people have put in office printers over the years.
If it takes even the tiniest bit of administration, management, or oversight to save £50, I can tell you it’s just cost more than £50 to save you £50, which is the exact opposite of improving your bottom line.
The difference is that the £50 saving on photocopy paper is visible, but the £200 of management time to unlock that saving is invisible, because you pay your Procurement Manager a straight salary. So nobody ever calculates the true RoI.
On the flipside, a 1% saving on a £5million spend is worth spending a bit of time on.
But, a remarkable proportion of the time, you would save more money by firing your Procurement Manager and putting their salary cost back on the bottom line even if, in the same breath, you accept that you might spend an extra £50 a year on printer paper every now and again.
Internal staff resources aren’t free, so their costs need to be factored into any RoI.
However, asking yourself “what’s the upside?” will help you take an early go/no-go decision about most things you might be tempted to do.
Yes, even when that comes to collecting data, improving systems, and finding marginal gains.
I know it sounds simple. Almost overly-simple, perhaps, but the fundamental challenge to manage your bottom line effectively is not to spend any more than you have to in carrying out any activity inside your business.
While not every idea someone in your business has will be a sure-fire home run, at least “what’s the upside?” will keep the lid on the number of times someone sets out to deliver a project which has a low expected return, relative to the cost incurred to get there.
Unless that upside is a significant multiple of the cost of making any change, you’re probably better not collecting that data and not making that change. Being able to wrap that idea up in a story about how the British cycling team became world-beaters still doesn’t make it a good idea if the impact on your bottom line isn’t big enough.