The nonsense of numbers

Numbers are the language of business. But sometimes...often, I might argue...the numbers we are presented with are varying degrees of nonsense.

Admittedly, I'm a spreadsheet wrangler by profession. Which is a bit like being a butcher who knows how the sausages are made.

I want to be very sure of the provenance of any numbers presented to me (or was it sausages...I forget which...) in case someone has been playing fast and loose with the recipe.

Nearly always, you have to go beyond the numbers to find the truth. You can't take them at face value.

You see, numbers alone are too simplistic.

They're too easy to manipulate (deliberately or unintentionally).

And even if faithfully prepared, they're too often misunderstood, leading to unwise decisions and potential calamity.

"Is this guy crazy, or what?"

You might be wondering by this point if I'm completely crazy. (Although if we know one another IRL, you may have stopped wondering about that quite some time ago...)

But it's true. I'm an accountant who doesn't completely trust any numbers presented to him.

Especially when they're presented by people with an agenda they would like to pursue, who are using numbers to support their case.

Even if they're as honest as the day is long, it's natural for people with an interest in a particular outcome to want to find plenty of evidence to support their proposal, and not be quite as diligent in finding data which suggests they'd be mad to attempt it.

A bigger danger, though, is people who think they understand numbers because they were good at maths as a kid or can do rapid-fire mental arithmetic in their head.

Those people are generally easy to persuade if the maths of the dataset they're presented with is good, even if the underlying premise makes no sense (a quick shout-out to a former colleague I'm not going to name here, who elevated that to an art form).

These people revel in the purity of the calculation, bathe in the warm bath of multiple pivot tables, and relax into the comforting embrace of an SQL query.

Their Achilles heel is that pretty much nothing in the world runs in practice the way it's supposed to run in theory.

But the biggest danger of all, though, is people who take every number at face value and don't dig any deeper.

A few examples

Much of this is counterintuitive, but I've seen all of these at one point or another in my own career...

Increasing sales is always good news!

In an ideal world, we'd all like to see the graph of monthly sales figures going up and to the right.

But what if the business had allowed itself to be boxed into a corner on price by a customer, with the result that margins are non-existent.

Or the sales growth came from an unwise move into a different product or market without factoring in the extra cost and additional complexity.

Or the extra sales were booked this year, but the 10% rebate the customer demanded as part of the deal won't hit the accounts until next year.

Reducing cost is always good news!

All things considered, we'd all rather have lower costs than higher costs.

But what if the business was only fixing breakdowns that stopped production, intending to leave the ruinously expensive repairs for the business planning to acquire them.

Or if buying cheaper materials meant they were likely to damage their reputation for rock-solid reliability, resulting in lower unit sales prices as an undifferentiated player in a competitive market in the short term, and higher warranty costs in the medium term.

Or if the reduced cost had been achieved by firing all the people who knew what they were doing and replacing them all with trainees. History suggests that rarely has a positive outcome.

More cash in the bank for the year-end accounts is always a positive!

Yes, more cash is usually better than less cash. But it's not that simple.

I've seen plenty of businesses "manufacture" strong year end cash balances just by delaying their end-of-the-month payment run for 24/48 hours from 30th/31st of one month until the 1st or 2nd of the following month.

A large company can easily add ￡millions to its year-end cash balance without changing a single other thing in the business, just by doing that.

Or have they just stopped paying suppliers for a month or two and built up cash that way.

Or are there big staff bonuses due to be paid out a couple of months into the new year which are still sat in the business bank account at year end.

Year-on-year profit growth is always good!

At some level, of course increased profit is a good thing.

But what about if the business has taken some extra costs to get out of a project where it became apparent there was no prospect of making a decent return. That's probably a net positive, not a negative.

Or if the business has scrabbled around to find anything they possibly can to increase the sales line, but has turned a blind eye to any extra costs or liabilities which might be floating around the business. (Be particularly sceptical of businesses which report profits this year just a few ￡000s up on last year - with rare exceptions, they've probably strained every sinew in their bodies to achieve that.)

Or if the year end numbers have been manipulated using provisions for things like work-in-progress or projected bad debts at year end.

For the unscrupulous, this is not especially difficult to do and while the auditors will be looking out for it, I've found unscrupulous people tend not to draw a line at telling porky-pies to the auditors after spinning all manner of stories to the board, their bank, and every other stakeholder. So it can be hard for even the most diligent auditors to spot.

The reality

The reality is that numbers alone tell you very little.

Just because something appears in an Excel spreadsheet doesn't mean it's true. Even if it is factually accurate, it doesn't necessarily mean you're seeing "the truth, the whole truth, and nothing but the truth".

Think of it like this.

A factually accurate statement might be that a sprinter ran 100m in under 9.8 seconds. The sophisticated timers between the start line and the finish line in a stadium can accurately record their time to the nearest 100th of a second.

You can put that number in a spreadsheet, plot it on a chart, and use sophisticated analytics to compare their performance with all the other world-class sprinters.

But your spreadsheet isn't going to tell you that they achieved their time with the aid of performance-enhancing drugs.

So when it comes to your business, you need to make sure nobody has been dabbling with the "funny powder" in the off-season, hoping they don't get caught.

Three ways to check your numbers

Some strategies I use are very situation-dependent, but here are three strategies that work pretty much anywhere for interrogating a set of numbers before you decide whether to believe them or not.

1 - What's the presupposition?

Almost every set of numbers you get presented with has an in-built presupposition (sometimes several of them). Work out what that is and test it every way you can think of to see if it's robust.

A presupposition might be that increasing sales can only be a good thing, or that the same is true of cost reductions. As we saw above, both of those can be true, but they are not necessarily true.

In this context, a presupposition is usually the weakest point in the argument being presented to you. It's the single most likely part of the proposal to scupper the whole project, even if every other element runs perfectly according to plan.

So identify the presuppositions and unpick every aspect of them until you can be sure it's a justifiable position to take. (Word of warning: the people presenting to you generally won't know what their presuppositions are. We are usually blind to our own presuppositions, so you need to work this out for yourself.)

2 - What's the story?

I find if you get people off the numbers and make them explain their proposal to you in words a moderately smart teenager could understand, some of the holes in their thinking become a lot easier to see.

When I hear something like "as you can see X is 17% of Y, so our best strategy is to do Z", I might ask them to explain why Z is a good strategy without using the numbers presented as their collateral.

Recently a number of high-profile businesses have crashed and burned because their valuation assumptions were based on the idea that some huge percentage of the world's population would find their products highly attractive.

The problem here was not usually the numbers. The maths was invariably accurate.

But the underlying story - something like "of course 25% of the world's population are going to use electric ride-share scooters as their only source of daily transport", for example - might be complete garbage.

3 - Look for links

No numbers exist in isolation. There is always corroboration somewhere.

If, for A to happen, B would need to happen first...but B has not yet happened...then it's unlikely A is going to happen either.

Although they go to great lengths to disguise this, I'm sure, Apple can't manufacture a new iPhone without buying up a gazillion of the new microchips they need to build those devices.

So if the chip manufacturers aren't turning out a gazillion of their newest microchips, odds are the next generation iPhone isn't getting launched any time soon.

Go upstream and see if you can find some early indications to corroborate the fact that A is likely to happen. If you can't find any, it probably won't.

And the final kicker

Just because something has worked in the past, that doesn't mean it is necessarily still true today, or will continue to be true in the future.

When I've seen organisations in real trouble, almost inevitably their approach did work once upon a time (even if purely by chance).

And it worked well. That's why they kept doing it.

But they haven't wanted to change a model they knew and thought they understood.

The senior team often didn't listen to people who were seen as naysayers, or "negative" (even though history shows they were, in fact, right all along).

And some form of groupthink has usually taken hold, where people convince themselves that they are right and the rest of the world is wrong, or their customers are stupid, or their staff are feckless.

Usually, like that haven't been challenging the numbers they were presented with as robustly as they might have done, especially when those numbers confirmed a worldview they would desperately have liked to be true.

Most of the businesses I've seen in deep trouble have had mostly green boxes on their RAG-rated monthly KPI reports immediately before their worlds came crashing down.

So don't get taken in by the nonsense of numbers.

You might not notice until it's too late.