If it says it’s going to be sunny on Saturday, don’t go sunbathing in the rain on Wednesday
Even with the global investment in supercomputers and statistical analysis, it is widely accepted that a useful weather forecast beyond 5 days becomes almost meaningless due to the exponential tail-off in mathematical predictability.
The chance of any business being able to produce more accurate forecasts than weather-supercomputers is questionable at best, without dedicated statistical mathematicians.
Business involves so many variable permutations, forecasting becomes an extraordinary ask, and then further risk to the hope or decisions that may rely upon them.
The irony of forecasting is that it can often increase business risk by placing emphasis on the forecast, and losing focus on the historic analytics and optimisation to actual data.
By creating extraordinary work working on guesstimation and spending hope with diminishing returns from these management entertainment endeavours, further expert time-costs are also lost from performance analysis and reactivity.
Just give me your best guess then
Even the most powerful super computers [1] can only predict the weather in 5 days with 75% accuracy. [2]
If we say the weather will be the same tomorrow as it is today, we will be right 70% of the time. [3]
So, even with a super computer and dedicated team, 75% accuracy is the upper-most benchmark for 5 days time – but if we can manage with 70% accuracy – then we can far more quickly and cheaply make a simple prediction with an educated guess.
Every variable that does change, also then reduces forecasting accuracy exponentially – and the inevitable multiple people and variables involved in any business does mean that change in them is ongoing and often very quick – so resources are better deployed to managing change than they are writing horoscopes.
Evidence-based decision making
The starting point should be looking at actual, proven historic data – for the same previous period from a reasonable sample or records, and looking to exceptions – the spikes and troughs – in their context.
Then, you can reasonably make decisions based on the assumption that we will generally be within a 30% range of that again, when there is no other change in the environment in which those events occurred – saving your budget from supercomputers, and your analysts’ sanity!
Our business has grown over many years with almost no forecasting – because the symptoms and actions they would attempt to address are already being worked on first, with a deeper market and performance analysis of every area.
Report checks and responses are happening every single day as part of the routine of what we need to do in ecommerce particularly.
It’s a simply and healthy habit for everyone that can affect performance in any area, to have access to these analytics and glance at them for a few minutes a day, just as it is to look out the window before leaving the house.
What forecasts should I use?
There is only one forecast that should be integral to your financial accounts and that is a cashflow forecast.
However, cash-flow forecasts are not something that should ever be based on non-realised sales – it should only contain all currently confirmed income and expenses. Adding speculative data, would make it a business plan – which is fine, but the distinction must be declared and compared for context.
Healthy habits create healthy businesses
The way to work on future cash positions with that cash-flow forecast is to input future expenses on the date they are committed to be paid, and future sales on the date they are contracted to be paid – this then means your regular accounting software can do the rest for you.
If you want more guarantees for the future, then you need to ask for more forward orders – which can be done in many ways – but starts with offering that facility in your website and sales process.
You can even offer incentives on price, and security over inventory, for these forward orders. This gives you safer planning, and the commitment from your customers and partners to give you actual numbers to work – which is just more secure and safe to plan production with than forecasts – and builds a stronger business for everyone.
Make this a policy document too, so that everyone knows the right way to work in their training, long after you’ve left your growing expert team to manage themselves – because you respect your experts, and they respect your commitment to structure and standards.
Stop selling and spending forecasts
The bigger danger in forecasts is that because they might look reasonable or reliable because they are in a pretty, and sometimes complicated looking spreadsheet or report.
However, there could be temptations in financial planning to then spend them as if they are some kind of guarantee or somehow over-promising, leveraging or financing based upon them – all of which inevitably will back-fire sooner than expected because we are now in the realms of gambling and distortion.
If you’re being asked for a forecast by anyone that has anything to do with sales, whether selling products, services, plans or dreams, then double your caution because there is a habitual tendency for positive bias because that is mostly what the role involves anyway, which is great in many ways – but also commands equal caution from those buying into the information and those feeding it.
The absolute biggest danger of all though, being self-bias becoming unsustainable risk-taking.
We all want to hope for the best – but, as my mother says; “hope for the best, plan for the worst, and expect something in-between”. Meaning, don’t forget your risk management – what happens when the forecast is wrong?
The best way to get closer to a guaranteed income and expenses in business, is with complete risk-management awareness and documentation, terms, policies, procedures, and agreements on all.
Ultimately, focus on the business operations and development in-hand, that can generate extraordinary performance, can only result in positive cash growth – a much nicer problem to then have to debate.
But I need a forecast because I was asked for a forecast
Well – do you really?
What is going to be the result of that forecast if it says good things?
What is going to be the result if is says bad things?
Now you have a risk management situation based on the reaction to forecasts too, and now *work* is being created instead of real efficiency or value to offer to the world.
Try charging for forecasts with explaining the opportunity-cost of the time value of those that have to create them – ask what the requester would like you to postpone while you work on them – and then see how important they really are.
Or just send your forecast requester the link to this page – and maybe ask them to create the forecast template that they require for populating first – if they aren’t already asking for analytics summary or access they really need.
Then, you can find out if they really know what they are asking for, and how much work they are prepared to do or forsake for this extraordinarily risky and expensive request – before they spend your valuable time on fortune-telling instead of fortune making.
I cannot recall a single success story, autobiography or masters in business administration study that placed forecast creation and sharing as the highest priority for success in any business – but I do know many that will strongly advise confidence in numbers, pattern-analysis, bias, and studying your results to learn and improve from them.
So if I don’t need forecasts, what do I need?
Simple: Analytics first, then key performance indicators, benchmarks, market research, monitoring, alerts, planning, strategy – and most important of all, agreements!
Focus on getting agreements on every one of those areas, for who will create it, and who is responsible for monitoring it, and what you will do with the results.
This is Agile Project Management – tried, tested, proven and used by the fastest moving operations in all industries.
Use self-updating reporting systems
Your business is only as good as your agreements, capabilities and review by the outside world – no matter how distracting, unfair or impatient you might think these things might be.
They are all you have in the event that someone unknown looks at your business and asks to see these things – and this happens all the time, without notice.
If you are not already ahead on this, you are risking a negative view, that might be completely unfounded, but you allowed to happen by taking the shortcuts in the first place.
Summarising focuses the mind too
More-often, what someone really needs is a summary view of accurate data, and to come to their own conclusions.
Summaries can then recommend some targets to either work towards, or agree you’ll to work towards.
So, every time you are asked for a forecast, have a look at the reporting you already have – and how it can be made more accessible to others to self-serve the data they really need for the decisions they are looking to guide.
Forecasts are not targets
If someone is looking to turn a forecast into a target, often the best people to set targets will be the people actually working on striving for those targets every day.
It is the doers in an organisation that have much more meaningful experiential data to call on in their minds, and should easily be able to beat that 30% accuracy weather volatility benchmark with just the computing power a brain full of memories and feedback.
Generally, people are happier to work on a target that they were consulted on, thought through, and set both for an area or interest – and for their own personal achievement goals.
Less haste, more speed
However exciting it is to get going in business – just take care of the planning and agreements first.
Look-after and ask for advice from those with a daily hand on experiential data, and you will find that the only forecast you need to make is when you will be meeting next or on holiday.
Now – go and get on with that – and send the forecast addicts here for a sanity-check because they might need a holiday too ?
