Most businesses rely on forecasting to make informed business decisions. Business decisions rely on predictions of profitability, customer acceptance, and other vital metrics. Over the years, business executives have developed predictable approaches to forecasting their business performance. But of late, the unpredictable and uncertain business environment has rendered such methods obsolete. 

The COVID-19 pandemic plunged the world into an uncertain future. The business environment since then has remained fluid. At the pandemic’s start, companies had to transform to work-from-home mode overnight. The pandemic gave way to a recessionary environment, coupled with a supply chain crisis and high inflation. Geopolitical tensions in many parts of the world compounded the uncertainty. Today, monetary policies, exchange rates, trade negotiations, and health all cause grave concerns. There is no end in sight to such uncertainties. And things change fast. For instance, a lockdown may come anytime, anywhere, for the climate, if not public health reasons.

Side-by-side, tech fragmentation has set in, leading to a lack of standardisation for enterprise tech stacks. Also, the rapid evolution of technology leaves a short shelf-life for new technology. Skill developments must catch up with tech developments to overcome the talent crunch. The skill shortages have delayed transformation and eroded the competitiveness of many businesses. Most businesses face looming uncertainties if their skilled workforce decides to quit.

Customer expectations also changed overnight. The fast-paced world has made them impatient, and they have become fickle by nature. They expect instant results and engagement through the channel of their choice.

Here are how to make realistic predictions in a fast-paced, fluid, and rollercoaster reality.

1. Invest in data and analytics

Data drives decision-making in today’s digital age. As enterprise processes become digitised, each transaction and engagement generates data. Using such data to make predictions offers rich insights.

The conventional approach to forecasting includes measuring trends and preparing performance reports. Weighing these factors higher or lower depended on predictable business trends. The basic approach still holds good but in today’s unpredictable world,

• Give a broader range of metrics to cope with the uncertainty.
• Broad-based data sources and co-opt global data. Use extraneous data that can better predict changes in the business environment.
• Be flexible to change. Metrics considered irrelevant earlier might become important. Metrics believed essential earlier might turn irrelevant.
• Dig deeper. Insight need not always come from one big thing. A collection of little things can also offer great insights.
• Apply the latest technology. For instance, use predictive AI to factor economic and behavioural volatility into plans. Apply correlation analysis and econometric modelling tools to create economic baseline forecasts. 

There is no crystal ball that can make accurate predictions every time. But the above tools and approaches improve accuracy, with little chance of big misses.

Here are the new metrics IT needs to capture the effectiveness of digital transformation.

2. Prioritise the metrics that matter for growth.

The success of predictions depends on the metrics used to predict trends. Each business has specific factors that influence growth.

Build a model that makes it easy to test what happens when such factors change. A good model makes it easy to decipher how leading indicators, quality metrics and outputs, impact the business.

For the short term, only two metrics matter – cash in and cash out. In the long run, other metrics matter more, and such vital metrics often change during times of uncertainty. For instance, in normal times, the marketing team may look at traffic growth. But during unpredictable times, one can influence the traffic with certainty. As such, focus on alternate metrics such as signups and optimised content. These metrics set the stage for increased traffic and growth.

• Identify how external factors influence the business. Augment internal data with external market signals. Examples include consumer sentiment and behaviour, macroeconomic drivers and supply chain shifts.
• Leverage econometric modelling to quantify uncertainty and co-opt it into the planning process.
• Detail specific initiatives or deliverables to prioritise. For instance, in the content strategy, one plan could be to increase case study volume rather than focus only on blogs.

3. Focus on controllable metrics 

It is hard to control outcomes during times of uncertainty. Government mandates, panic reactions, irrational thinking, and other out-of-normal scenarios foil predictions. In such a situation, enterprises need to buckle down and control what they can.

• Control the purse string. Spending is one thing still within the control of enterprises. Enterprises may have to make contingencies for unexpected expenses. An example is the unplanned spending on PPE kits, masks, and sanitisers during the COVID-19 pandemic. But on the large, controlling spending is a way to ensure the enterprise stays on track and channels resources to a growth trajectory.
• Align teams around critical metrics. Doing so makes it easy to prioritise activities and initiatives that matter. Also, in the absence of such alignment, work teams will default to outcomes which are hard to control during times of uncertainty.
• Obsess over outputs rather than outcomes. For instance, salespeople are rarely in complete control of closing deals. Several factors outside their control influence the deal, even at the best times. During times of uncertainties, such outside-the-control factors increase. To cope, change the focus activities within the focus of control. Salespersons could focus on organising meetings and continue pitching the product. Content marketers could publish more content and help sales book more meetings.
• Focus on quality. Another factor in one’s control is the quality of output. Businesses may focus on moving things fast during busy times, and quality may take a beating. Consider the support team. In the best of times, the focus may be to maximise the number of customers and minimise response time. During uncertainties, customers who may also be facing uncertainties need more help. It makes sense to measure quality. 

4. Update forecast models regularly

In good times, when external factors rarely impact performance, it may be enough to update the model once per quarter. But during uncertainty, which is often a crisis, review such models more frequently. Depending on the business situation, it may make sense to review the model weekly or daily.

• Choose the right forecasting technique relevant to the business.
• Have regular reviews to identify the metrics that matter. Track the metrics that move and focus on what the data reveals.
• Make sure all functional teams align with the approach. Get the support of sales, marketing, operations, accounts, IT, human resources, and other teams.
• Revisit and iterate plans with higher frequency than before. Change from monthly forecasts or even real-time live monitoring.

5. Keep an ear on the ground.

As the adage goes, “an ounce of action is worth a ton of theory.” Customer-facing employees, employees who work in the field, and other people on the ground remain connected to reality. Experts and analysts often work in air-conditioned ivory towers and lose their pulse with reality. Relying only on analytics, for instance, could conjure up probabilities that have no relation to reality.

• Talk with people in the field. A field agent engaged with customers could identify trends through conversations or negotiations. Such insights would otherwise take weeks to materialise through reports and metrics.
• Communicate with stakeholders and employees often. Frequent, in-depth communications give a deeper and more current perspective on business opportunities.

Change is happening faster than anyone imagined. Enterprises equipped to cope with uncertainties gain valuable competitive advantage.