In eCommerce, forecasting is valuable for predicting a company’s cash flow and growth; doing so in a way that is accurate and free of bias, however, is critical to the viability of a business in the long run. Explore essential forecasting metrics, account for all information and variables, and combat behavioral errors by engaging in these five steps for making better projections:
Use the right tools
Forecast growth and revenue conservatively
Understand your costs
Use the Right Tools
Relying on a set of tools to experiment with your assumptions will not only standardize your forecasting process but also allow you to focus on hard data. The most commonly utilized and valuable instruments of projection are pro forma financial statements:
Income statements: a forward-looking profit and loss statement, which takes into account different assumptions and scenarios to project future estimates.
Balance sheets: a forward-looking reporting of a company’s assets, liabilities, and shareholder’s equity at a given point in the future.
Cash flow projections:a forward-looking breakdown of net cash usage from operations, investments, and financing activities.
Though projections are likely to be error-prone, the process of creating them forces forecasters to really think through the full spectrum of possibilities, opportunities, and potential challenges surrounding future revenues, costs, and cash flows. These tools are forward-looking estimates of profitability and financial position, and they assist you in answering essential questions such as:
When am I going to run out of money?
Can I afford to grow this fast?
How much does growth cost?
How is cash being utilized in my operation?
Do I need to raise more money?
Forecast Growth & Revenue Conservatively
Estimating revenue is a vital part of developing accurate projections. There are different ways to estimate revenues, and selecting the most relevant formula will depend on your business model and the availability of historical data. Here are some common methods:
Revenue = (units sold x price/unit) + shipping – returns
Revenue = revenue per visitor (RPV) x number of visitors
In order to get this piece of your pro formas right, consider estimates — preferably based on actual historical data — for important metrics like unit sales, price per unit, number of visitors, conversion rate, and AOV. Keep in mind: revenue estimates are almost always wrong. Even when you have access to good data from the past, there is no guarantee that the future is going to look similar. Things frequently change in eCommerce as they do in business more generally. If you are going to err, be on the conservative side — aggressive assumptions are dangerous if you start to believe them.
Understand Your Costs
Forecasters must factor in the full gamut of likely costs — not just the obvious ones — when projecting future expenditures. Consider the following:
Products: Don’t overlook the costs associated with researching, developing, manufacturing, and shipping your products. These costs almost always come in heavier than anticipated, especially for start-ups; however, they are often misjudged by experienced practitioners as well.
Marketing: Generating sales costs money. Whether you are driving organic or paid traffic to your store, a significant investment in marketing is required. Quick growth generally requires paid campaigns of some kind. A few metrics to keep in mind:
CPM (cost per thousand or cost per mille): a metric used to describe the price of 1,000 advertisement impressions on one web page. CPMs depend on competition, time of year, time of day, and perceived quality, among other things.
CTR (click-through rate): a metric that evaluates the number of clicks advertisers receive on an ad per impression.
CPC (cost per click): the amount an ad pays for each click on an advertisement.
CPP (cost per purchase): the cost required to convert a sale on average.
Fulfillment & Shipping: Fulfillment and shipping are important and costly elements of running an eCommerce business. Speaking with a fulfillment provider like IronLinx will allow you to best understand all of the costs and assumptions associated with accounting for fulfillment- and shipping-related variables. A few things to keep in mind:
Picking, packing, and shipping
Returns: Even experienced eCommerce sellers often underestimate the impact of returns and exchanges on financial performance. In certain industries like fashion, for instance, return rates are often in excess of 30%. A few things to keep in mind:
Return shipping fees
Inspection, rework, and restocking fees
Technology: Technology is an often-overlooked piece of forecasting. Business owners need to account for important technological elements such as:
Software: storefronts, fulfillment integration software, and add-on apps like free shipping bars, social proof elements, and other interactive features on your website.
IT Infrastructure: computers, printers, mobile devices, servers, network infrastructure, IT troubleshooting, and/or IT managed services contracts can add up quickly.
Online brand assets: websites, social media pages, and other online brand assets are the lifeblood of most eCommerce operations — and they take consistent and steady investment of money, time, and other resources.
Finance: Don’t overlook hidden fees and costs associated with financing your operations. A few to keep in mind:
Merchant services fees
Bank account fees
Accounting software licenses and tax preparation fees
Sales platform fees (Amazon selling fees, etc.)
Timing: Everything tends to take longer than it should: what should take a day often takes a week; what should take a week often takes a month; what should take a month often takes a quarter. In the end, this tends to push back revenues, while overhead-oriented costs continue to accrue unabated.
Perspective: The forecasting errors that we make tend to be on the aggressive side – with the result being that actual revenues and profits are likely to lag expectations. If you are working on a limited budget, this can be a real problem as you are therefore susceptible to running out of money sooner than anticipated — this can easily be the difference between success and failure.
Scenario Analysis: Use scenario analysis to evaluate the effects of external variables, like market conditions, on your business. Not everything is under your control and planning for the worst can help you to survive it.
Sensitivity Analysis: Using sensitivity analysis allows you to identify the variables which are particularly important to the success and well-being of your business. For those variables which offer little to no room for error, sellers would be wise to do whatever possible to mitigate uncertainty (contracts, diversification, etc.)
In the developing world of the behavioral social sciences, an enormous literature has been compiled which highlights the errors inherent to human biases, heuristics (rules of thumb), and framing tendencies. For instance:
Excessive Optimism: the evidence suggests that we as individuals are prone to overestimating the likelihood of good outcomes while simultaneously underestimating the likelihood of bad outcomes — biasing us towards aggressive assumptions.
Overconfidence: the evidence suggests that an individual’s subjective confidence is often greater than their actual performance — again, biasing us towards aggressive assumptions.
Illusion of Control:the evidence suggests that individuals tend to overestimate their ability to control independent events — which, once more, creates an aggressive bias.
Anchoring & Adjustment:the evidence suggests that we are likely to anchor ourselves to our initial pre-conceptions — adjusting them only inadequately to new data as it is presented.
None of these behavioral issues are doing you any favors as each of the above (as well as others) cause us to be overly aggressive in our projections — which is likely to cause pain when realized numbers come in below expectations.
Conclusion: How do we make better projections?
In order to create useful forecasts, decision makers need to put a lot of critical thought into their assumptions. It’s easy to plug and chug with random numbers, but the devil really is in the details. A few questions to help you get thinking along the right lines:
What does it actually take to get to x amount of orders per day or month?
Do you have the resources to get to x amount of orders per day or month?
Is there any historical data to assist in making my assumptions?
Is there reason to believe that the descriptive statistics of the past are going to hold in the future if you are projecting to scale to a materially higher level?
Do you have the budget to generate a certain amount of traffic?
Can your supply chain keep up with your projections?
Remember: projections are only as good as the assumptions on which they rest.