Forecasting with Accuracy

By: Jason Marcus

Whether we’re deciding what to wear in the morning or we’re at the supermarket choosing the best ingredients for a new recipe, our adult brains are making upward of 35,000 decisions each day. This is according to Eva Krockow, lecturer at the University of Leicester in the UK. Most of these decisions are small, but some are so large that they can change the course of our lives or even the course of human history (if you’re a globe-bestriding colossus).

To make decisions as effectively as possible, we innately measure the benefits, costs, tradeoffs and risks of each option. Similar to how we optimize decision-making in our personal lives by anticipating value (benefits minus costs), so too can we anticipate the value of each possible path in our work lives, especially when it comes to drawing roadmaps for new initiatives. The key difference between personal and work decisions is that work decisions often require building presentations to convince others to go along with our ideas, and standard benefits vs. costs analyses are not persuasive enough. It takes trust, and trust in our decisions needs to be built and maintained.

What really helps elevate those prioritization stories? Quantifiable business impact forecasts that align to the metrics senior leaders care most about.

This post offers a step-by-step approach to accurate forecasting. Intuition or pulling forecasts out of thin air only gets product managers so far. A more scientific approach to forecasting and learning via follow-up is needed.

Seeing the World Through Users’ Eyes

First, you must learn everything about the people who use your product — your customers. Here is a checklist of insights you’ll want to consider gathering:

· How many users are there total?

· What are the key segments?

· How many users are in each key segment?

· What are the personas?

· Which personas are primary vs. secondary?

· Do you have empathy-building research?

· Do you have funnel data?

You’ll need to deep dive all the data above in detail to gain a crystal clear view of who comprises your universe of users. At a broad level, you’ll need an understanding of who all the users are, how your product fits into their lives and all the ways they’re currently interacting with it.

Quantifying the Forces

Next, you’ll need to understand the forces that push users to and from your desired business outcomes. Typically, motivation level, value proposition/differentiators and incentives are the forces that propel users toward a “conversion.” On the flip side, friction and anxiety tend to pull users away from a “conversion.” You can think of these as driving and hindering forces.

Finally, you’ll need to review your backlog of initiatives and view each item through the eyes of your user base. You’ll need to anticipate how the proposed change increases or decreases user motivation, value proposition strength, incentive level, friction and anxiety as users attempt to complete their jobs to be done. Once you determine which levels are increasing or decreasing, and by approximately how much, it should become clear how the change will affect baseline key performance indicators (KPIs).

To illustrate a possible forecasting scenario, here is a hypothetical issue:

Discovery research finds that when customers open an account online, users enter the flow from search and paid ad channels directly onto product detail pages (PDPs), without being provided with a simple, intuitive way to navigate to the comparison product pages once they land on-site. Analytics data shows that 60% of overall site visitors enter the flow this way, and usability data confirms the hypothesis that there is an inter-page navigation problem. This is also confirmed via an exit rate report that shows 70% of visitors are exiting from PDPs without converting. In reference to persona research data, it’s clear that customers who are less tech savvy feel the problem at a higher rate than more tech savvy personas.

The cross-functional team decides to add “breadcrumbs” near the top of the page to simplify the navigation for users. This change is prioritized based on a forecast of how the target audience will perceive the enhancement’s utility, and the leadership team agree that it would reduce friction because it would provide an easy way to navigate to the comparison page so a fully informed decision could be made. The change would also help answer the immediate “Where am I?” question that visitors feel being dropped on a product page fairly far down the funnel. The overall baseline conversion rate is 24%, and the team believe that reducing friction for 60% of visitors will impact the overall conversion rate by +6%, raising the total conversion rate to 30%. Revenue is forecasted to increase linearly as well, by +6%.

After launching the “breadcrumb” enhancement as a production A/B experiment, it was observed that conversions within the test increased to 32%, which would turn into success for the company if the increase held true as expected when annualized. Together the group assessed WHY conversions increased by more than expected within the experiment. It was determined that tech savvy visitors received a more-than-expected amount of value from the feature, and thus they drove up the numbers in an unexpectedly positive way.

The most important part of the process is actually the last paragraph above — continual improvement.

As you assign forecasts to improvement opportunities when they’re in the backlog, it’s critical to maintain a culture of experimentation in which you monitor for actual business impact post-deployment and compare against the forecast. If the forecast was accurate, celebrate it! But if it wasn’t, do your best to understand why. This is key to fine-tuning your forecasting process. This exercise of continual monitoring and improvement works best at the team level, as forecasting and retrospectives can be done as a group to encourage team growth and skill development.

After you and your team have done this exercise hundreds or even thousands of times, you’ll collectively become highly skilled at forecasting accurately. Just give yourself and your team grace if you’re just getting started, and know that over time the skills will improve. It’s never too late to get started.

Putting It All Together

This exercise should ideally yield an accurate forecast — especially if you’ve done this many times — and it becomes the quantifiable benefit of the proposed enhancement. If you add up all the benefits, both tangible and intangible, and subtract all the costs, both tangible and intangible, you should start to understand the overall value of the idea. Skilled product leaders break up roadmaps into swim lanes and objectively prioritize the highest value improvements that ladder up to outcomes and key results. Alignment should be the North Star.

Just like how we quickly and naturally optimize decision-making in our personal lives, so too can we extend that logic to our professional lives as we attempt to improve business outcomes. This forecasting method of building deep customer understanding and assessing how the forces that push users in different directions through the funnel change will take some practice, but over time it should become more innate and accurate. Just take the first step if you haven’t already.

For Informational/Educational Purposes Only: The opinions expressed in this article may differ from other employees and departments of JPMorgan Chase & Co. Opinions and strategies described may not be appropriate for everyone, and are not intended as specific advice/recommendation for any individual. You should carefully consider your needs and objectives before making any decisions, and consult the appropriate professional(s). Outlooks and past performance are not guarantees of future results.

Any mentions of third-party trademarks, brand names, products and services are for referential purposes only and any mention thereof is not meant to imply any sponsorship, endorsement, or affiliation.

Like what you’re reading? Check out all of our opportunities here