The Jelly Bean Jar: A Lesson in Incremental Discovery

Planning a digital transformation, replatforming your commerce engine, or standing up a B2B portal? Project estimation in these contexts can feel like guessing how many jelly beans are in a jar—until you start doing the work. What seems like a simple task quickly reveals hidden complexity, with success depending upon how much you really know. Let’s unpack why project estimation mirrors the classic jelly bean guessing game—and how to turn uncertainty into clarity.

--------

Picture yourself in a classroom, staring at a jar of jelly beans across the room. You’re asked: “How many are inside?” You’ve seen similar jars and know the size of a jelly bean, but your guess is based on limited information. Let’s face it—at this point, your guess probably isn’t going to be very accurate.

Now imagine holding the jar. You weigh it, measure its dimensions, and shake it to gauge how tightly the beans are packed. With each new analysis, your estimate improves. Still not perfect, but better than before.

Next, you start counting. A quarter of the way through, your guess has sharpened. As the pile in the jar begins to dwindle, your estimate of the total remaining grows ever more accurate. Yet, until every bean is counted, you can still only estimate the remainder.

Project estimation works the same way. Initial estimates are just rough sketches. Only through hands-on exploration can assumptions solidify into actionable plans.

Until the final deliverable is complete, the remaining level of effort is just an increasingly well-informed estimate.

Discovery Is Your Secret Weapon

Like a flashlight in a dark room, discovery doesn’t create scope—it reveals it. Using the jelly bean analogy, the total number of jelly beans is the total scope, and that remains fixed, regardless of the discovery methods that are used, or the estimates that are provided. Discovery is your chance to gather data, test assumptions, and clarify needs—transforming vague guesses into precise estimates.

At Zaelab, discovery goes beyond the basics. It’s not just about gathering requirements; it’s about validating direction. That means using strategic roadmapping to align business goals with implementation, conducting platform fit assessments to ensure the right technologies are in play, running proof-of-concepts for CPQ or headless builds to test ideas in the real world, and performing tech stack evaluations to identify gaps and future-proof decisions.

This combination makes discovery the foundation of accurate estimation. The more information you have, the more accurate your estimates will become, and the fewer surprises you’ll face during implementation.

Should I Pay for Discovery?

The decision to invest in an upfront discovery phase boils down to one question: Do you prioritize confidence or speed?

• A paid discovery phase offers clarity upfront, providing valuable deliverables such as a well-defined scope, precise estimates, and a roadmap grounded in research. Paid discovery is ideal for clients who wish to reduce risk and wish to execute implementation with minimal surprises.
• Unpaid discovery accelerates timelines by shifting uncertainty into execution. Note that unpaid discovery does not eliminate the need for discovery—an implementation plan must still be created for each piece of a project—but this will happen alongside development, rather than before contract signature. Unpaid discovery is ideal for clients who want to move quickly, and are tolerant of the increased risk of change orders, delays, and budget overruns.

Consider this—without compensation, is an agency more likely to diligently learn the full requirements of the scope, or are they more likely to make assumptions? Cutting corners leaves clients to bear the cost of undiscovered risk. Ultimately, it’s the client—not agency sales people—who suffer the consequences of unrealistic promises made during sales, especially under an hourly billing agreement.

Trust is built on proof, not promises. Before signing a contract, ask:

• Has the agency demonstrated specific solutions to your challenges?
• Has the agency shown working proofs of concept, or has there just been lots of talking, hand-waving, and slides that claim they can meet your needs without proving anything?
• Has the agency openly addressed gaps in their expertise?
• What makes you confident that the agency can meet your needs?

Paid discovery protects both parties. Clients gain confidence, while agencies avoid ambiguity. If you understand the risks of unpaid discovery, and you’re willing to absorb them, then proceed confidently knowing that the details will come later. Otherwise, invest in the confidence that upfront discovery provides.

Uncertainty always has a price. The choice is whether you pay it upfront or later—with interest.

Iterative Discovery: The Antidote to Analysis Paralysis

How do you know when to move from discovery to delivery? Gathering too little detail leaves risks buried, while gathering too much detail wastes time. You don’t want to get mired in the details of discovery, unable to make decisions or move forward.

You will find that discovery has diminishing returns. Early insights drastically improve estimates, while later ones offer only marginal gains. The goal of discovery is to obtain just enough information to confidently make decisions—then act. 

A process of iterative discovery—a cycle of probing, learning, and refining, conducted during a series of working sessions between the client and the agency—allows you to achieve these goals:

1. Start high-level, then dive deeper in future discussions as needed until the desired level of confidence is achieved.
2. After each session, ask:
   
   • What did we learn today that we are adding to the scope?
   • How does that impact the estimate?
   • Do we now have enough confidence in the scope to sign a contract?

These checkpoints prevent endless deliberation. Either confirm readiness or identify gaps—keeping trust and progress intact.

Turning Estimation into Execution

Clients need to feel that the agency truly understands their goals, pain points, and constraints—and that they’re capable of delivering. Ultimately, the level of detail that the client requires in order to feel comfortable signing a contract is determined by the client, and until those needs are met, discovery will continue.

Eventually there is nothing more to be gained from additional discovery. At this point, the best next step will always be to begin work. If the path is clear, start building. Remember: The best way to refine an estimate is to begin the work. Once development starts, you’re no longer guessing jelly beans—you’re counting them, one by one.