Discovery at Project Start

Core question — Are all three lined up: AS-IS / TO-BE / GAP?

The goal of discovery isn't "to listen" — it's to line up all three: AS-IS (current state), TO-BE (target state), and GAP (the difference = the issue). Move on to estimating or requirements definition with even one missing, and you'll always come back later. The very habit of checking each time whether all three are lined up is the core of this course.

The client speaks of only one of the three

The client starts speaking from either pattern A or pattern B. A client who brings all three lined up from the start is virtually nonexistent. The remaining two are the discovery side's responsibility to draw out.

Defining GAP — it's not a list of AS-IS flaws

GAP = the issue is not "a list of complaints about AS-IS." Define it as what you must clear to reach TO-BE — that is, the difference of TO-BE minus AS-IS. Keep this definition from drifting and requests unconnected to TO-BE won't sneak in.

• ─✗ Listing out AS-IS gripes as-is → requirements unrelated to TO-BE slip in
• ─✓ Set TO-BE first, then extract issues as the difference → connects directly to estimate line items

The resolution at which you articulate GAP

Don't stop at "vaguely struggling" or "want to improve operations." A GAP only becomes a GAP once you can articulate what, how much, and when it's a problem. Coarse granularity makes both the estimate and the requirements coarse.

• ─What is the problem — which task / which owner / which form
• ─How much of a problem — count, minutes, frequency (in numbers)
• ─When it's a problem — the conditions of occurrence: monthly / on exceptions / peak season

Issue list — don't finish it in one pass, keep cycling

Don't try to fill in the entire issue list (Excel template) that organizes AS-IS / TO-BE / GAP in the first discovery session. Leave the unfilled cells as "?" and fill them in the next alignment. The people who try to finish in one pass are the ones who fill cells with their own guesses and cause accidents.

Issue list (3-sheet structure):

  Sheet 1: AS-IS list
    - Task name / owner / frequency / effort / pain point
  Sheet 2: TO-BE list
    - Ideal state / source (who said it) / confidence
  Sheet 3: GAP list
    - TO-BE No. / AS-IS No. / difference to clear / rough scale

The reflex that doesn't stop at "that sounds tough"

The most common failure in field discovery is "empathizing right after hearing an issue and then moving straight to the next topic." Whether you can reflexively return the next two questions the moment you hear "it's tough" determines the density of your discovery.

• ─① "How many minutes does it take each time? About how many times a month?" (quantify the burden)
• ─② "Ideally, how would things look so it helped you?" (draw out the seed of TO-BE)
• ─If you record it as just "tough," no information remains when you read it back later. Always keep it as a pair of a number and a TO-BE

Conditions for handing off to estimating

Move on to estimating without all three lined up and the rationale for the line items disappears. The good/better/best three options covered in the next course, "How to Estimate," can't be laid out unless GAP is lined up. Always confirm all three are lined up before handing off to estimating.

The discovery lead is a "translator"

The discovery lead is not someone who records the client's voice. They're a translator who organizes the client's words (fragmentary, skewed by pattern) into the three of AS-IS / TO-BE / GAP and returns them. The moment you can return them lined up as three, the client's own understanding aligns too. That's the true starting point.