Resources›Operational Walkthrough
Discovery FrameworkAI Variant

Find where AI actually fits
in your prospect's workflow.

Stop demoing AI features nobody asked for.

The Operational Walkthrough is a 5-phase structured discovery process for AEs and SEs - built around the buyer's process, not your feature list.

Read time 9 minFor AEs & SEsUse it pre-demo

Measures of success

60min
Ideal duration
5 - 7
Process steps
3
Pain points
80%+
Prospect talk time
The OpWalk difference

Demo what moves the needle. Skip what doesn't.

Your prospect's workflow will look different. The principle is the same.

01Intake / triage02Assign / investigate03Escalate04Enrich05Validate06Resolve07Document / Close01Intake / triage02Assign / investigate03Escalate04Enrich05Validate06Resolve07Document / Close

← scroll to see full flow →

STEP
Where AI moves the needleHigh volume, consistent input, clear output. Demo these.
STEP
Not the focusLow AI signal. Acknowledge and move on.

The OpWalk is operational discovery and just an entry point.

The Operational Walkthrough surfaces how your prospect's team actually works. That is necessary to contrast with a better future state, but not sufficient to win a deal.

Prospects who see value in your product still need a reason to buy it now. Continue building decision confidence through Value Discovery - understanding the Key Business Results they are accountable for, and the KPIs they use to measure progress. That is where deals close.

5 phases
Phase 1

Context Setting

10 min

Understand who you are talking to, what they own, and which process is worth mapping in detail. You are choosing the right thread to pull - not pulling all of them.

Questions to ask
Walk me through what your team is responsible for - end to end.
Where does the process start - what triggers it?
Which process would have the highest impact if it ran faster or more consistently?
How many people are involved in this process? Across which teams?
Failure mode

The prospect jumps straight to a specific tool or pain point and you follow them - skipping the process map entirely.

Do this insteadRedirect with: "Before we go there - help me understand the full picture of what your team owns. I want to make sure we're looking at the right part of the workflow."

Phase 2

Current State Map

15 min

Build a shared, accurate picture of how the work flows today - step by step, system by system. This map becomes the foundation for every AI conversation that follows.

Questions to ask
Walk me through exactly what happens from trigger to output.
What does the handoff look like between teams at that point?
What systems or tools are involved at each step?
Where do things slow down - or stop - most often?
Is this the same every time, or does it vary by case?
Do you need different workflows in other geographies?
Failure mode

The prospect describes how things should work, not how they actually work today. You end up mapping a fiction.

Do this insteadAnchor them with: "Walk me through the last time this actually happened - what did you do, step by step?" Real instances beat general descriptions every time.

Phase 3

Process Deep Dive

20 min

Surface the AI signal data hidden inside each process step - volume, consistency, input quality, and decision complexity. What looks like a workflow conversation is actually a readiness assessment.

Questions to ask
Who else is involved at this step?
How often does this happen - and at what volume?
Where does this information come from? How consistent is it?
Does this step require human judgment, or is it mostly rule-based?
Are there dependencies on other teams?
What would you do if this constraint didn’t exist at all?
What to capture vs. what to ask

Listen for

  • Who else is involved at this step?
  • How often does this happen - and at what volume?
  • Where does this information come from? How consistent is it?
  • Does this step require human judgment, or is it mostly rule-based?

Write down

  • Tools and systems used at each step
  • Manual, repetitive steps - high-volume tasks are AI candidates
  • Handoffs between teams - where data changes hands
  • Wait times and delays - especially approval or review bottlenecks
  • Data re-entry and format inconsistency - signals poor AI readiness
  • Frustrations - around speed, scale, or consistency
  • Tasks done the same way every time - strongest AI signal
Failure mode

You get surface answers - "it varies," "depends on the case" - and move on. The AI readiness data never materialises.

Do this insteadStay on the step. "When it varies, what does that look like? Give me the two or three most common versions." Specificity is the whole point of this phase.

Phase 4

Pain Point Deep Dive

10 min

Turn a surface-level frustration into a quantified business problem. You need frequency, impact, and stakeholder breadth before you can build a credible business case or a focused demo.

Questions to ask
What happens immediately upstream and downstream of this step?
When this step is delayed or wrong, which downstream processes also break?
In a typical month, how many times does this happen?
What’s the variance between best and worst case?
Is the input consistent enough to automate, or does it vary too much?
What’s the impact when this goes wrong?
What is the financial exposure of failure here - revenue risked, penalties, or cost of workaround?
When this comes up in your leadership meetings, who raises it - and how is it framed?
Who would need to sign off on a change to how this works - and which of them would object?
Who in your organization would need to support a change to how this works?
Failure mode

The prospect can't quantify. You accept "it takes a while" or "it happens a lot" and move on. The business case stays vague.

Do this insteadEstimate together. "If you had to guess - ten times a month? Fifty? And when it goes wrong, are we talking hours of rework or days?" A rough number is more useful than no number.

Phase 5

Wrap-up & Next Steps

5 min

Leave the prospect feeling heard, and leave yourself with a clear brief. A confirmed summary prevents misalignment in the demo stage. A committed next step prevents the deal from going quiet.

Close the loop

Summarize

Recap what you learned. Confirm you understood correctly before proposing anything.

Validate pain points

So the top challenges are: 1) … 2) … 3) … Did I get that right?

Next steps

Schedule follow-up. Ask who else to include. Set expectations on what comes next.

Failure mode

You summarize, the prospect agrees, and you close without a specific committed next step. The deal goes quiet because nobody owns what happens next.

Do this insteadName it explicitly. "So the next step is X - I'll send the summary by end of day. Can we get thirty minutes in the calendar for [specific person] by [specific date]?" Vague agreement is not a next step.

A demo built before a walkthrough is a guess.

A demo built after a walkthrough is a response to a problem the prospect has already told you matters.

Don't guess.
The point of view
This is such a consultative approach, and our prospects love it! They open up in a way they don’t in a typical discovery call. This is my go-to approach before validation demos now.
Senior Sales EngineerEnterprise SaaS
What to look for

AI Readiness signals & risks

Sort what you hear into two columns as you go. By the end of the call, you should know which signal each step belongs to - and whether this is a deal that's ready for a demo.

+Strong signalGreen-light a demo

Strong AI candidate

High volume, consistent inputs, rule-based steps, clear output. The process runs the same way every time.

Transformation opportunity

Something they can’t do at all today - scale, speed, or language coverage - not just doing the same thing faster.

Executive alignment

The pain connects directly to a KBR the CEO or board is already tracking - revenue, cost, or risk. Someone at the top has a reason to care.

!Risk signalPause before demoing

Weak AI candidate

Low frequency, high variability, heavy editorial judgment, or data that is inconsistent and unstructured.

Data readiness risk

Inconsistent taxonomy, unstructured content, or data spread across disconnected systems. AI will not fix a messy foundation.

Change readiness risk

Strong process ownership by one person, previous failed automation attempts, or no executive sponsor for change. Without a sponsor, the deal will stall.

In-room habits

Good practices

The walkthrough is a buyer-centric conversation. These habits keep you on that side of the table - and out of demo-mode reflexes.

01

Don’t pitch AI

Resist the urge to map every pain point to an AI feature. Some processes aren’t ready. Say so.

02

Use their language

They won’t say “automate.” Listen for “faster,” “consistent,” “scale,” “always manual.” That’s where AI fits.

03

Watch for gaps

Note what they don’t mention. A team not talking about scale often hasn’t imagined what’s possible.

04

Ask “What if”

“What would you do if this step happened instantly at any scale?” surfaces transformation, not just efficiency.

05

Read emotional cues

Sighs and pauses around repetitive tasks signal AI opportunity. Hesitation around change signals adoption risk.

06

Qualify the use-case

Volume + consistency + clear output = strong AI fit. One of those missing = a conversation before a demo.

Remember, don't stop at operational discovery!

Operational clarity opens the door. Use what you learn here to drive Value Discovery conversations with managers and executives - their KPIs and KBRs are what turn a compelling demo into a signed deal.

When you walk their workflow with them [your prospects], they start connecting dots themselves. I’ve had prospects hand me 80% of the business case before my SE and I demoed to their team.
Enterprise AEEnterprise SaaS
Overheard in discovery

Red flags worth pausing for

If you hear any of these, slow down. Each one warrants a conversation before you build a single demo slide.

We usually just use Excel for this
Lots of copy-paste between systems
I have to check with [person] first
This part takes forever
The data is never quite consistent
Every case is a bit different
We’d need IT to sign off on any changes
We tried automating this before…
After the call

Post-interview actions

The walkthrough is only valuable if you act on it. Four moves between the call and the demo.

1

Document the flow

Capture the process map and flag each step as strong, weak, or unclear AI candidate.

2

Send the summary

Thank you note with your initial read on where AI fits - and, importantly, where it doesn’t.

3

Tag in CRM

Log AI readiness signals: volume, data quality, integration constraints, change readiness.

4

Build the demo brief

Map the 2–3 strongest AI use-cases to specific capabilities. That’s your demo. Nothing else.

We ran an Operational Walkthrough with a prospect and identified exactly where they were losing time. By the end of the call we were already in a financial conversation. No demo, no pitch. The deal moved faster than anything else in our pipeline that quarter!
Senior Sales EngineerEnterprise SaaS

Take it into your next meeting

Stop winging discovery. Run the OpWalk.

Five-phase structured template for immediate use.

FigJam

Five-phase board, AI signals legend included.

Download FigJam template

Is your team struggling to position your AI capabilities?

Let's talk through it.

A focused one-on-one for sales leaders. When discovery is built around your prospect's processes - not your feature list - your demos land differently. Prospects arrive at next steps with confidence, not hesitation.

What you'll leave with

  • Clarity on where your team's discovery is leaving decision confidence on the table
  • Clarity on which changes to make - and in what order
  • Clarity on whether structured, reinforced coaching is the right decision for your team
Loading booking calendar...