Discovery workshops: what they are, how to run one, and what they should deliver

A discovery workshop is a facilitated working session where business stakeholders, technical specialists, and, when needed, people who talk to customers every day agree on four things: what the problem is, why it’s worth solving, what constraints exist, and where to start. It happens before you approve a quote, choose a vendor, or hire someone. It’s not an extended brainstorming session. You leave with documented decisions and a plan a team can put into action on Monday. This guide covers when a workshop makes sense, who to invite, how each stage works, and what the final document should include.

At a glance

• A discovery workshop turns ambiguity into an actionable plan, with goals, priorities, constraints, and first steps put in writing.
• It works when the room includes people who rarely talk to one another: management, technical teams, marketing, sales, and customer support.
• Formats range from half a day for a narrow use case to one or two days for complex projects.
• The output isn’t a slide deck: it’s a roadmap, a prioritized backlog, user stories with acceptance criteria, and measurable KPIs.
• For artificial intelligence projects, the workshop needs to assess data availability, financial returns, and AI Act requirements right away.

Contents

• What a discovery workshop is—and isn’t
• When it makes sense
• Who needs to be in the room
• How it works, step by step
• Half a day or two days? Comparing formats
• What you should leave with: the deliverables
• Discovery for AI projects: data, returns, and the AI Act
• Mistakes that derail a workshop
• Our perspective
• From the plan to the right team
• Frequently asked questions
• Sources

What a discovery workshop is—and isn’t

“Discovery” refers to the exploratory phase before design begins. The workshop is where that exploration happens together, in a structured way, with a facilitator guiding two alternating modes: first broadening the discussion (problems, ideas, perspectives), then narrowing it down (choices, priorities, trade-offs).

What sets it apart from an ordinary meeting is the outcome. A meeting produces minutes and a few good intentions. A discovery workshop produces an actionable document you can hand to a development team, a creative agency, or a group of freelancers—and use to support a budget decision.

Three things a discovery workshop is not:

• It’s not a vendor presentation. If someone spends half the time walking through their solution, that’s a sales demo.
• It’s not team building. The group leaves more aligned, but the goal is to make decisions, not bond.
• It’s not a handoff of predefined requirements. If the company arrives with a solution already chosen and just wants an estimate, the workshop becomes a quote in disguise.

In Italy’s public sector, the same principle has a specific name: co-design. The Designers Italia operational manual describes it as a way to gather perspectives from everyone involved in a service and surface problems before designing. The AgID design guidelines describe co-design workshops with a diverse group: designers, users, and representatives from public agencies and private companies. The same logic applies to a business.

When it makes sense

Not every project needs a workshop. If you’re redesigning a flyer with a clear brief, the brief is enough. A workshop pays for itself in these situations:

• Different departments describe the same problem in different ways. Marketing says, “We need more leads,” sales says, “The leads are low quality,” and support says, “Customers leave after their first order.” Before investing, you need to understand the real problem.
• The first step isn’t clear. You have five possible initiatives and the budget for one.
• The project involves technology that’s new to the company, such as an AI-powered assistant or a marketing automation platform.
• You’re about to bring in outside help: an agency, freelancer, or consultant. The workshop puts your expectations in writing, making it possible to compare proposals meaningfully.
• A previous project stalled and no one can explain exactly why.

If the problem is clear and the scope is narrow, a lighter approach is more than enough: a well-written creative brief or, for brand identity, a brand sprint, with an even tighter timeline and more focused questions.

Who needs to be in the room

A workshop’s strength lies in bringing together people who rarely cross paths at work. An effective group has 5 to 10 people: enough to cover the key perspectives, but small enough for everyone to speak.

Support and sales can join for just part of the session. They’re closest to the market, and their presence helps guard against confirmation bias: the project team’s tendency to look only for evidence that supports its original idea.

When there are many participants and opinions diverge, establish in advance who owns each decision and where it will be recorded. A useful reference is this guide to stakeholder alignment meetings, built around one owner and one record for each decision.

How it works, step by step

Every facilitator has their own approach, but the sequence that works is almost always this one.

1. Preparation (one or two weeks ahead). Conduct brief interviews with participants, gather existing data (analytics, support tickets, sales reports), and define the workshop’s central question. A good question is specific: “How do we reduce churn after the first purchase by 20%?” works better than “How do we improve the customer experience?”
2. Opening and goals (30 minutes). Clarify what needs to be decided by the end of the day and what falls outside the scope.
3. Map the current state. Processes, tools, workflows, and pain points. Use a physical or digital whiteboard and write everything down.
4. Define the problems. Group the symptoms you’ve identified and reframe them as challenges. This is where you separate cause from effect.
5. Generate solutions. Now—and only now—start exploring ideas. Everyone works individually first, then shares, so the most talkative person doesn’t dominate.
6. Prioritize. Evaluate each idea by impact and effort using a simple matrix. Keep the list short: no more than three initiatives.
7. Metrics and constraints. For each selected initiative, define KPIs, an estimated budget, timing, technical dependencies, and regulatory requirements.
8. Wrap-up and next steps. Who does what, and by when. The final document usually arrives within a week.

The rule behind it all: a workshop is for making decisions, not collecting ideas. If you haven’t ruled anything out by the end of the day, you haven’t really made a choice.

Half a day or two days? Comparing formats

Duration depends on complexity, not ambition. An overly long session on a simple problem wears the group out; an overly short one on a complex project produces a superficial plan.

Remote workshops work well when split into short blocks: a full day on a video call reduces the quality of the final hours. In person, a single-day session remains the most energizing format.

What you should leave with: the deliverables

A workshop is judged by the document it leaves behind. At a minimum, it should contain:

• Problem statement: one or two sentences everyone agrees on, including the scope.
• Goals and KPIs: measurable numbers, with a baseline and a target.
• Stakeholder and dependency map: who’s involved, who decides, and which systems the project touches.
• Roadmap: the phases in order, with estimated timelines.
• Prioritized backlog: the list of tasks, from most to least important.
• User stories with acceptance criteria: for digital projects, a description of what users need to be able to do and how you’ll know it works.
• Risks and assumptions to validate: what you don’t yet know and how you’ll find out.
• Required expertise: the roles needed to execute the plan and the skills missing in-house.

That last point is the one most often forgotten. A plan tells you what to do, but without saying who will do it, it stays on paper. A portfolio approach helps here: if the workshop produces several initiatives, evaluate them together using shared criteria, as this guide to an innovation project portfolio explains. That keeps you from funding everything a little and nothing fully.

Discovery for AI projects: data, returns, and the AI Act

Today, a growing share of discovery workshops starts with a specific request: “We want to use artificial intelligence.” In Italy, that momentum is real. According to Istat’s 2025 Enterprises and ICT survey, the share of companies with at least 10 employees using AI rose from 5% in 2023 to 8.2% in 2024 and 16.4% in 2025. The same release, however, reveals a widening gap: the difference between large companies and small and midsize businesses has grown to 37 percentage points. Among the obstacles smaller businesses report, skills shortages and regulatory uncertainty stand out.

The most common application of AI is in commercial activities: the Istat statistical report identifies marketing and sales in 33.1% of cases. We explore what’s changing in practice in this article on AI for marketing in 2026.

In an AI project workshop, three checks need to happen right away, before discussing models or vendors:

1. The business problem. “We want a chatbot” is a solution, not a problem. The problem might be “40% of support requests are about order status.” From there, you can determine whether a chatbot is needed or a well-designed tracking page would be enough.
2. The data. Does it exist? Is it structured and accessible? Can it be used in compliance with the GDPR? A lack of organized data is often the first obstacle that surfaces in the room.
3. The return. How much time will the project save, or how much revenue will it generate? If you can’t estimate that even roughly, the project isn’t ready.

Then there’s the regulatory framework. The European Union’s Artificial Intelligence Act entered into force on August 1, 2024, and classifies systems by risk. High-risk systems, including those used in recruitment, must meet requirements for data quality, transparency, and human oversight. Under the Digital Omnibus package revision, enforcement of many of these requirements is being pushed back to December 2, 2027. A project that starts today and goes into production a year from now needs to build compliance in from the workshop stage, not add it later. If your use case involves recruiting, credit, healthcare, or public services, bring your DPO or data protection lawyer into the room and have the system’s classification checked before approving the budget.

For small and midsize business leaders starting from scratch, this overview of AI and project management in small and midsize businesses helps distinguish tools that are ready to use from those still in the experimental stage.

Mistakes that derail a workshop

• Jumping to the solution. The group starts discussing which software or model to use before defining the problem, KPIs, and constraints. It’s the most common mistake—and the most expensive.
• Inviting only management. Without the people who work within the process, the current-state map becomes an idealized version of reality.
• Skipping preparation. Without data and preliminary interviews, half the day gets wasted piecing together information that already existed.
• Too many priorities. Leaving with ten “important” initiatives is the same as leaving with no priorities at all.
• A biased facilitator. If the person leading the session has an interest in selling a specific solution, the workshop will tend to steer toward it.
• A late final document. If the report arrives a month later, the group’s momentum is already gone. One week is a reasonable limit.
• No one responsible for execution. The plan exists, but no one is tasked with moving it forward. This is where a freelance project manager or a fractional manager can make a difference.

Our perspective

In our work with companies and startups, we often see the same pattern: a good workshop, a clear plan, and then weeks lost searching for the right people to execute it. The document calls for a performance marketer, a designer, and a developer who can work together. But finding them individually, vetting their skills, and turning them into a team becomes a second project—often longer than the first.

That’s why we believe discovery should close with one more question: who, specifically, will carry out this plan? Listing roles isn’t enough. Cultural fit with the company matters. So does a shared working language. And someone needs to coordinate the work so the company can focus on strategy instead of managing five separate freelancers.

It’s why we carefully hand-pick talent, one person at a time, and build custom teams rather than offer a standard roster. A plan built through thoughtful discovery deserves a team chosen with the same care.

— Marco

From the plan to the right team

If your workshop produced a roadmap and you’re now missing the expertise to execute it, that’s where we can help. Through Talent as a Service, HOOOP Studio, and Fractional Management, we provide hand-picked professionals in marketing, design, branding, content, and digital development, organized into a custom team with centralized operational management. If you need part-time senior leadership to move the plan forward, read about the benefits of a fractional manager for small and midsize businesses.

To figure out where to start, begin with our business assessment: the first step toward clarifying your goals and the expertise you need.

Frequently asked questions

How long does a discovery workshop last?

It depends on the project’s complexity. A single use case or process generally takes 2–4 hours, while a new digital product or a project involving multiple departments needs one or two days, potentially split across several sessions. Preparation matters more than duration.

Who should facilitate the workshop?

An experienced facilitator with no vested interest in pushing a specific solution. They can be internal, provided they have the authority to keep management on track too, or external. What matters is that their role is to guide the process, not make decisions for the group.

What’s the difference between a discovery workshop and a design sprint?

A discovery workshop helps you understand which problem to solve and what to prioritize. A design sprint starts with a problem that’s already been chosen and produces a prototype to test with users within a few days. The first often sets up the second.

Can you run a discovery workshop remotely?

Yes, with a shared digital whiteboard and the session split into blocks of about two hours. Asynchronous work beforehand, such as filling out a process map or answering brief interview questions, becomes even more important. A full day on a video call significantly reduces the quality of the final hours.

What should my company prepare before the workshop?

The data you already have (analytics, sales reports, support tickets), a list of participants and their roles, and an initial version of the question you want to answer. If the project involves AI or personal data, also bring information on where the data is stored and who is responsible for it.

Sources

• Istat, Enterprises and ICT, 2025 (press release)
• Istat, Enterprises and ICT 2025 Statistical Report
• European Commission, AI Act enters into force
• Key4biz, postponement of AI Act requirements for high-risk systems
• Designers Italia, Operational Design Manual: co-designing solutions
• AgID, Design Guidelines for Digital Public Services

Recommended reading

• Brand sprints: what they are, how they work, and when you really need one
• Content workflows: how to create, review, and approve without the chaos
• How to write an effective creative brief: structure, examples, and checklist