Back to all articlesCustomer onboarding

Run an onboarding kickoff that ends with decisions, not a presentation

Alactic Editorial10 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

A kickoff can end with everyone understanding the product and nobody knowing whether implementation can begin. The customer expects a broad rollout. The implementation lead expects a small pilot. A manager has offered to help, but nobody has accepted responsibility for preparing the customer’s data.

Run the kickoff around a decision worksheet rather than a presentation. Confirm the intended outcome, settle the choices that affect the first implementation step, verify who can approve and complete the work, and record the next commitment with evidence of completion. Use slides only when they help someone make one of those decisions.

The practical test is simple: after the meeting, could someone who missed it explain what will happen next, who owes what, and what would stop the work? If not, the kickoff produced context rather than an implementation agreement.

1. Define the decisions the kickoff must produce

Start with the implementation work you want to unlock. Then identify the choices required to begin it. This keeps the kickoff from becoming either a product tour or a second discovery call.

A useful kickoff decision set covers:

  • Intended outcome: What should become observably different in the customer’s operation?
  • First scope: Which workflow, team, or use case will be implemented first, and what is excluded?
  • Readiness: Which inputs and approvals must exist before configuration or testing can proceed?
  • Responsibility: Who prepares, approves, performs, and checks each immediate deliverable?
  • Next commitment: What will be handed over or demonstrated next, under what conditions?

Do not force every implementation question into this meeting. A choice belongs in the kickoff if leaving it unresolved would change the first scope, block the next task, or leave acceptance unclear. Detailed training preferences and later expansion ideas can wait unless they affect that immediate work.

Where a choice is reversible, a bounded provisional decision may be enough. Where changing it would require substantial rework, identify the necessary approver before treating the choice as settled.

2. Send a pre-read that exposes gaps, not just background

A useful pre-read gives the customer something to confirm or correct. An overview deck explains what your team knows; a decision pre-read reveals what still needs agreement.

Include a proposed outcome, a proposed first scope, the inputs already available, and the unresolved choices. State which attendees need to approve which decisions. Ask for corrections to assumptions rather than a general request to review the material.

Separate established information from proposed interpretations. A goal recorded during the commercial handover is context, not necessarily an implementation acceptance criterion. Likewise, a named customer contact is not automatically authorised to approve scope or commit another team’s time.

If the pre-read receives no response, do not treat silence as agreement. Bring the assumptions into the kickoff explicitly. If a required approver cannot attend, either obtain their decision in advance or limit the meeting’s authority: identify which choices can be made and which must remain pending.

3. Use this filled kickoff decision worksheet

Fictional worked example: The organisation, people, quantities, schedule, and acceptance criteria below are hypothetical. They illustrate a recommended operating design, not customer results or verified product capabilities. The example concerns a service business preparing a shared enquiry-handling workflow.

Purpose and proposed outcome

Customer: Cedar Service Group, a fictional business.
First workflow: New service enquiries handled by the weekday service desk.
Intended outcome: During the pilot review, the service lead can identify the responsible person and next action for every enquiry in the agreed review set without asking individual team members for an update.
Not an outcome: Completing a product demonstration or creating user accounts. Those are activities, not evidence that the workflow works.

Pre-read items and readiness

  • Workflow sketch: Received from Maya, the customer service lead. Shows receipt, assignment, follow-up, and closure. The handling of disputed enquiries remains unclear.
  • Proposed pilot roster: Four service coordinators identified. Maya still needs to confirm that each can participate in testing.
  • Sample records: Not yet received. Maya will provide ten fictional enquiry records rather than live customer information for the initial walkthrough.
  • Draft status definitions: Sent for review. The distinction between awaiting customer information and awaiting internal action needs a decision.
  • Authority: Leena, the operations sponsor, can approve first-phase scope. Maya can approve the workflow definitions. Omar, the implementation lead, can confirm implementation readiness after checking the proposed design against the available workspace and plan.

Unresolved choices brought to the kickoff

  1. Pilot boundary: Include only weekday service enquiries, or include weekend work too? Recommendation: weekday work first because the weekend handover owner has not been confirmed.
  2. Ownership rule: Is the team responsible collectively, or must each enquiry have a named responsible person? Recommendation: require a named person so the intended outcome can be checked.
  3. Waiting work: Use one waiting status or distinguish customer and internal dependencies? Recommendation: distinguish them because the next action belongs to different people.
  4. Acceptance authority: Can Maya accept the pilot workflow, or does Leena need to approve every detail? Recommendation: Maya accepts operational behaviour; Leena approves changes to scope.
  5. Start condition: Begin configuration immediately or wait for agreed definitions and sample records? Recommendation: allow preparation, but hold configuration that depends on unresolved definitions.

Decisions recorded during the kickoff

  • D1 — Scope confirmed: Weekday service enquiries only. Weekend handling and historical backlog are excluded from the first phase. Approved by Leena. Revisit weekend scope after the pilot review.
  • D2 — Responsibility confirmed: Each enquiry must have a named responsible coordinator. Maya resolves assignment disputes. This is a workflow requirement; Omar will verify how the available setup can support it before promising a configuration.
  • D3 — Waiting states confirmed: Distinguish awaiting customer information from awaiting internal action. Maya will provide a plain-language definition and a sample case for each.
  • D4 — Acceptance confirmed: Maya accepts the operational walkthrough. Leena is needed only if the proposed solution changes scope or requires additional customer resources.
  • D5 — Conditional start confirmed: Omar may prepare the configuration plan. Configuration dependent on workflow definitions begins only after Maya supplies and approves those definitions.

Open item, dependency, and next commitment

Open item: Who owns disputed enquiries when Maya is unavailable?
Decision owner: Leena.
Due: Tuesday at noon.
Effect if unresolved: Ordinary assignment examples can proceed, but the disputed-enquiry path cannot be accepted. No fallback owner has yet been authorised.

Customer commitment: Maya will send the ten fictional enquiry records, the status definitions, and the confirmed pilot roster by Tuesday at noon.
Implementation commitment: Omar will review the pack by Wednesday at noon and return either a readiness confirmation or a specific list of gaps.
Next checkpoint: Thursday’s workflow walkthrough remains provisional until readiness is confirmed.
Completion evidence: An agreed review sheet showing the responsible person, current state, and next action for each sample enquiry, with exceptions recorded.

This worksheet is intentionally narrower than a full implementation plan. It contains enough agreement to unlock the next piece of work without pretending that later phases are already designed.

4. Use meeting time to resolve differences

Open with the proposed outcome and ask the customer to correct it. Avoid asking only whether it looks right. Ask what evidence would make them say the first workflow is usable, and what could still be wrong even if the setup were technically complete.

Then move through unresolved choices. For each one, present the recommendation, the alternative, and the consequence. In the fictional example, excluding weekend work reduces initial scope but leaves a known handover outside the pilot. That is a trade-off to approve, not an omission to hide.

Use demonstrations selectively. A short illustration may help distinguish two workflow options. A broad tour that introduces unrelated possibilities is more likely to reopen scope than settle it. When a new request appears, ask whether it changes the intended outcome or belongs in a later phase.

Record decisions where participants can inspect them. Read back the practical consequence, not just the label: “Weekday enquiries are in scope; weekend handovers will not be tested in this phase.” This gives the customer a chance to catch a misunderstanding before it becomes work.

5. Verify ownership through deliverables and authority

A name beside a task is not sufficient. The person may lack the authority, time, inputs, or understanding needed to complete it.

For each immediate commitment, verify:

  • Deliverable: What specifically must be supplied or completed?
  • Owner: Who has accepted responsibility for producing it?
  • Approver: Who can say it is adequate, where approval is needed?
  • Dependency: What must happen first?
  • Completion evidence: What can the next person inspect before starting?

“Customer to provide data” fails this test. In the fictional worksheet, Maya owes a defined sample pack, and Omar owes a readiness response. Neither side has to guess whether an attachment alone means the dependency is cleared.

Keep this proportional. Do not build an elaborate responsibility matrix for work that needs a simple handoff. Add detail where a missed approval or ambiguous deliverable could stop implementation.

6. Recover when the kickoff cannot settle a choice

An unresolved choice is manageable when its consequences are explicit. It becomes dangerous when the recap quietly converts it into agreement.

If the approver is absent: Record the recommendation as pending, name the approver, and specify which dependent work must wait. Do not ask an attendee to approve beyond their authority just to preserve the schedule.

If the desired outcome is too broad: Narrow it to an observable first workflow. Keep the broader ambition as context, but avoid accepting an entire transformation as the first implementation test.

If a product capability is uncertain: Assign a verification task. Check the actual workspace and current plan before committing to a design. Record what the workflow requires separately from whether the available setup supports it.

If inputs are late or incomplete: Return a concrete gap list and revise the dependent checkpoint. Continuing with invented assumptions may produce visible activity while increasing rework.

If participants disagree: Capture the competing choices and their effects, then route the decision to the appropriate owner. A smaller authorised scope is often more useful than a broad agreement that different people interpret differently.

7. Send a recap that preserves the decisions

The customer recap should be short enough to use as a working reference. Separate confirmed decisions, open items, and commitments. Do not bury them in a chronological account of the meeting.

Hypothetical customer recap for the fictional worksheet:

We agreed that the first phase covers weekday service enquiries only. The pilot will test whether Maya can identify a responsible coordinator and next action for every enquiry in the agreed review set.

Confirmed: named responsibility for each enquiry; separate waiting states for customer information and internal action; Maya accepts the operational workflow. Weekend work and historical backlog are excluded.

Maya will send the ten fictional records, status definitions, and confirmed pilot roster by Tuesday at noon. Omar will confirm readiness or return specific gaps by Wednesday at noon.

Leena will name the backup owner for disputed enquiries by Tuesday at noon. That path remains unapproved until she decides.

Thursday’s walkthrough is provisional, subject to readiness confirmation. Please flag any correction to these decisions before dependent work begins; missing approval will remain pending rather than being inferred from silence.

8. Check whether the kickoff unlocked implementation

Measure the kickoff by the quality of the next handoff, not by whether the presentation finished. After the next checkpoint, inspect whether the agreed inputs arrived, whether they met the recorded criteria, and whether the next owner could proceed without reopening a supposedly settled choice.

Useful operating signals include decisions reopened because the record was unclear, commitments missed because ownership was ambiguous, and work started before its dependencies were accepted. These signals point to specific repairs in the worksheet rather than a vague need for better communication.

For your next kickoff, prepare one filled decision worksheet before creating slides. Leave each unknown visible. End the meeting by reading back the next deliverable, its owner, its completion evidence, and the condition that would delay it. The immediate result to look for is a handoff that the receiving person can either accept or reject for a stated reason.

Customer onboardingImplementationKickoff decisionsCustomer success

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications