K
KowalAI

GuidesStop rebuilding

Automate client onboarding without the email chase

Client onboarding breaks when sequence and visibility are missing, not when you lack a branded portal. Write the path from signed to kickoff — intake, countersign, pay, assets, internal handoff, first working session — and put status where the team can see it without searching email. Automate reminders and checklists after the stages are named. Buy a portal only when that path is stable and the remaining pain is the client experience, not your own confusion. The $997 assessment can rank whether onboarding is the first leak. It does not include building the portal.

Who this is for

This playbook is written for agencies and service firms: studios, consultancies, implementation shops, retainers with a kickoff, and any owner-operated company where “you’re signed” and “we have started” are weeks apart. The chase is familiar — W-9, brand assets, login, deposit, kickoff agenda — and it lives in threads that only the account lead can parse.

If the pain you feel is the packet itself, start at /document-workflows. If the pain is the quiet after a good sales call, read pipeline follow-up first. Onboarding starts when the commercial yes exists and the delivery yes has not.

Symptoms

If several of these are true in the same month, this is a live operational leak — not a tooling preference.

  • “Are they onboarded?” is a Slack question, not a field the team can see.
  • Intake forms arrive half-complete and restart in email with a different file name.
  • Countersign, deposit, and asset collection happen in an order nobody wrote down, so each client is a custom chase.
  • The person who sold the work dumps a folder on delivery and disappears into the next deal.
  • Kickoff gets booked before payment or access is real, then slips, and the client thinks you are disorganized.
  • A portal was bought last year. The team still emails the checklist because the portal stages do not match the work.
  • New team members ask the same six questions every time a client is signed.

What done looks like

Done is a working operating change, not a purchased seat or a dashboard nobody opens.

  • Stages are named and shared: signed, intake, countersign, pay, assets, handoff, kickoff.
  • Each stage has an owner, a blocker rule, and a definition of complete.
  • Anyone on the account can answer “where is this client?” without opening the founder’s inbox.
  • The next three clients can run on a checklist even if the usual coordinator is out.
  • Reminders go out from the stage, not from memory.
  • A portal, if it exists, mirrors those stages. If it does not, it is not the system of record.

Sequence and visibility first

Automation that emails a client “please see the portal” while the internal path is still tribal knowledge creates a second chase. The first job is to write the sequence the business already uses on a good week, then make the current stage visible. Visibility is a board, a CRM field, or a single running doc. Pick one. The tool is less important than the rule: no stage exists unless a person who did not sell the deal can find it.

Sequence is not a slogan. It is the order of blockers. Do not book kickoff as if it were a courtesy if you cannot work without access or a deposit. Do not collect fifty asset types before the statement of work is countersigned if the scope still moves. Write the dependency, then the reminder. Reminders without dependencies are how clients get nagged for things you are not ready to receive.

The stages: signed to kickoff

Signed

Commercial yes is real: a signed proposal, a verbal yes with a written confirmation, or a PO — whatever you already treat as permission to spend delivery time. If “signed” is fuzzy, onboarding will stay fuzzy. This is also where leftover sales work belongs: the notes that should become the kickoff brief. That merge is call notes into a document, not a new onboarding tool.

Intake

Ask only for what the first two weeks require. A form that asks for the company’s life story gets abandoned. Name the fields, the owner who reviews them, and the rule for “good enough to proceed.” Incomplete intake is a stage, not a personal failure.

Countersign

If you need a countersigned SOW, MSA, or change note, it is a stage with a link and a status. Do not hide it inside “paperwork.” The person chasing it should be able to see whether it is with you, with them, or with legal.

Pay

Deposit or first invoice. Money has a high blast radius. Drafts and reminders are fine. Auto-marking a client “paid” from an emailed “we sent it” is not. Tie the stage to the billing system you actually use, even if the first version is a human checking the inbox against a field.

Assets

Logins, brand files, analytics access, domain records — the minimum set for kickoff to be real. List them per service line. A branding client and a media buyer do not share an asset list. Store the receipt of each item, not a vibe that “they sent stuff.”

Handoff

Internal. Sales tells delivery what was promised, what was deferred, and who the client thinks their day-to-day is. A fifteen minute structured handoff beats a folder named “NEW.” If this stage is skipped, kickoff becomes a second sales call.

Kickoff

The first working session with an agenda that assumes the prior stages are complete. If they are not, kickoff is not late — the prior stage is blocked. Saying that out loud to the client is kinder than a vague delay.

DIY checklist before a portal

Run the next three clients on a written checklist. Same stages, same owner names, same “complete” tests. Put status in the CRM or a shared board. Use the email and e-sign tools you already pay for. Measure only this: can someone else answer where the client is, and does the chase still depend on one memory.

A portal earns a place when the path is stable, clients repeat, and the remaining pain is them hunting for files or you re-sending the same links. It does not earn a place because a vendor showed a white-label login. If the stages still change every sale, a portal will freeze last quarter’s confusion and call it experience.

Document packets — welcome PDFs, kickoff decks, access checklists — should be templates with a review pass, not a custom essay each time. That work sits with document workflows. Do not wait for the portal to start using a reviewed template.

Why the email chase survives good intentions

Onboarding threads mix five jobs: collecting facts, collecting money, collecting access, scheduling humans, and soothing a client who is already paying and has seen nothing. When those jobs share one inbox conversation, every reminder looks like nagging and every reply looks like progress. Split the jobs in the status view even if the client still receives ordinary email.

The chase also survives because sales and delivery do not share a definition of “ready.” Sales hears a yes and books a ceremonial kickoff. Delivery needs access and a deposit. Write the ready test as a sentence both sides can repeat. If they cannot repeat it, you do not have a sequence. You have two calendars.

A third reason: assets are treated as a pile. “They sent stuff” is not a stage. Name the minimum set for the service line. A missing analytics login is a blocker. A nice-to-have brand video is not. Clients complete lists they can finish. They abandon lists that look like a scavenger hunt.

Who owns which stage

Sales can own signed and the commercial handoff. A coordinator can own intake, countersign, pay, and assets if those are checkable. Delivery owns kickoff quality. The founder should not be the default owner of every stage unless the firm is still one person — and even then, the stages should be written so a second person could run them.

Escalation is a rule, not a mood. If pay is blocked for more than one reminder cycle, a named person calls. If assets are blocked, delivery does not pretend kickoff is on track. If scope moved after signed, send the work back to a change note instead of silently stretching onboarding. Quiet stretching is how kickoff becomes a second sale with none of the pricing.

Write coverage. When the coordinator is out, the board should still show the stage and the last ask. That is the test the DIY checklist is for. If only one person’s laptop can answer “where is this client?”, you have not onboarded the process. You have onboarded a hero.

When a portal is still the wrong buy

Buy a portal when clients repeat, the stages have been stable across several jobs, and the remaining pain is them finding files or you re-sending the same links. Do not buy one because the current week feels chaotic. Chaos is a sequence problem.

Skip the portal when every sale is a custom scope, when legal still rewrites the path, or when the team will not update a board they already have. A portal that the team bypasses is worse than email: it creates a second place the client might have uploaded the thing you are still chasing.

If you already own a portal, map it to the written stages or stop using it as the system of record. Half-migrated portals are how status questions return to Slack. The live question is not “do we have a portal?” It is “can someone who did not sell the deal see the stage?”

DIY vs hire

DIY the sequence, the status field, and the three-client checklist. Most owner-operated firms can do this with current software. The hard part is agreeing that kickoff is not a courtesy calendar hold.

Hire for the later build: a client portal, billing write-back, automated access provisioning, or an agent that chases assets across email. Those are separately scoped. The $997 assessment can tell you whether onboarding belongs in the first two to four DIY items or on the later roadmap. It will not implement the portal inside that fee.

Controls before automation

Recommendations that touch customers, money, contracts, private data, or systems of record need human review, limited permissions, a log, and a fallback. Do not automate a broken path because the tool is ready.

  • Do not mark paid, signed, or access-complete from a model’s guess. Read the source system or a human check.
  • Client-facing reminders stay drafts until the stage and the ask are correct.
  • Store credentials in the password tool you already trust, not in the onboarding chat.
  • If scope is still moving, freeze onboarding at intake and send the change back to sales.
  • Keep an email fallback. Portals go down. Threads should still show the last clear ask.
  • Limit who can skip a stage. Skips should be visible, not folkloric.

The first three clients are the test

Do not judge the checklist on the messiest account you already have. Run it on the next three signed clients. If stage names keep changing, you are still discovering the path. If the same blocker appears three times, write it into the sequence. If a stage is never blocked, it may not be a stage.

After three, you will know whether a portal would remove work or add a login. You will also know whether sales is still booking kickoff as a courtesy. That evidence is enough to rank onboarding against the rest of the week.

Related ranking

Onboarding is one room in the fourteen mid-market AI asks. If every new client also reopens “what did we promise,” you have a proposal and knowledge problem upstream. Rank those together rather than buying a portal to hide them.

Soft next step

If ranking this work would help, start with the assessment.

$997 is a live consultation and a written analysis. You leave with 2–4 improvements you can run yourself, plus a later roadmap quoted only if you want us on it. Implementation is not included.