What to Automate During the First 100 Days After an Acquisition

A new owner sees spreadsheets, inboxes and manual workarounds and understandably wants to modernize them. But those workarounds often contain undocumented customer promises, employee knowledge and exception handling. Replacing them before understanding them can destroy the very cash flow the buyer acquired.

Days 1-30: stabilize and listen

Do not lead with a software rollout. Confirm payroll, banking, insurance, backups, access, customer commitments and critical deadlines. Meet employees and major customers. Shadow how orders become cash.

Good early automation is administrative and reversible:

• Calendar reminders for statutory, customer and supplier deadlines.
• Backup verification and access-review tasks.
• Daily cash, receivable and service-backlog visibility.
• A transition issue log with owners and due dates.

Avoid changing CRM, accounting, pricing, scheduling or customer communication solely because the existing tool looks old.

Days 31-60: map the real process

Map five flows: lead-to-cash, purchase-to-pay, service delivery, customer issue resolution and monthly financial close. For each step record owner, system, input, output, exception and failure consequence.

Now identify repeated friction:

• Duplicate entry between inbox, spreadsheet and accounting system.
• Missing information that creates back-and-forth.
• Approvals waiting on the former owner.
• Tasks remembered by one employee.
• Customers surprised because status is invisible.
• Reports assembled manually but rarely acted upon.

Ownership change is immediate in employees' and customers' eyes. That makes the first weeks a poor time for broad experimentation and an excellent time for careful observation.

Days 61-100: test one meaningful improvement

Choose a workflow that is frequent, measurable, low-consequence and owned by an employee who wants it fixed. Typical candidates:

1. Standardized lead intake and routing.
2. Missing-document reminders.
3. Quote assembly from approved components.
4. Accounts-receivable reminder queues.
5. CRM follow-up tasks.
6. Internal weekly exception summaries.
7. Document classification with human review.

Run beside the current process. Use a test set. Record exceptions. Require human approval before external actions. End the pilot with a written continue, revise or stop decision.

Worked example: service scheduling

What not to automate in the first 100 days

• Employee performance or termination decisions.
• Credit, pricing or refund authority without controls.
• High-value supplier selection.
• Customer complaints with legal or safety implications.
• Processes nobody can explain.
• Anything dependent on unreliable data that has not been reconciled.
• A full system replacement without a tested migration and rollback plan.

The sequencing matters: document reality first, improve it second and automate stable work third. That is how systems reduce owner dependence without destabilizing the relationships and judgment the business relies on.

Observe, stabilize, document, simplify, then automate

The order matters. Observe the actual workflow at the frontline. Stabilize payroll, billing, customer service and access. Document the current state, including exceptions. Remove unnecessary approvals and duplicate entry. Only then automate the smallest bounded step. A new owner who automates immediately can hard-code the seller's workaround before learning why it exists.

Every early automation needs a rollback

Name the person who can pause it, preserve the manual path and define the trigger: error rate, missed deadline, customer complaint or data-quality failure. Measure adoption as well as technical success. A workflow that functions perfectly but is bypassed by the team has achieved nothing except a software invoice.

Borrow mechanisms, not famous-company theatre

Cross-industry thinking is valuable when the underlying mechanism matches. Dyson's industrial-cyclone observation mattered because clogging was the shared constraint.

Three examples from The Book of Elon make the method concrete. Tesla looked at the way inexpensive toy cars could be cast as large pieces and asked whether a full-size vehicle could eliminate assemblies the same way. Musk used LEGO's inexpensive but precise fit as evidence that low cost need not mean loose standards. SpaceX questioned disposable rockets by comparing them with reusable airplanes, cars and boats. The lesson is not to imitate cars or rockets. It is to borrow the mechanism: combine steps, define a visible tolerance, or challenge an expense the industry has learned to accept.

For a maintenance company, that could mean one standardized intake replacing three re-entries, completion photos with defined tolerances instead of “looks fine,” and reusable diagnostic histories rather than starting every service call from zero.

A plumbing company might apply remote visual triage from telehealth or technical support: ask the customer to show the leak safely by video, identify the likely skill and parts required, and reserve an emergency visit when risk is high. The goal is not to diagnose recklessly; it is to arrive prepared and reduce avoidable travel.

After the repair, the same system can record asset type, age and recommended maintenance, then offer a relevant follow-up rather than a generic promotion. That turns one emergency call into a permission-based service relationship. In the first 100 days, test this with one technician and one call type before changing dispatch for everyone.

General operational information only. Acquisition transitions may require legal, privacy, employment, cybersecurity, accounting and industry-specific advice.