I'm the sole IT person for a 10-person CPA firm that has operated on premises for nearly 30 years. Most employees are not very technical, and management wants a flat-rate proposal to move the environment to a bare cloud server rather than using an all-in-one hosted solution.
Microsoft 365 accounts and Exchange mailboxes are already configured. The project would include designing the cloud infrastructure, creating Purview DLP policies, setting up Intune, supporting QuickBooks and Drake, migrating about 4 TB of client files from a NAS, backing up and disposing of roughly 10 workstations, configuring VPN and RDP, provisioning roles and permissions, testing, documentation, user training, and general cutover support.
There is also no written WISP currently, although the firm is interested in creating one. Their QuickBooks environment includes Premier Accountant editions from 2019 through 2024, so version compatibility may need to be reviewed.
I was considering estimating around 100 hours, but I'm concerned that I'm underestimating the infrastructure, permissions, migration, compliance, and training work. For anyone who has handled a similar project, what kind of effort or budget would you plan for, and how would you structure the quote?
4 Answers
I wouldn’t offer a flat rate unless the scope is extremely well documented. There are too many unknowns here, especially around permissions, undocumented workflows, old software, data quality, and user training. An hourly project with a not-to-exceed estimate is safer. If you do use a fixed price, spell out assumptions and define change-order triggers for anything outside them.
If you want to present a number, separate the design and migration into a paid discovery phase first. Produce an inventory, dependency list, permission matrix, application plan, backup and recovery plan, and a detailed implementation estimate. Then quote the build using clearly stated assumptions, with hourly billing for scope changes.
For a small firm, a cloud desktop or hosted virtual-desktop design might land somewhere around a five-figure project depending on complexity, but the number should follow discovery. The customer is paying for a secure, supportable outcome—not just the hours spent copying files.
I’d also avoid promising a deadline before testing the applications and confirming the data structure. A staged pilot and user-acceptance period can prevent a painful cutover.
The data copy is only one part of the job. Budget separately for rebuilding permissions and access groups, because small firms rarely have a reliable record of who should access which client files. You may spend several days interviewing the partners before touching the technical migration.
Also verify the QuickBooks versions and hosting requirements before committing to a price. Older unsupported versions can turn into an upgrade or application-remediation project. Since this is a tax and accounting firm, review their security and compliance obligations as well. Creating a WISP, assigning responsibility, documenting incident response, and implementing the related controls should be explicitly scoped rather than treated as free extras.
They currently have QuickBooks Premier Accountant editions from 2019 through 2024, and there is no WISP. The firm has also had recent ownership changes, so I expect the permissions interviews to be difficult.
One hundred hours sounds very light for doing this properly. Break the project into separate work packages: discovery, architecture, identity and access, security policies, application compatibility, cloud configuration, data migration, endpoint work, testing, training, documentation, and cutover. Estimate each one generously and include contingency. A three-month timeline may be more realistic than trying to complete everything in a week or two, especially if this is being done alongside your regular job.
That’s more realistic than the one-week estimate I initially heard. This is also my first project of this type, and I can only dedicate a few hours most days.
That’s my biggest concern. I tend to estimate confidently, then discover more work and end up absorbing the extra hours.