Signing a Scope of Work? The Party It Leaves Out Usually Sets Your Schedule

The scope of work ran eleven pages, and by the standards of the market it was a good one. A regional distributor with about ninety employees had decided to replace a payroll and time-tracking setup that had been assembled in pieces over a decade, and it hired an implementation firm to do the work for a fixed fee across a fourteen-week schedule. Deliverables were listed. Acceptance criteria were written out. Payment was tied to milestones rather than to elapsed time, which is the arrangement most buyers are told to want. The project finished four months late, and nothing in the document was wrong.

The eleven pages described two parties and needed three

What the scope described was a relationship between the distributor and the firm it had hired. The distributor would provide access, name a project sponsor, and review deliverables within five business days. The firm would configure the new system, map the data, run two parallel payroll cycles, and train the staff who would use it. Both sides did roughly what they had promised. The thing that decided the schedule was a party that appeared nowhere in the eleven pages except as background: the incumbent payroll provider, whose historical records had to come out before anything could go in, and who had no contractual reason to hurry.

That provider's standard export gave twelve months of detail and summary totals for anything older. The distributor needed several years of wage and deduction history at the individual level, both for its own reporting and because the Department of Labor is responsible for the federal recordkeeping requirements that sit behind payroll. Getting it meant a custom extract, a request form, a fee, and a queue. Nobody had checked the incumbent's terms before signing, because the scope of work had not asked anyone to.

Where the cost actually landed

The firm was not idle during the wait, which is part of what makes this pattern expensive. It configured what it could, built the mapping against the twelve months it had, and moved on to training. When the full extract arrived seven weeks later in a layout that did not match the sample, the mapping had to be rebuilt, the first parallel run had to be scrapped, and the training that had already happened was training on a version of the system that no longer existed. None of that was covered by the fixed fee, because the fixed fee had been priced against a scope that assumed clean inputs would be available in week two.

What followed was a sequence of change orders that together came to a meaningful fraction of the original fee, plus four months during which the distributor's controller was running two systems and reconciling both. The firm behaved reasonably throughout. The change orders were properly documented and fairly priced. That is the uncomfortable part of the case: no one misbehaved, and the buyer still paid substantially more than the number it had negotiated, because the negotiation had covered the wrong risk.

Which parts of the deal were still open, and which had closed

Five years out, the choice of payroll system will look like the smallest decision in this project. Systems get replaced. What does not get replaced easily is the data that came across during the migration, because every subsequent year of records sits on top of it, and gaps become permanent features of the file. The reversible parts of a professional services engagement are the ones people spend the most time arguing about: fee level, milestone dates, which staff member leads the work, whether training happens in person or remotely. All of those can be renegotiated mid-flight, and often are.

The irreversible parts are quieter. Once the historical detail was summarized rather than itemized, no later change order could recover it. Once the parallel run was scrapped, the distributor lost its one clean chance to compare old and new output on the same pay period with both systems live. Once the staff had been trained on a configuration that changed, the second training landed on people who had already decided the new system was confusing. Judgement about scopes of work is largely the habit of asking, before signing, which of these consequences can be bought back later and which cannot.

The four questions that would have caught it

The first is simply who else has to do something for this to work. Not who is paying and who is performing, but every party whose cooperation is on the critical path: the incumbent vendor, the outgoing bookkeeper, the bank that has to approve a new debit authorization, the landlord, the county reviewer, the insurance carrier, the software company whose license terms govern the export. Write the list, then check what each one is contractually obliged to give you and on what timetable. The distributor could have answered this in an afternoon by reading its existing service agreement.

The second is who is responsible for chasing them, named as a role rather than implied. A scope that says data will be provided by the client has assigned the work to whoever has least authority over the party that holds it. The third question is what happens to price and schedule if the third party is slow, which is where a good firm will tell you plainly whether the fee assumes inputs on time. The fourth is what the firm needs to see before it commits to a fixed number, because a firm willing to look at the incumbent's export format during the proposal stage is telling you something about how it works.

What a better version of the same document looked like

The distributor's next engagement, an inventory system replacement eighteen months later, used a scope built the other way around. It opened with a section listing dependencies by party, each with the name of the person accountable for obtaining it, the contractual basis for expecting it, and the date it was needed. Two of those dependencies sat with outside organizations, and both were requested before the engagement letter was countersigned. The fee was structured with a discovery phase priced separately and a fixed price for delivery that was set only after the inputs had been inspected. That project ran two weeks over, absorbed inside the contingency, and the controller kept her evenings.

The habit worth building is not suspicion of the firm you are hiring. It is a working map of everyone whose slowness can cost you money, drawn before you sign, and a plain answer to the question of who is holding each of them to a date.