The verdict in three sentences
A 50-employee company that asks 3 vendors to quote from a short brief receives bids that typically differ by 30 to 60 %, because each vendor imagines a different application. A 12-chapter requirements specification, with user stories prioritised using MoSCoW and measurable acceptance criteria, brings that gap under 20 % and limits change requests. Scoping costs 5,000 to 15,000 EUR in Dublin in 2026, about 5 to 10 % of a typical business application budget.
The 12 chapters of a custom software specification
A business application (field service management, quoting and invoicing, production tracking, scheduling) touches the company's internal processes. The vendor does not know them: the specification exists to pass them on without ambiguity. Here are the 12 chapters, grouped in pairs, with the mistake we see most often.
| Chapters | Expected content | Common mistake |
|---|---|---|
| 1 and 2: context, scope and out of scope | Stakes, measurable goals, what is not included | Leaving out the out-of-scope list, a source of change requests |
| 3 and 4: users, roles and current processes | Profiles, volumes, diagram of existing processes | Describing the dream tool without describing reality |
| 5 and 6: target processes and MoSCoW user stories | Target journeys, 40 to 120 prioritised user stories | Marking everything as a Must |
| 7 and 8: business rules, data and migration | Calculations, statuses, approvals, files to migrate | Underestimating the clean-up of spreadsheet data |
| 9 and 10: integrations and non-functional requirements | Accounting, payroll, ERP, security, GDPR, hosting | Not stating who pays for third-party APIs |
| 11 and 12: acceptance, timeline, budget, exit terms | Test scenarios, milestones, budget envelope, code ownership | Forgetting exit terms and source code access |
In volume, 15 to 30 pages are enough for an SME, appendices included. Chapter 6 alone accounts for nearly half the work.
Prioritising with MoSCoW: a field service example
MoSCoW sorts each need into Must (required at launch), Should (important but with a workaround), Could (nice to have) and Won't (excluded from this release). In a healthy first release, Musts represent about 60 % of the effort, Shoulds 20 %, Coulds 15 % and Won'ts the rest. Here is an extract for an HVAC maintenance company.
| User story | Priority | Estimated effort |
|---|---|---|
| As a scheduler, I assign a job to a technician based on area and certifications | Must | 6 days |
| As a technician, I complete the job sheet on mobile and the client signs on screen | Must | 8 days |
| As an accountant, I export billable jobs to the accounting software | Must | 4 days |
| As a manager, I see the first-time fix rate per technician | Should | 3 days |
| As a client, I receive an SMS the day before the visit | Should | 2 days |
| As a technician, I view the equipment photo history | Could | 3 days |
| As a client, I book appointments online | Won't (release 2) | 7 days |
Each user story must carry a testable acceptance criterion, for example: the signed job sheet is emailed to the client as a PDF within 2 minutes. Without it, acceptance testing turns into a negotiation.
Scoping budget and its effect on quotes
Three ways of producing the document coexist in Dublin. The table below compares their cost with the price gap typically seen between the quotes received afterwards, as a 2026 order of magnitude.
| Scoping approach | Estimated cost | Duration | Average gap between quotes | Overrun during the project |
|---|---|---|---|---|
| One or two page brief | 0 EUR | 1 week | 30 to 60 % | +25 to 40 % |
| Internal write-up without a method | 10 to 15 internal days | 4 to 6 weeks | 20 to 35 % | +15 to 25 % |
| Independent business analyst | 5,000 to 9,000 EUR | 3 to 4 weeks | 10 to 20 % | +5 to 15 % |
| Paid discovery workshop with a vendor | 8,000 to 15,000 EUR | 2 to 4 weeks | 10 to 20 % | +5 to 10 % |
| Discovery + clickable prototype | 12,000 to 15,000 EUR | 4 to 5 weeks | 5 to 15 % | Under 5 % |
A paid discovery with one vendor does not oblige you to hire it for the build, provided the document legally belongs to you. Get that in writing.
Mini case study
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Ciarán runs a 50-employee HVAC maintenance company in Dublin. His first tender, run from a 2-page brief, produced 3 quotes at 55,000, 88,000 and 125,000 EUR: impossible to compare. He then paid for 9,000 EUR of scoping, delivering 85 prioritised user stories and 40 acceptance criteria. The new quotes ranged from 80,000 to 94,000 EUR.
The maths: with the brief, a 30 % overrun on an 85,000 EUR base would have cost 25,500 EUR. With the specification, an 8 % overrun costs 6,800 EUR. The estimated saving reaches 18,700 EUR, more than twice the scoping fee, before counting a faster launch thanks to a well-bounded Must release.
FAQ
Should the specification describe the user interface?
No, describe needs and business rules, not screens. Wireframes of 5 to 10 key screens are enough to clear up most ambiguities.
How many vendors should we approach?
Three is right for an SME: beyond that, assessing bids takes more than 5 days and vendors invest less effort. Give them 3 weeks to respond.
What if the quotes exceed the budget?
Move Shoulds and Coulds to release 2. A first release limited to Musts often represents 55 to 65 % of the total budget and can launch in 10 to 14 weeks.
How do we write useful acceptance criteria?
Each criterion must be measurable: a time, a volume, an expected result. For example: the accounting export of 500 jobs runs in under 30 seconds.
Who owns the source code at the end?
Chapter 12 settles it: assignment of IP, repository access and documentation. Without a clause, switching vendors can cost 20 to 40 % of the initial budget in reverse engineering.
Let's scope your project. Share your brief or current processes, and we will propose a priced discovery, an estimate of the build budget and a release-by-release plan. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
