Skip to content
IT proposal & estimate review IT project specification Client-side IT project oversight Software acceptance & UAT IT project recovery
The IT project is delayed How to compare software vendors The scope keeps growing The vendor wants sign-off How to define your IT project
How I work Insights About Contact PLBook a consultation

Issue · Preparation

How to prepare an IT project before requesting estimates

Sending a few sentences to a development company usually produces a broad estimate or incomparable proposals. Each vendor fills the gaps with assumptions and prices a different project.

You do not need a complete architecture or every screen specified. You do need clarity on the problem, processes, users, priorities and constraints.

Good preparation helps with estimates and tests whether a new system is the best response to your organisation's need.

1. Define the problem and business goal

Do not start with features. Describe what fails today, its consequences and the intended change. “Implement a CRM” is too vague. A better goal is faster lead handling, a reliable contact history and a shared view of data.

The goal should allow you to assess business value later, not merely confirm that software went live.

2. Describe the current process

Check how work actually happens, not just the documented procedure. Who starts it? What data is used? Where are exceptions, manual workarounds and supporting spreadsheets? Who decides?

User interviews often reveal inconsistent rules and fragmented ownership as well as missing software.

3. Identify users and stakeholders

  • People performing daily operations.
  • Process owners and managers.
  • IT and security teams.
  • Data, legal and compliance owners.
  • Customers, partners or suppliers using the solution.
  • Management and the project sponsor.

4. Gather and prioritise requirements

Requirements work is not adding every idea to one list. Separate essential, important and optional needs, and explicitly defer items beyond the first phase.

Priorities should follow value, risk, dependencies and cost, not simply which stakeholder speaks loudest.

5. Establish data, integrations and constraints

  • What data will be processed and where it comes from.
  • Whether migration, cleansing or deduplication is needed.
  • Which systems the solution must connect to.
  • Whether current API documentation exists.
  • Applicable security and legal requirements.
  • Expected availability and performance.
  • Existing technology and supplier standards.

6. Define budget and time constraints

Hiding the budget does not always improve negotiation. Vendors need to know whether to propose a simple first phase or an organisation-wide solution.

A budget range is enough to start. Explain deadlines driven by real needs, such as a seasonal peak or regulatory change, rather than an arbitrary date.

7. Define success and acceptance criteria

State how you will judge the investment: shorter processes, fewer errors, more self-service, better data completeness or less manual work.

Define feature acceptance criteria separately. Business success and technical acceptance are related but not identical.

8. Prepare vendor selection

  1. Give candidates consistent materials and questions.
  2. Set evaluation criteria before seeing prices.
  3. Request assumptions, exclusions and the proposed team.
  4. Run workshops with finalists.
  5. Compare total cost and exit terms.
  6. Carry important proposal commitments into the contract or schedules.

Minimum pack before requesting proposals

  • Problem and goal.
  • Key process maps.
  • User groups.
  • Prioritised requirements.
  • Integrations and data.
  • Constraints and assumptions.
  • First-phase scope.
  • Vendor evaluation criteria.
  • Client-side decision-making arrangements.

FAQ

Frequently asked questions

Should we disclose the budget?

Not necessarily an exact figure, but a range helps vendors propose an appropriately sized solution. Without it, proposals may vary almost arbitrarily.

Should we choose the system or vendor first?

It depends. With an off-the-shelf product, assess both software and implementation partner. With custom software, the team and delivery approach matter more.

How long does preparation take?

It depends on the number of processes and stakeholders. You can start with a short discovery and develop documentation in stages.

Can we start without knowing everything?

Yes, if unknowns are explicit and the work to resolve them is planned. Uncertainty is normal; pretending it does not exist is dangerous.

A good idea, but most of the scope is still in people's heads?

Let us put it into a clear form before you approach vendors.

Let us prepare the project