Skip to content
EncryptTechBD
Home Industries Our Work Blog Pricing About Contact Discuss Your Project
বাংলা (BN)
Software Engineering 6 min read

Buying software project discovery: what to require and how to judge it

A buyer guide to software project discovery, covering deliverables, vendor categories, a weighted scorecard, pricing questions, pilot scope and exit terms.

E
EncryptTechBD · 0 comments
Code on a laptop screen in an office
Photo: Christina Morillo / StockSnap

Software projects rarely fail at the first line of code. They fail earlier, when a budget is approved against a vague description and a vendor is asked to estimate work nobody has defined. Discovery is the paid, time-boxed stage that turns an ambition into a decision: what will be built, what will not, what it depends on and what evidence would justify the next spend.

Buying discovery is different from buying development. You are not purchasing features; you are purchasing clarity, options and a defensible estimate.

What discovery is for

Discovery reduces uncertainty in four areas: the problem, the solution, the delivery approach and the commercial case. It should end with a decision, not a document that restates the brief. A good provider will be willing to recommend a smaller build, a process change or no build at all. If every discovery ends in the same large proposal, the stage is being used as sales collateral.

The deliverables you should receive

Ask every provider to price the same output. The format can vary, but the substance should not.

  • Problem statement: the current workflow, who is affected, the cost of the status quo and the outcome being sought.
  • Users and roles: primary users, administrators, approvers and anyone whose work changes.
  • Scope map: in-scope journeys, explicit exclusions and dependencies on other teams or suppliers.
  • Requirements: user stories or use cases with acceptance criteria, plus non-functional requirements for performance, availability, security and privacy.
  • Process and data models: current and future flow, key entities, integrations and data ownership.
  • Options: at least two delivery approaches with trade-offs, not one predetermined answer.
  • Estimate range: effort by phase, assumptions, exclusions and a confidence level.
  • Risk register: likelihood, impact, owner and mitigation for the top risks.
  • Delivery plan: phases, decision gates, team roles and client responsibilities.

Requirements to settle before you brief vendors

You do not need a specification to start discovery, but you do need constraints. Write down the business outcome, the budget range you can actually approve, the deadline and why it exists, the systems that must integrate, the data classification and the people who can make decisions. Name one client-side owner with authority to settle scope questions. Without that person, workshops become interviews and the report becomes a wish list.

Four vendor categories

Independent business analyst or product consultant

Strong for process-heavy problems and for organisations that already have a delivery team. The risk is a report that is analytically sound but technically naive. Set a fixed end date and a defined decision gate.

Product studio with design and engineering

Useful when the solution is uncertain and needs prototyping. Studios can test a journey before committing to a build. The risk is that discovery drifts into a design engagement with a large production proposal attached. Put the end date and the decision gate in the contract.

Offshore delivery firm

Often price-competitive and able to staff a full team. The risk is distance from users and estimating from a brief rather than observed work. Insist on user access, recorded walkthroughs and named people who stay on the project.

Internal team with an external facilitator

Best when the business has domain knowledge and engineering capacity but lacks a structured method. An external facilitator can keep the sessions honest and write the decision record. The risk is that internal day jobs crowd out the workshops. Protect the time in calendars before you start.

A weighted scorecard

Score each provider from one to five against the same criteria. Weights reflect what matters for your project; the list below is a starting point.

  • Domain evidence (25%): examples of similar workflows, not just similar industries. Ask how they handled an exception.
  • Facilitation quality (20%): who runs the sessions, how they handle disagreement and how they stop scope from expanding.
  • Technical literacy (20%): can they discuss integration, data quality, security and operability without hand-waving?
  • Artefact quality (15%): review a redacted sample. Are requirements testable? Is the estimate traceable to assumptions?
  • Handover (10%): will the delivery team receive a usable pack, or is the knowledge locked in the author?
  • Price transparency (10%): clear days, rates, expenses and what happens when the scope changes.

A score below three on domain evidence or technical literacy is a serious gap.

Pricing questions that reveal the model

  • Is the fee fixed for a defined output, or time and materials with a cap?
  • How many workshops and interviews are included, and who attends from your side?
  • What happens if the stakeholders are unavailable or the scope changes mid-stage?
  • Who owns the artefacts, and will you receive editable files?
  • What is excluded from the estimate, and how will those exclusions be priced later?
  • Can you stop after discovery without losing the work or triggering a penalty?

Compare the total cost of the decision, not the day rate. A cheap discovery that produces an unusable estimate costs more than a thorough one.

Run a paid pilot before the full programme

If you are unsure about a provider, commission a one-week problem framing sprint. Define the output as a problem statement, a journey map, a draft scope and a risk list. Set acceptance criteria in advance: the client owner must be able to explain the problem and the proposed next step without the consultant in the room. A pilot that fails this test is useful evidence, not a wasted fee.

Exit and handover terms

Discovery should make you more independent, not more dependent. Require editable source files, a decision log, a glossary and a walkthrough for the delivery team. Confirm that the provider will answer clarification questions for a defined period after handover. If the same firm is bidding for the build, ask them to separate the discovery fee from the build proposal so you can compare it with alternatives.

Red flags

  • A proposal that lists deliverables but no decision the stage will enable.
  • Estimates with no assumptions, no range and no exclusions.
  • Workshops run only with managers, never with the people doing the work.
  • A fixed solution presented before the problem has been observed.
  • No named team, no availability commitment and no handover plan.

A six-week buying sequence

  1. Week 1: write the outcome, constraints and decision owner; define the discovery deliverables you will require.
  2. Week 2: shortlist three providers with different models; send the same brief and ask for a redacted sample.
  3. Week 3: interview the people who will do the work; score against the weighted criteria.
  4. Week 4: commission a one-week framing sprint with the preferred provider.
  5. Week 5: review the output against the acceptance criteria; check the estimate assumptions with your technical adviser.
  6. Week 6: decide whether to proceed, reshape or stop; sign the build contract with the discovery pack attached.

Getting help with discovery

If you need a discovery stage before committing to a build, start with Custom Software & Integrations. The pricing page explains how we separate discovery, delivery and support. For the wider decision, read build versus buy for custom software.

Sources

Services that deliver this

Want this implemented properly?

Pricing →

Every engagement starts with a written scope. If a requirement is outside it, we quote it separately before starting.

Responses (0)

No comments yet. Add the first one below.

Leave a response

Your email address will not be published. Required fields are marked *.

Keep reading

More field notes