RFP Feedback and Optimization Priorities

Derived from:

• Issue Tree
• Priorities
• Win/Loss Picture

Purpose

This document summarizes what feedback already exists across recent rejected bids, what can reasonably be concluded from it, and what Datopian should optimize first.

The aim is to move from a raw win/loss picture to a more actionable diagnosis.

Summary

We already have enough evidence to make a first pass at diagnosis.

The evidence is still partial and fragmented, but it is not so weak that we must wait before acting. The current pattern suggests Datopian should prioritize improving:

1. proposal quality against scored questions
2. proof of domain / buyer-specific fit
3. compliance and pricing discipline
4. post-submission process discipline

This does not yet justify a full redesign of the end-to-end RFP process. It does justify a focused first wave of improvements in how bids are selected, written, checked, priced, and followed through.

Rejected Set Reviewed

This note focuses on the rejected issues in the recent submitted RFP set:

• #653 DfT Local Transport Data
• #646 Executive Council EEIA
• #663 DHSC MedTech Compass
• #615 Leicester Air Quality
• #657 UKRI SDR UK Data Catalogue
• #678 3PODS
• #705 Energy Poverty Dashboard
• #661 LG Inform 2026
• #721 Scotland open data platform
• #450 Ofwat Programme Ocean
• #471 IDB Open Data Platform
• #775 Code for Canada Workflow Services
• #776 Code for Canada Notification Services
• #780 UAE Ministry of Investment

What Feedback We Already Have

Strong / actionable feedback already available

#653 DfT Local Transport Data

The feedback is clear: the response did not demonstrate enough depth on the relevant criterion. The evaluator wanted one convincing example rather than several shallow references.

Implication: answer structure and evidence density were weak relative to the scoring format.

#615 Leicester Air Quality

The winning supplier had stronger operational and domain-specific fit:

• AQMS hardware coordination
• DEFRA reporting and validation
• real-time sensor telemetry integration

Implication: this was not just a general platform contest; operational domain credibility mattered materially.

#657 UKRI SDR UK Data Catalogue

This is currently the richest single source of learning. The documented weaknesses include:

• insufficient academic research fit
• response exceeded limits, so useful content may not have counted
• methodology was too generic
• risks were not tailored enough to the context
• timeline lacked enough delivery detail
• innovation and social value were weak / not measurable

Implication: this bid exposes several repeatable improvement opportunities at once.

#678 3PODS

The comments suggest two concrete issues:

• pricing was considered too low / not credible
• financial stability was a concern

Implication: pricing problems do not only point in the "too expensive" direction. Credibility and commercial logic also matter.

#450 Ofwat Programme Ocean

Recorded feedback indicates weaknesses on:

• data transformation leadership
• metadata management plus cultural change
• specificity of agile methodology
• capability-building approach

Implication: responses appear to have been too technical and insufficiently strong on change, operating model, and organizational delivery dimensions.

#661 LG Inform 2026

This is a different kind of loss. Datopian was shortlisted, but then effectively lost because:

• the next stage required in-person presentation in London
• remote presentation was not accepted

Implication: some losses happen after proposal quality is already good enough to stay in contention. Later-stage process execution matters.

#780 UAE Ministry of Investment

From the project notes and offsite discussion, the strongest known issues are:

• missed mandatory criteria
• commercially uncompetitive pricing

Implication: both compliance control and pricing discipline need attention.

Some signal, but weak / incomplete

#646 Executive Council EEIA

No detailed debrief was provided, but the reply suggests our response may not have been sufficiently aligned to what the solicitation was actually asking for.

#705 Energy Poverty Dashboard

We know the bid was rejected. A debrief was attempted through partners but did not produce anything useful.

#775 and #776 Code for Canada bids

We have rejection notices, but no substantive debrief yet.

#471 IDB Open Data Platform

We know it was lost and a feedback request was made, but no meaningful debrief is captured in the issue comments reviewed so far.

Worth trying to get more feedback

Priority follow-up candidates:

• #775 Workflow Services
• #776 Notification Services
• #705 Energy Poverty Dashboard
• #471 IDB Open Data Platform
• possibly #663 DHSC MedTech Compass, if no breakdown ever arrived
• possibly #721 Scotland, depending on how good the linked feedback file is

What Seems True So Far

1. Proposal responses are often too generic

This is the clearest repeated pattern.

The feedback from #653, #657, and #450 points toward the same issue in different forms:

• not enough depth against the specific scored question
• too much generic material
• insufficiently tailored methodology
• too little concrete proof

This is not about writing style alone. It is about evaluator usability: can the buyer quickly see why Datopian deserves points on this exact criterion?

2. Domain / buyer-specific fit is not always being demonstrated strongly enough

The feedback from #615, #657, and #450 suggests a second pattern: Datopian may have broad technical capability, but broad capability is not always enough.

What seems to matter:

• relevant delivery context
• recognizable stakeholder environment
• domain vocabulary
• evidence of having solved similar problems for similar buyers

This is especially important when competing against incumbents or firms with narrower but stronger perceived fit.

3. Compliance and pricing discipline are inconsistent

The strongest evidence here comes from:

• #780 mandatory criteria + pricing
• #678 pricing too low / not credible

This suggests the commercial issue is not simply “we are too expensive.” The broader issue is that pricing and bid logic are not always clearly calibrated and defended.

Possible failure modes:

• bid price too high for the buyer/context
• bid price too low to appear credible
• mandatory criteria not properly controlled
• too little explicit commercial rationale

4. Some losses are caused by process discipline after submission

#661 is the clearest example: the proposal was good enough to progress, but the bid still failed because the team could not support the next-stage in-person requirement.

This suggests the process should not conceptually stop at submission. The real bid process continues through:

• presentations
• clarifications
• negotiation readiness
• award monitoring

Optimization Priorities

Priority 1: Improve response quality against scored questions

This is the highest-confidence intervention.

Focus:

• answer each scored question directly
• use one strong relevant example rather than several shallow examples when space is tight
• cut generic boilerplate
• structure responses so the evaluator can map them easily to the scoring criterion
• stay rigorously within limits

Why first:

This appears in the broadest range of documented losses and is the most reusable improvement across bids.

Priority 2: Improve fit-testing and fit-proof

Before and during the bid, ask:

• do we actually have strong enough domain fit?
• can we prove it with the right case studies?
• do we need a partner to strengthen credibility?
• are we clearly addressing the specific stakeholder environment, not just the technical stack?

Focus:

• better case study selection
• clearer domain framing
• more explicit statement of why Datopian is credible in this exact context

Why second:

Technical capability without visible fit appears insufficient in several rejections.

Priority 3: Add lightweight compliance and pricing controls

Focus:

• mandatory criteria checklist before submission
• explicit commercial rationale for pricing posture
• sanity check for price credibility in both directions
• clear sign-off on any high-risk pricing or qualification assumptions

Why third:

This is a high-impact control that may prevent avoidable losses.

Priority 4: Extend process discipline beyond submission

Focus:

• identify presentation / interview requirements early
• assess travel / presence constraints
• track clarifications, stage transitions, and award timing
• record outcome and key learning after the result

Why fourth:

This is important, but less urgent than proposal quality and compliance controls.

What We Should Probably Not Do Yet

Based on current evidence, it is too early to:

• redesign the whole RFP process end to end
• invest heavily in new tooling
• conclude that pricing alone is the main problem
• conclude that selection alone is the main problem

The current evidence supports focused improvements, not a sweeping process rewrite.

Recommended Next Moves

Immediate

1. Review the strongest feedback documents in detail:
   • #657
   • #721
   • #450
   • #653
2. Turn those into a simple checklist for future bids:
   • question-by-question response quality
   • fit proof
   • compliance check
   • pricing rationale

Next

1. Continue trying to close the rejected bids where useful debrief may still be obtainable
2. Select 3–5 representative bids for full retrospective review
3. Use those reviews to refine the first wave of changes

Bottom Line

We do not yet know everything, but we do know enough to act.

The evidence already supports a first improvement wave focused on:

• more specific, evaluator-friendly proposal responses
• stronger proof of domain fit
• tighter compliance and pricing controls
• better post-submission discipline

That is the most defensible next direction from the feedback currently available.