Tender Summary

• Client: West Yorkshire Combined Authority (WYCA), covering five councils:
  • Bradford
  • Calderdale
  • Kirklees
  • Leeds
  • Wakefield
  Data must hold up by district.
• Owning directorate: Inclusive Economy, Skills and Culture. Audience is policy-focused, not technical.
• Strategic hook for the bid: the work evidences the Local Growth Plan's skills pillar and the Mayor's "Region of Learning and Creativity" ambition. Frame the methodology around this.

Rest of the section is scene-setting boilerplate (transport, housing, sister programmes) — not scope.

• Core problem: no consistent or comparable destination data across providers. Solution is a single standardised survey instrument to produce comparable datasets.
• In-scope provision (pilot) — three unlike populations, compared like-for-like:
  • Essential Skills (Level 2 English, Maths, Digital)
  • ESOL (English for Speakers of Other Languages)
  • Construction Skills (any level)
• Population / sampling frame: ~14,500 learners, adults aged 19+.
• Four outcome measures (the framework everything hangs off):
  • Into work (incl. self-employment) if not previously working
  • Progression to higher wages
  • Progression into further learning
  • Progression in well-being / mental or physical capacity toward any of the above softer, needs a validated scale

• Single overriding objective: a statistically robust report on the effectiveness and value for money of each provision type at producing the four outcomes.
• Two distinct analytical outputs (not one):
  • Effectiveness — comparative: how well each provision (Essential Skills / ESOL / Construction) produces the four outcomes.
  • Value for money — effectiveness per pound; must be framed in Green Book / Magenta Book terms.
• "Statistically robust" is the phrase you'll be held to. Links directly to the SDs / standard errors / confidence levels requirement, and back to matched-pair sample size.

• Coverage: survey as many of all ~14,500 learners as possible. Census ambition, not a sample.
• Two surveys per learner (matched pair):
  • Baseline = learning intent (near start)
  • Destination = outcome (later)
  The matched-pair count is the true analysis sample; securing both surveys from hard-to-reach learners is the central delivery risk.
• Multi-channel outreach (supplier's choice of mix), doubling as an inclusion mechanism:
  • WhatsApp
  • SMS
  • Email
  • Telephone interviews (for demographic spread; key for ESOL / low digital-literacy)
• Rolling cohorts over up to 14 months — continuous baseline enrolment plus destination follow-up, not a one-off issue.
• Ongoing project management scope (easy to under-price):
  • Data cleanliness & timeliness of Learning Provider data (upstream dependency and risk)
  • Survey transmission
  • Response management with follow-ups
  • Response data management & analysis
  • Ongoing response-rate reports (feed the dashboard)
  • Final reports on outcome performance by provision

Four contractual outputs. Each maps 1:1 onto a paid milestone, so every deliverable is a paid gate. Three of the four are yours or heavily data-led; only the inception paper is shared PM work.

D1 — Inception paper

• Agreed delivery strategy, target survey respondent numbers, and full methodology.
• Produced after the inception meeting.
• This is where your target matched-pair numbers get formally committed, so the sample-size / power thinking must be done up front, not retrofitted.

D2 — Shareable dashboard

• Ongoing intelligence on both baseline and destination survey responses.
• Must show numbers of matched pairs, broken down by Learning Provider and by provision — and note the wording is "at least", so that is the minimum granularity, more cuts expected (e.g. by district).
• Squarely your build. "Shareable" + "ongoing" points to Power BI published to WYCA, running live across the 14 months. Price it as build + run, not a one-off. The "at least" is an open invitation to add value with extra dimensions.

D3 — Final report (the £-weighted, highest-bar deliverable)

• Intelligence on effectiveness and value for money of outcome production.
• For each of the four outcomes, cut by provider and by provision.
• Full statistical analysis: standard deviations, standard errors, confidence levels.
• Plus "any other data analysis deemed useful by the provider" — room to demonstrate expertise.
• The by-provider AND by-provision AND by-outcome cross-tabulation is exactly where thin matched-pair cells will bite. Methodology must address low-n cells: suppression thresholds, aggregation, or explicit confidence-interval caveats.

D4 — Roll-out recommendations & methodology

• A recommended, efficient and cost-effective process for extending robust outcome measurement to all Adult Skills provision (45,000+ learners).
• The strategic pay-off: the pilot exists to prove a scalable, affordable method. A distinct deliverable in its own right, not a footnote. Ties back to the value-for-money theme.

A brief for how the final report must look and land. Plays directly to data-visualisation strengths and is a genuine differentiator on the 50% quality score.

Audience & purpose

• Report must be clear and engaging for a policy-focused audience, not analysts.
• Findings will directly inform decision-making and policy, so must be concise, accessible, impactful.
• The statistical rigour of D3 has to be translated, not dumped. The readers decide funding, not methodology.

Mandatory format

• A one-page visual summary using graphics to highlight key insights, conclusions and recommendations — a hard requirement.
• An executive summary — separate from, and in addition to, the one-pager.
• Actively use charts, diagrams, infographics and alternative visual formats to maximise clarity.
• Purpose is effective dissemination across WYCA — outputs travel beyond the immediate project team.

• Bi-weekly steering group to update WYCA's project team — meetings virtual and in-person as required.
• Additional updates on request; respond to reasonable requests within three working days.

Only two things matter for pricing: (1) "bi-weekly" is ambiguous — likely fortnightly, worth an enquiry to WYCA; (2) in-person sessions over ~13 months carry travel cost/time — reflect in pricing and propose a sensible virtual/in-person split.

• Budget: up to £35,000 excluding VAT for the whole engagement.

• Proposal submitted on WYCA's supplied Template; evaluation 50% quality / 50% cost.
• The completed proposal is incorporated into the contract — promises made are contractually binding.
• The Pricing Schedule was a mandatory submission — not using it meant exclusion from scoring.

Three qualifying criteria the consultant had to meet.

• Experience — minimum 4 years delivering similar work; demonstrable understanding of the interventions, the delivery challenges, and the direct/indirect benefits behind the outcome data.
• Expertise — proficiency in qualitative and quantitative survey and research methods; specifically experience of low-cost methods for reaching large numbers of clients with varied characteristics, including different languages and reading abilities.
• Sector knowledge — detailed understanding of the Green and Magenta Books, Government Departmental Guidance, and best practice in appraisal and evaluation.

Constrains how the data platform may be built and hosted.

Framing

• WYCA = data controller; supplier = data processor — personal data processed on their instructions, not ours.
• Must follow the contract's Data Protection Appendix instructions.
• Must assist with a DPIA pre-award where appropriate (very likely, given data volume and multi-channel contact details).
• International transfers: any processing outside the UK/EEA must be declared with safeguards (Standard Contractual Clauses, adequacy, Article 46 / Binding Corporate Rules).
• Must comply with WYCA's minimum security standards and confirm compliance.

Security standards that bite for a cloud data platform

The full table has 13 rows; the postal / paper / physical-asset controls are omitted here as not relevant to the build.

• Storage: no saving CA documents locally to a device's C:\ drive; use secure cloud (SharePoint, AWS) or technical measures such as network segregation and firewalls.
• Removable media: encryption to AES256 / FIPS 140-2 (USB), or WinZip / 7zip for CD/DVD.
• Infosec / Technical / Cloud Security: supported OS; MFA, anti-virus and Microsoft-best-practice passwords on any device handling data; cloud services must follow the NCSC 14 Cloud Security Principles before any CA data is stored.
• Email: special category data only to secure addresses, with encryption/password sent via a separate channel.
• Data Protection: a DP policy aligned to the data-protection principles, plus procedures for subject-rights requests, privacy notices and incident handling.