Havish M Consulting
Customer Case Study · Sales & Projects · Tender Intelligence
Lighting and Electrical Distributor × Havish M Consulting

Win the right projects, at the right price

In HMC's GenAI programme, the distributor's sales and projects leaders built a pipeline radar for the UAE construction market: every live project on one map, an animated pipeline board, a bid optimiser that finds the discount sweet spot, and a GenAI tender brief that turns a new enquiry into a bid plan.

ClientLighting and Electrical Distributor FunctionSales & Projects ThemePipeline & Tender Intelligence · GenAI use case
Live projects on the radar
0projects
across seven emirates and seven segments
Tender win rate
0%
▲ from 24% before the use case
Average discount on won bids
0pts lower
price where it wins, not everywhere
Time to a tender brief
0hours
from about two days

Figures indicative; validated metrics available on request.

The Challenge

Lighting is won or lost long before the tender

In projects business, the brand a consultant specifies, the contractor who buys and the discount offered decide the outcome. The distributor's leaders wanted to see the whole pipeline and bid with evidence, not instinct.

Pipeline in many heads

Project leads sat with individual salespeople and spreadsheets, with no single view across emirates.

Specification blind spots

Which consultants specify the distributor's brands, and where competitors are written in, was not tracked.

Discounting by habit

Deep discounts were given to win, even where specification meant the order was likely anyway.

Slow tender preparation

Reading tender packs and finding comparable past bids took days for each enquiry.

HMC Approach

A programme use case, built by the people who sell

Developed by the distributor's sales and projects leaders during the GenAI programme, with HMC coaching the build.

Select a step to see what happened.

Discover · What wins a lighting package?

Leaders mapped the decision.

  • Who specifies, who buys, who approves alternatives.
  • Where the distributor wins and loses today, and why.
  • The questions they want answered every Monday.

Deliverable: Win-loss hypotheses.

Design · Pipeline model and bid logic

Turned hypotheses into a model.

  • Project stages from specification to supply.
  • Win probability by discount, specification and segment.
  • A tender brief template GenAI fills from the documents.

Deliverable: Data model and prompt templates.

Build · Radar, optimiser, brief

Built with GenAI-assisted coding in the hackathon.

  • Map and pipeline board.
  • Bid optimiser trained on past tenders.
  • GenAI tender brief with comparable bids.

Deliverable: Working prototype.

Adopt · Monday pipeline review

Into the sales rhythm.

  • One screen for the weekly review.
  • Every bid above a threshold run through the optimiser.
  • Briefs drafted before the bid meeting.

Deliverable: Review cadence.

Scale · Earlier in the cycle

Where the next gains are.

  • Consultant engagement plans to get specified.
  • Win-loss reasons captured on every tender.
  • Connection to CRM and ERP data.

Deliverable: Six-month roadmap item.

The Solution

See the pipeline, price the bid, brief the team

Four working modules on one synthetic year of UAE projects. Press play to watch the year unfold.

01 · Pipeline radar

Every live project in the UAE, by stage

Bubble size is the lighting package value, colour is the stage. Play the year, filter by segment, scroll or pinch to zoom into Dubai, and click any project for its details and a GenAI next-step.

Jan
Month

By emirate

Synthetic projects, values, consultants and contractors; locations approximate. Boundaries: Natural Earth.

02 · Pipeline board

Watch projects move from specified to supplied

The same year as a board. Cards slide between stages as the month advances; column totals show the lighting value at each stage. Click a card to see it on the map.

03 · Bid optimiser

Find the discount that maximises expected margin

A win-probability model fitted on 260 past tenders. Choose the context, slide the discount, and see win probability, margin and expected gross profit. The heat-map shows win rates by consultant and segment; click a cell to use that context.

SegmentSpecification
Won tenderLost tenderWin probability (fitted)Expected gross profit
Discount offered
Package value

Win rate · consultant × segment

Synthetic past tenders; logistic model fitted in the page.

04 · GenAI tender brief

From enquiry to bid plan in one pass

Pick an incoming tender. The assistant reads the enquiry, retrieves comparable past bids, scores win probability with the bid model, and drafts the brief for the bid meeting.

✦ Tender assistant
Pick a tender and press "Generate brief".

Comparable past tenders

Retrieved comparables appear here, ranked by similarity.

Impact Dashboard

Specified earlier, priced smarter

With the pipeline visible and bids priced on evidence, discounts tightened where specification already favoured the distributor, and effort moved upstream to consultants.

The distributor's rank among lighting suppliers

By segment, share of awarded lighting value, quarter by quarter

Discount on won bids

Distribution by segment

Figures indicative; validated metrics available on request.

Before vs After

How projects are pursued now

DimensionBeforeAfter
Pipeline viewPer salesperson, in spreadsheetsOne map and board for all emirates
SpecificationDiscovered at tenderTracked by consultant from the start
DiscountingDeep by defaultSet from win probability and margin
Tender preparationAbout two days per briefDraft brief in about two hours
LearningWin-loss reasons lostEvery bid feeds the model
Weekly reviewStatus updatesDecisions on where to focus
Tools & Tech Used

Built in the programme

GenAI assistants tender reading and briefs GenAI-assisted coding map, board and optimiser Logistic regression win probability Excel pipeline data projects and past tenders HMC safety framework no confidential data in public tools
Key Takeaways

What any projects business can borrow

01

Win at specification

Being written into the consultant's spec moves win probability more than any discount.

02

Price to expected margin

The best discount maximises win probability times margin, and it is rarely the deepest one.

03

Let leaders build

A tool built by the people who sell gets used in Monday's review.