Deciding whether to price a tender went from 6 days to 1
At Kristian’s construction company, a main contractor building and refurbishing high-end homes for private clients, a tender invitation used to sit for about 6 days before anyone decided to price it. Now it’s decided the day it lands, and the company wins as many jobs from fewer tenders.
Key results
- 6 → 1 dayFrom invitation to a decision
- 210 → 45 minsTo log, read and decide each one
- 6 → 1Tenders a year dropped part-way through pricing
- The business
- High-end homes
- Service
- Tender Intake
- Live in
- 4 weeks
- Paid back in
- Approx 6 weeks
- Size
- 80+ staff
Background
The company builds and refurbishes high-end homes for private clients, mostly through architects. About 80 tender invitations came in a year, by email, from portals and from architects’ Dropbox links. The team priced about 25 of them and won about 1 in 4.
The challenge
Each invitation took about 3.5 hours before anyone decided anything: logging it, downloading the documents, reading enough of the pack to know what it was, then checking it against the work already on.
Most of that reading landed on Kristian as MD, between clients, cash and site. So invitations waited, often for 6 days, and a tender with a 4-week return lost 6 days before the estimators started.
About 6 a year got dropped part-way through pricing, once the contract amendments or a programme clash with a job already on came to light.
6 tendersa year dropped part-way through pricing, about 3 days of estimating gone on each
“The invitations weren’t the problem, it was me being the bottleneck on every one of them.”
The solution
Code does the logging, filing and checking against the company’s own rules, so nothing is typed twice. AI reads the pack, because every architect lays one out differently, and writes a 1-page summary the MD can decide from.
What we built
- Logs every invitation. It reads the email, pulls out the client, value and return date, and adds it to the tender register.
- Gets the documents. It downloads them from the portal or link, files them, and lists anything missing.
- Writes a 1-page summary. It covers the scope, the contract and its amendments, the programme and anything unusual, each with its page.
- Checks the fit. Value, location, workload in the return window, and how past jobs with the client and architect went.
- Sets it up. Once the MD says yes, the dates go in calendars, the query log starts, and the estimator gets the folder.
Tech stack
It runs on the company’s own Microsoft 365 and Azure accounts. It works alongside the tools the company already uses.
The results
- 01
Priced fewer, won the same
About 20 tenders a year instead of 25, and still about 6 won. That’s 5 tenders of estimating time back each year, and the win rate up from 1 in 4 to nearly 1 in 3.
- 02
Decisions on day 1
The summary is in Teams the day the invitation lands. The estimators start about 5 days earlier on the ones worth pricing.
- 03
1 tender dropped part-way, not 6
The amendments and programme clashes show up in the summary, before anyone starts pricing.
- 04
Every decline goes back the same day
A prompt, polite reply keeps the architect onside.
“I read one page now and make the call. If I want more, it shows me the page in the pack.”
More case studies
See all case studies
Main contractorThe final tender check went from a day to 30 minutesRead the case study
Main contractorPricing a variation went from 14 days to 3Read the case study
Main contractorPackaging a tender went from 19 hours to 5Read the case study
M&E contractorReading an M&E spec went from 6 hours to 2.5Read the case study
Housebuilder40 hours a week back for a commercial and projects team of 10Read the case study
Find what nobody’s pricing in your business
Answer a few questions about your business and you’ll get your top three priorities, the changes that would put the most margin or hours back, and where to start. It costs nothing.