Example project — read-only
Missed delivery windows
A finished project, shown exactly as yours will look when all five phases are locked. The notes below were captured by the Guide during the conversation — nothing here was written by hand. It's an operations problem on purpose: you don't need to be an engineer for this method to work.
Empathize & Define
LockedWho is affected
Suburban customers of a regional delivery company who wait at home for a promised two-hour window. The window is missed roughly 30% of the time. Dispatchers field the complaints; drivers feel the pressure but can’t control traffic or stop density. The company already tried adding buffer time to routes and hiring two extra drivers — buffer made total routes longer (deliveries got later overall), and extra drivers were too expensive to keep.
Problem statement
Customers can’t rely on the promised delivery window, and the obvious fixes (slack in the schedule, more drivers) have either made deliveries later overall or cost too much to sustain. Point of view: the at-home customer needs to trust the window enough to plan their day around it.
Isolate Contradiction
LockedImproving parameter
Reliability per stop — the chance that any single delivery lands inside its promised window. (In TRIZ language: reliability.)
Worsening parameter
Total route time — every minute of buffer added to protect one stop pushes every later stop further out. (In TRIZ language: loss of time.)
Contradiction type
A technical contradiction — two different things trade off (the reliability of one stop vs. the total route time), rather than one property that must be two opposite values at once. That tells us the resolution will lean on inventive principles, not on separating a single property in time or space.
Where & when it collides
It bites on the road, in the back half of the run: each minute of buffer added to protect an early stop pushes every later stop further out, so the conflict concentrates in the afternoon, on the last deliveries of the day. The morning stops barely feel it. Pinning the exact zone and moment keeps the fix aimed where the pain actually is.
Contradiction statement
Adding slack makes each stop more reliable but makes the whole route slower — so the last deliveries of the day get later, and the company trades one kind of lateness for another.
Envision the Ideal
LockedIdeal final result
Every parcel lands inside its promised window — with no extra route time, no extra drivers, and no added cost. In the perfect case the “window” problem stops existing at all, because each parcel is already exactly where the customer wants it, when they want it, and the customer barely has to think about the logistics. (Stated by the benefit, not the mechanism — we are not yet saying how.)
What blocks it
One obstacle stands in the way: door-to-door delivery forces every parcel onto the same time-boxed live route, so protecting one window necessarily steals time from the next. Remove just that — let the at-risk parcels arrive without riding the live route — and the ideal opens up. That single obstacle is exactly what ideation has to dissolve.
Resources in play
Things already on hand, no new spend: the fleet’s idle overnight hours; neighbourhood convenience stores already open long hours and dotted across every postcode; the customers themselves (some would happily collect if it were closer and easier than waiting in); and the company’s own complaint log, which already flags exactly which stops are most at risk. A solution should try to reach the ideal using these before buying anything new.
Inventive Ideation
LockedCandidate moves
Because this is a technical contradiction (two different things trading off), the Guide reached for inventive principles rather than separation. Prior Action — do part of the work before it’s needed (which here doubles as a separation in time: shift the delivery effort to the night before). Segmentation — split one job into independent parts (a separation between parts: handle the at-risk stops apart from the rest). A third idea, Dynamics (let the system change shape mid-run), was raised and dropped — re-routing mid-day confused dispatch more than it helped.
Adapted directions
Prior Action → stage parcels at neighbourhood pickup points overnight, so the morning route only covers customers who insist on door delivery. Segmentation → a small fast van runs only the stops flagged at-risk that morning; the main trucks run the rest. The Guide also floored the obvious framing with one inversion — what if the parcel never rides the live route at all? — which is what makes the pickup-point idea click.
Ideality ranking
Ranked by benefit against cost and new problems — not by which felt nicest. Pickup points score highest: they spend only resources already on hand (existing shops, idle overnight hours) and remove the at-risk stops from the route entirely, at the modest cost of asking some customers to collect. The segmented fast-van scores lower — a second vehicle and extra dispatch are fresh cost and a fresh problem. So pickup points lead; the fast van is the fallback if opt-in is weak.
Prototype & Validate
LockedPrototype scope
Two-week pilot in one postcode: one neighbourhood pickup point (rented shelf space in an existing convenience store), parcels staged the evening before, customers offered the choice of pickup or their usual window.
Validation plan
Test with the twenty customers who complained most in the last quarter. Success signal, read against the ideal of “inside the window at no extra route time”: missed-window complaints drop by half AND total route time does not grow. Open risks: customers may treat pickup as a downgrade — watch the opt-in rate; if it stays under 30%, try the segmented-fleet direction next.
In tribute — Genrikh Altshuller (1926–1998)
The TRIZ half of this method is the life's work of Genrikh Altshuller. As a young navy patent examiner in 1946 he noticed that the strongest inventions don't trade a problem away — they resolve the contradiction at its heart, and they do so in a surprisingly small number of recurring ways. Studying tens of thousands of patents, he distilled those patterns into a teachable system built on one radical conviction: inventing is not a gift of genius — it is a skill anyone can learn. He paid for his independence in Stalin's Gulag, where he kept teaching fellow prisoners, and reached a wider audience as the science-fiction writer "Altov." Every contradiction, ideal result, and separation principle in this tool descends from his work.
The official Altshuller Foundation site →Your problem, same path.
Start a project →