Creative Destruction Lab application 2026: how CDL works
CDL is not an accelerator: no demo day, no standard equity, and mentors can stop working with you between sessions. Here is how the application really works.
Creative Destruction Lab application 2026: how CDL works
A Creative Destruction Lab application is a bid for a seat in a nine-month, objectives-based program, not an accelerator batch. CDL runs full-day mentor sessions roughly every eight weeks (CDL program page), partner sites such as HEC Paris take no equity and charge no fees, and mentors decide between sessions whether to keep working with you.
Treat CDL like a Canadian Y Combinator and your application will read wrong. There is no demo day, no standard term sheet, and no guarantee you reach the end of the cycle. CDL is a sequence of full-day mentor sessions where objectives get set, and between sessions the mentors choose whether to keep going with you.
The scale is real. CDL hosted over 630 companies in its 2024/25 program year across more than 20 specialized streams, and listed 230+ graduating ventures (CDL 2024/25 graduate companies). CDL does not publish a session-by-session attrition rate, so do not read a survival percentage into those two numbers.
How to apply to Creative Destruction Lab in 8 steps
The application is a stream decision first and a writing exercise second. Pick the wrong stream and your strongest possible application lands in front of mentors who cannot help you.
- Check you are a technology company, not a services company. CDL is built for seed-stage science and technology ventures aiming at massive scale. If your defensibility is a sales team, apply somewhere else.
- Pick the stream before the city. CDL ran more than 20 specialized streams in 2024/25 (CDL 2024/25 graduate companies). The stream decides which mentors sit in your room, and the room is the entire product.
- Check the current cycle on CDL's own site. Windows and start dates are set per stream and per location, so check the program's site for the current cycle instead of trusting an aggregator listing.
- Write the application around one falsifiable objective. Name what you will have done in eight weeks. Do not name a market you plan to "explore."
- State your hardest technical risk in the first paragraph. Say what would falsify your approach and what evidence you have so far. Burying it does not make it disappear at session one.
- Show the science, not the deck. Link the paper, the benchmark, the pilot data, the patent filing. Everything else compresses.
- Name the mentors you want and why. Streams publish mentor lists. If you cannot name two people you specifically want across the table, you picked the wrong stream.
- Budget nine months, not three. The cycle runs roughly nine months with full-day sessions about every eight weeks (CDL program page). Do not apply if a founder cannot be in the room each time.
The Creative Destruction Lab objectives process, explained
The objectives process is the whole product, and everything else is scaffolding. CDL's operating rhythm is mentor-led, objective-setting full-day sessions where short-term objectives are set and progress assessed (CDL program page).
In practice: you present, the mentors argue, and you leave with a small set of objectives due at the next session. Then you either hit them or you explain why not, to the same people, in the same room.
Write objectives a mentor can also be held to. Vague ones give nobody anything to check, and an objective nobody can check is an objective nobody will help you hit.
✅ Good: "Sign two paid hospital pilots and clear IRB review before the next session." Dated and falsifiable, and a mentor can open one specific door for it.
❌ Bad: "Improve go-to-market and refine the pitch." Nothing to verify in eight weeks, so no mentor takes ownership of it.
CDL program structure: nine months of checkpoints
CDL program structure is a series of hard checkpoints, not a continuous batch. Sessions run full-day, roughly every eight weeks, across a cycle of about nine months (CDL program page).
Between sessions, mentors decide whether they want to keep working with a venture. That is the mechanism that removes companies mid-cycle, and it is the single biggest difference from a fixed-length accelerator batch.
How often that happens is not published. CDL releases cohort and graduate counts, not attrition rates, so treat any specific drop percentage you see quoted as unsourced.
Creative Destruction Lab equity and fees: what CDL takes
CDL is not buying your equity. It is spending your time. That is the actual price. HEC Paris states plainly that the CDL program at its site takes no equity and charges no fees, and that it is funded through philanthropy (HEC Paris, Creative Destruction Lab Paris).
There is no single global equity FAQ covering every location, so verify terms on the page for the specific stream and site you are applying to. If a term is not published for your site, it is not published: ask before you commit, and get the answer in writing.
Which CDL streams and site to apply to
Pick the stream by its mentor list. The city is a logistics decision. CDL started at the Rotman School of Management at the University of Toronto and by 2024/25 ran across 13+ locations and dozens of streams (CDL 2024/25 graduate companies).
| Decision | Get this right | Common mistake |
|---|---|---|
| Stream | Match the mentor pool to your hardest unsolved problem | Choosing the broadest-sounding label |
| Site | Go where your customers and follow-on investors already are | Choosing the closest campus |
| Timing | Apply when a mentor could move something inside eight weeks | Applying at idea stage to "get feedback" |
Is the CDL accelerator worth nine months?
CDL is worth it if your bottleneck is judgment and access to technical operators. It is a bad trade if your bottleneck is a sales hire.
CDL reports that alumni have collectively generated over CAD $56 billion in equity value since inception (CDL 2024/25 graduate companies). Read that as a cumulative alumni figure, not a per-company expectation.
The program also publishes a research dataset covering roughly 15,000 applicants, 9,000 founders, and 2,000 mentors across 27 advanced-technology domains (CDL data and research). A program that instruments itself that heavily is telling you what it optimizes for: objectives that can be measured on a schedule.
When this matters for your raise
CDL's payoff for a raise is the mentor list, not a demo day, because there isn't one. If you are applying to more than one program, the how to apply to Entrepreneur First and how to apply to Techstars guides cover the same ground for those programs, and seed deep tech VCs lists the funds most likely to lead a round after a program like this one.
Most founders applying here are raising either way, and the ones who do not get in still need a round. Causo matches you to the investors most likely to fund your stage and sector and drafts the outreach, so a rejection does not cost you a quarter.
Before you commit to one program, check what each accelerator takes in equity, 2026 accelerator application deadlines and published accelerator acceptance rates, then see how the terms compare against the South Park Commons application.
Run the raise in parallel
Applying is not a funding plan. Programmes like this one accept a small share of applicants and run on their own calendar, so the founders who come out ahead are the ones who kept a funding process moving while they waited. CDL runs nine months with no cheque attached at most sites, so a funding process has to run alongside it.
Start from deep-tech seed funds actively deploying and work the list while your application sits in the queue. If you get in, you arrive with warm conversations already running. If you do not, you have not lost a quarter.
FAQ
Does Creative Destruction Lab take equity? HEC Paris states plainly that the CDL program at its site takes no equity and charges no fees, and that it is funded through philanthropy (HEC Paris, Creative Destruction Lab Paris). There is no single global equity FAQ published across every CDL location. Confirm the terms on the page for the specific stream and site you apply to, and get the answer in writing before you commit.
How does the CDL objective-setting process work? CDL's program page describes mentor-led, full-day sessions held roughly every eight weeks, where short-term objectives are set and progress is assessed (CDL program page). You present, mentors set objectives with you, and at the next session you report against them. Objectives are meant to be specific enough that both you and a mentor can be held to them.
Can mentors drop a company from CDL between sessions? Yes. Mentors decide between sessions whether they want to keep working with a venture, which is how companies leave before the program year ends. CDL does not publish a session-by-session attrition rate, so treat any specific drop percentage you see quoted as unsourced.
Which CDL stream should I apply to for AI / climate / health? Pick the stream whose published mentor list contains people who can unblock your hardest technical or commercial problem, not the stream with the broadest label. CDL ran more than 20 specialized streams across 13+ locations in its 2024/25 program year (CDL 2024/25 graduate companies). If you cannot name two mentors in a stream you specifically want in the room, it is the wrong stream.
Is CDL the same as an accelerator or is it different? It is structurally different. CDL runs a roughly nine-month cycle of full-day objective-setting sessions rather than a fixed batch ending in a demo day, and mentors decide between sessions whether to continue with a venture. CDL hosted over 630 companies in its 2024/25 program year across more than 20 specialized streams (CDL 2024/25 graduate companies).
Related on the hub
- How to apply to Techstars in 2026 (application guide) — Related accelerators guide.
- How to apply to a16z Speedrun in 2026 — Related accelerators guide.
- How to apply to 500 Global in 2026 — Related accelerators guide.
Run this raise inside Causo.
Match to the best-fit partner at 1,000+ funds, draft a hyper-specific email, and send from your own inbox, in one place.