Most vape brands don’t fail because their hardware is bad. They fail because their terpene sourcing is a mess, and by the time they notice, they’ve already burned through three suppliers and a chunk of their runway.
So here’s a walk-through of how a terpene sourcing journey actually plays out, month by month. This is an illustrative composite, not a specific named company. Think of it as a representative example built from the patterns you see again and again with early-stage vape brands. Call it a small vape startup with big plans and a modest budget. The numbers are approximate ranges, framed as typical, not exact figures pulled from one real client.
The point isn’t the story. It’s the lessons at the end of each phase, because those are the parts you can steal for your own cannabis terpene sourcing decisions.
Month 1: The Sample Hunt
The startup starts where everyone starts, by requesting samples. Lots of them. They reach out to five or six suppliers found through trade shows, LinkedIn, and a few Google rabbit holes.
What they learn fast is that “terpene supplier” means wildly different things. Some sell botanical terpenes rebottled from bulk. Some blend flavour compounds and call them strain profiles. A couple actually source from real cannabis analysis and reverse-engineer the profile with plant-derived isolates.
The samples arrive over three weeks. Some come with a certificate of analysis. Some come in an unlabelled amber bottle with a handwritten strain name and nothing else. That alone tells you a lot about who you’re dealing with.
The best-organised suppliers ask questions back. What’s your hardware? What viscosity are you targeting? Are you cutting distillate or running live resin? One provider even offers a structured sample evaluation process to start from, which turns a random bottle grab into an actual comparison you can act on.
Lesson: A sample without a certificate of analysis is just a smell. If a supplier can’t tell you what’s in the bottle at the compound level, you’re guessing, and guessing doesn’t scale.
Month 2: Testing In Real Cartridges
Smelling a terpene on a blotter strip tells you almost nothing about how it behaves in a cart. So the startup builds a proper test matrix.
They fill small batches of their actual hardware with each candidate profile and score them across five things:
- Aroma on the nose, cold and after a few draws
- Taste on the inhale and the exhale, because they’re often different
- Viscosity and how it flows into the coil
- Coil compatibility, checking for clogging, wicking issues, and burnt notes
- Two-week stability, watching for separation, colour shift, or aroma dropping off
This is where a lot of samples die. A profile that smelled incredible on day one goes flat by day ten. Another one is beautiful but so thin it leaks around the seals. One tests great on aroma and taste but throws a harsh note the moment it hits a warmer coil.
The two-week window matters more than people expect. Customers don’t vape a cart the day it’s filled. It sits on a shelf, in a car, in a pocket. If the profile can’t hold up for fourteen days in testing, it won’t survive real-world distribution.
By the end of the month, six candidates are down to two. And crucially, the startup keeps notes on why each one failed, which becomes gold later.
Lesson: Test in your own hardware, not on a strip. Aroma, taste, viscosity, coil behaviour, and two-week stability are five separate tests, and a profile can ace one while failing another.
Month 3: Consolidating To One Partner
Juggling multiple suppliers feels safe. It’s actually a hidden tax.
Every supplier means a different lead time, a different minimum order, a different COA format, and a different person to chase when something’s off. The startup does the math and realises the coordination overhead is eating hours every week that should go into selling.
So they consolidate. They pick the one partner whose profiles performed best in the Month 2 matrix and, just as importantly, who communicated like an actual partner rather than a vending machine. The supplier who asked about hardware and viscosity back in Month 1 is now the obvious choice.
The first scaling order is modest, somewhere around 1 to 2 litres across a couple of profiles. It’s the first time they’re buying to build inventory rather than to test. The per-unit price drops a little at this volume, but the bigger win is consistency. One COA format. One lead time to plan around. One relationship to deepen.
This is also the point where formulation flexibility starts to matter. Off-the-shelf profiles are fine to launch with, but the startup already knows they’ll want tweaks. A partner offering custom terpene formulation for scaling means they won’t have to switch vendors the moment they need something specific.
Lesson: Running four suppliers isn’t diversification, it’s overhead. Consolidate to the partner who both performs and communicates, and pick one who can grow into custom work later.
Months 4 to 6: Scaling And The Reformulation
Volume climbs. The startup moves from test batches to producing roughly 5,000 carts a month, spread across four or five profiles. This is where sourcing stops being a lab exercise and starts being a business function.
Then the customer feedback comes in, and it isn’t all good.
One profile, a fruity hybrid that tested beautifully, gets lukewarm reviews. Customers describe it as “flat” and “candy-like” in a way that reads artificial. Sales on that SKU lag the others by a wide margin. It’s the single most useful thing that happens all quarter, because it forces a real conversation with the supplier.
Instead of dropping the profile, they reformulate it. The partner pulls the original GC-MS data, adjusts the ratio of the dominant terpenes, dials back one compound that was reading as synthetic, and sends a revised sample. The startup runs it back through the Month 2 test matrix. Two rounds later, the profile lands, and that SKU climbs from worst-seller to middle of the pack.
That loop, feedback to data to reformulation to retest, is only possible because they’re working with a partner who has the underlying analytical data and the ability to change the formula. A rebottler can’t do this. They can only sell you what’s already in the drum.
Lesson: Customer feedback is free product development. But you can only act on it if your supplier holds the formulation data and can actually adjust the profile, not just resell it.
Month 7 And Beyond: Enterprise Lock-In
By month seven, the relationship changes shape. The startup is now a reliable, repeating account, and that gives them leverage they didn’t have as a sample-hunter. This is where the grown-up terms get negotiated.
Four things typically get locked in at this stage:
- A volume commitment, where the brand agrees to a minimum monthly or quarterly quantity in exchange for better pricing
- A price lock, protecting them from mid-year raw material swings that would otherwise wreck their margins
- Category exclusivity on their signature custom profiles, so the supplier won’t sell that exact formula to a direct competitor
- Quarterly reviews, covering quality trends, upcoming profiles, and demand forecasting
None of this is possible in Month 1. You earn these terms by becoming predictable and by proving you’ll keep buying. The exclusivity piece is especially worth chasing, because a custom profile that only you can sell is a genuine competitive moat, not just an ingredient.
The quarterly review also flips the relationship from reactive to planned. Instead of scrambling when a raw material shortage hits, the startup hears about it a quarter ahead and adjusts. That’s the difference between a vendor and a partner.
Lesson: Volume buys you terms. Price locks, exclusivity, and planning cadence are things you negotiate into, not features you find on a website. They come from being a serious, consistent account.
How The Costs Evolved
Here’s roughly how the economics shifted across the journey. These are approximate ranges for illustration, not a quote, and your numbers will vary with profile, volume, and hardware.
| Phase | Typical order size | Approx. spend | What changed |
|---|---|---|---|
| Month 1: Samples | Sample kits | ~$150 to $400 total | Paying to learn, not to stock |
| Month 3: First scale | 1 to 2 litres | ~$1,500 to $3,000 | Buying to build inventory |
| Months 4 to 6: Scaling | ~5k carts/month | ~$4,000 to $6,000/month | Volume pricing kicks in |
| Month 7+: Locked-in | Committed volume | Lower per-unit | Price lock plus exclusivity |
The headline isn’t that spend went up. Of course it did. The headline is that the per-unit cost came down while quality and consistency went up, because the relationship matured.
The Key Lessons, In One Place
If you skip straight to the end, here’s what the whole journey boils down to:
- A sample without a COA is just a smell. Demand compound-level data from day one.
- Test in your own carts. Aroma, taste, viscosity, coil compatibility, and two-week stability are five separate hurdles.
- Consolidate early. Four suppliers is overhead, not safety. Pick the one who performs and communicates.
- Pick a partner who can reformulate. Customer feedback is only actionable if your supplier holds the data and can adjust the formula.
- Volume earns terms. Price locks, exclusivity, and quarterly reviews come from being a consistent account, not from asking on day one.
The brands that get this right treat sourcing as a relationship that compounds over time, not a purchase they redo every quarter. The ones that struggle keep starting over with a new bottle and a new stranger.
Where To Start
If you’re at Month 1 right now, staring at a spreadsheet of suppliers and a pile of unlabelled amber bottles, the smartest move is to start with structure instead of guesswork.
Working with a partner who provides real GC-MS analysis, plant-derived accuracy, and cGMP-level consistency means the journey above happens faster and with far fewer dead ends. A good structured sample evaluation turns Month 1 from a chaotic hunt into a clean comparison, and a supplier built for custom formulation at scale means you won’t have to switch vendors the moment your brand outgrows off-the-shelf profiles.
Start the same way this composite startup did. Get real samples, test them properly, and build the relationship from there.
