8 min read · 1,464 words
At a glance
- Three terms to keep out of a bridge: a discount that compounds, a liquidation preference that stacks, and a board observer seat for every bridge participant.
- Three terms to keep in: a cap on the price, a target close on the next round with consequences if missed, and a defined use of proceeds the company will defend if asked.
- Two bridges and a down round can push the preference stack to $14M before common sees a dollar of exit value. The worked math is below.
Three terms decide whether a bridge protects the next round or quietly reprices it: the discount, the preference treatment, and the observer rights. A bridge done right is a useful instrument; a bridge done by template is a set of mechanics that the next lead will read, price, and hold against the company. Most bridge damage is signed in an afternoon and discovered eighteen months later, in diligence, by someone else.
When is a bridge the right call?
A bridge is the right call when it funds a named decision with a dated resolution. The company knows the one material call it has to make, knows what each path costs, and needs capital to execute the chosen path before the next round can price it. That is a bridge doing its job.
A bridge is a delay tactic when its real purpose is to avoid a conversation. The two most common avoided conversations are the down round that everyone can see and nobody will open (we walked the timing in The down-round conversation: how to time it) and the raise that has been “in process” for six months without a lead (the full anatomy is in How the stalled raise actually happens). A bridge taken to postpone either conversation does not remove it; it moves it twelve months out and adds a preference layer to it.
The test is one question: can you name the decision the bridge funds and the date that decision resolves? If the honest answer is “it buys us time,” the bridge is a delay tactic, and the next round will price it as exactly that.
What should stay out of the term sheet?
-
A discount that compounds. A 20 percent discount stacks against the next round’s price; if the round comes in flat or down, the discount converts into more dilution than the bridge dollar bought. Run the math once and the mechanism is obvious: a $1.5M bridge at a 20 percent discount converts as if it were $1.875M of the new round. Into a strong up round, that premium is tolerable; the price absorbs it. Into a flat or down round, the new lead reprices around the discount, the founder eats the difference in share count, and the bridge investors hold more of the company than the check warranted. Cap the discount, or replace it with a fee. A capped discount or a flat commitment fee keeps the bridge’s cost fixed in dollars instead of compounding in shares.
-
A liquidation preference that stacks. If the bridge converts as preferred, every dollar of bridge preference sits in front of the next preferred. Two bridges and a down round, and the founder discovers the cap table does not pay until $14M of exit value. The full math is in the next section; the fix belongs here. Negotiate the bridge to convert into the next round’s security, same class and same preference terms, rather than minting a senior layer of its own. Bridge investors get their conversion economics. They do not get a permanent position in front of everyone who funds the company after them.
-
A board observer seat for every bridge participant. If you are stacking bridges, you are stacking observers. A $1.5M bridge syndicated across five participants can add five people to every board call, none with a vote, all with opinions, each one a reason the room talks more carefully. Observers chill exactly the conversations a bridged company most needs to have: the kill decision, the down-round timing, the leadership question. Limit observer rights to the lead, and sunset the right when the bridge converts.
How does the preference stack reach $14M?
Walk one company through it. Series A raised $8M with a standard 1x non-participating preference. Eighteen months later the next round is not coming together, so the insiders write a $1.5M bridge that converts as preferred with its own 1x preference. Twelve months after that, a second $1.5M bridge on the same structure. The round that finally closes is $3M of new money, down, 1x preference.
Add the stack: $8M plus $1.5M plus $1.5M plus $3M is $14M of liquidation preference sitting in front of common. At any exit below $14M, the founders and every employee with options receive $0. And the stated math is the friendly version: if the two bridges carried 20 percent discounts into their preference price, the bridge layers mint roughly $3.75M of preference instead of $3M, and the clearing bar moves past $14.7M.
Now the part that surprises founders. A $12M acquisition offer, which would have been a real outcome for this company at Series A, dies quietly. Nobody on common has a reason to support it, management has no incentive to run the process, and the acquirer reads the cap table and lowers the offer further. Nobody signed up for a $14M floor. They signed three reasonable documents, eighteen months apart, and the stack compounded silently between signatures. That is the property to respect: preference stacks are never chosen, they accumulate.
What should stay in the term sheet?
A cap on the price. A cap gives bridge investors a defined best case and gives the company a known worst case in shares. Unlike a compounding discount, a cap is a number everyone can model on day one, and the next lead can price around it without repricing the company.
A target close on the next round, with consequences if missed. A named date for the round, and a defined step if the date passes: the cap steps down, or the rate steps up, by an amount both sides priced in daylight. The consequence is not a penalty. It is a forcing function that keeps the raise honest and keeps the bridge from quietly becoming permanent capital.
A defined use of proceeds the company will defend if asked. Not a gesture line in the recitals. A page the board can check. That page deserves its own section.
What does the use-of-proceeds page look like?
One page, and every bridge dollar named to a decision, not a runway extension. “Extending runway” is not a use of proceeds; it is the absence of one. Runway is what the company gets if the named decisions resolve well.
For a $1.5M bridge, the page reads like this: $600K to resolve the mid-market keep-or-kill by October 31, including the restructure cost if the answer is kill. $500K to take the enterprise pricing change live and through one full renewal cycle, with the restated NRR reported to the board in January. $400K to make two named hires against signed enterprise pipeline, released only after the pricing change holds. Each line has an owner, a date, and a number the board can check without asking for a meeting.
A bridge is a forward decision dressed as a capital event. Treat it like the decision: name the paths, put dollars on each, set the date, and write down what the company will do when the date arrives.
What is the Decisive Finance role in this?
The right time for a Diagnostic is before the bridge is signed, because the Diagnostic produces the raw material the term sheet needs: the three headline ratios restated, the material decision named with Keep/Kill/Restructure paths in dollars, and the use-of-proceeds page drafted against real numbers. Guaranteed 3x your Diagnostic fee in recoverable value, in 14 days. Typical 5x to 10x. The recoverable value matters here for a structural reason: every dollar recovered is a dollar of bridge the company does not raise, and a smaller bridge with named uses beats a bigger bridge with stacked preferences in every version of the next round.
Where to go from here
If a bridge term sheet is on the table this quarter, run the three exclusions against it before counsel marks it up.
Read the guarantee Book the call
Not sure whether you are bridging into a round or into a wall? Run the Next Call Self-Assessment and take the dashboard into the term-sheet conversation.
Related reading: