the founder takes the call on a tuesday. lead investor from the seed round, friendly tone, "we want to lead a bridge — same cap as the last round, $2M, you'll be back in market in nine months." it feels like a save. no priced down round, no signal damage, no new investor process.
the founder signs four weeks later. eighteen months after that, they raise a series a and the cap table comes back with numbers nobody warned them about. the bridge wasn't a bridge. it was a down round wearing a friendly name, and the cost showed up in the waterfall instead of the headline.
bridges in 2026 are not the bridges of 2018. the structure has shifted, the language has shifted, and most founders are still pattern-matching on a word that no longer means what it used to.
what the bridge actually looks like in 2026
three structures dominate, all called "bridges," all priced differently.
the same-cap safe with mfn and 2x preference. the lead offers a safe at the last round's cap. looks flat. then the side letter shows up — most-favored-nation clause that re-prices the safe to match anything cheaper that comes after, and a 2x liquidation preference that survives conversion. on a $20M exit with $4M raised, that 2x preference takes $8M off the top before common gets a dollar.
the convertible note with a discount. sometimes 15%, sometimes 20%, applied at the next priced round. founders model the conversion at the headline series a price. the discount means the note converts at 80-85% of that price, which on a $25M post-money is a $4-5M valuation gap that translates directly into 1-2 extra points of dilution.
the insider-led flat extension. existing investors put in $1-3M at the last cap, sometimes with a small bump for participation. no new lead. on the surface, this is the friendliest version. underneath, it tells every outside fund that nobody new wanted to lead, which prices the next round before the founder even walks in.
each of these is sold as "we're keeping the cap flat because we believe in you." what they're actually doing is moving the concession from the line founders look at, the cap, to the lines founders don't look at — preferences, discount, and signal.
why the founder agrees anyway
three reasons, all of them honest.
one. cash is low. a bridge that closes in four weeks beats a priced round that takes four months. when payroll is six weeks away, the headline cap is the only thing that registers — the founder reads "same cap" and stops reading. the side letter gets a quick scan and a signature.
two. the founder treats it as a favor. the lead is being supportive. the lead has reputation at stake too. it feels like betrayal to push back on terms when someone is bridging you. the founder doesn't realize that the lead's job is to underwrite a return, not to be a friend, and that a 2x preference is how they underwrite the return when the cap won't move.
three. nobody runs the math on the next round. a bridge looks like 10% dilution at signing. it actually adds 4-8 points of effective dilution at the series a, once the mfn, the discount, or the down-round signal plays through. founders model the bridge in isolation and never carry the cost forward to where it actually lands.
what the conversation should sound like
the bridge call comes in. before you answer, the question to run isn't "is the cap flat?" it's "is the waterfall flat?" — the cap, the preference, the participation, and the signal to the next lead, all priced together.
a founder who knows the math asks the lead directly. is this 1x non-participating preference? is there an mfn? is there a discount at conversion? would you lead the priced round at the same cap, or are you parking until someone else sets the price? the answers tell you whether this is a bridge to the next round or a structural concession written in a calmer font.
a bridge with 2x preference and an mfn is not a bridge. it is a down round priced in the waterfall instead of the cap, and the founder is the one paying.
what to push back on, and what to walk from
three specific levers, in order of importance.
preference must be 1x non-participating. anything above 1x is a flag. anything participating is worse. the lead gives this up easily if you ask, because they wrote it in expecting negotiation.
no mfn unless it has a sunset. mfn clauses that survive the series a re-price the entire bridge at the lowest valuation any future investor pays for the next 24 months. push for a 6-month sunset, or remove the clause entirely.
no discount, or a discount capped at 15%. discounts above 20% mean the lead expects a flat or down series a and is locking in the upside. if the lead expects that outcome, the founder should know before signing, not after.
if the lead won't move on any of these, the "friendly bridge" is a priced concession. the founder is better off taking a down round at a 25% discount with clean terms than a flat-cap bridge with 2x preference and an mfn, because the down round preserves the waterfall and the bridge doesn't.
how zift handles this
zift gives founders a live runway view that tells you exactly how many weeks you have before the bridge conversation becomes a desperation conversation. every monday morning, the briefing shows current burn, cash on hand, and projected zero-cash date. when a lead offers you a bridge, you know whether you're negotiating from twelve weeks of cash or four, and that delta is the difference between walking away and signing the term sheet.
if you're a finance lead at a series a team modeling bridge scenarios against post-conversion cap tables, zift handles the runway side — Carta handles the cap table itself.
the friendliest investor will still take a 2x preference if you don't ask. the cap is the headline. the waterfall is the price.