Seed Extension or Bridge: The Four Days a Hardware Startup Spent Deciding

Archi4 min read

startupsfundingfinance

Contents
  1. The Setup: A Camera Startup With Nine Weeks of Cash
  2. What We Measured First
  3. The Decision the Founders Actually Made
  4. Where the Numbers Landed
  5. What the Test Logs Showed About Both Paths
  6. The Counterfactual We Ran
  7. Three Things I Would Track Next Time
  8. The Part That Surprised Me

The Setup: A Camera Startup With Nine Weeks of Cash#

This is field notes from a two-person finance and ops team at a small computer vision company, 14 people, selling a $400 camera module to industrial customers. They asked me to sit in on the week they had to choose between a bridge round and a seed extension. I kept the scorecards, the calendar, and the bank statements. What follows is what the data showed, not what the deck said.

The company had 9 weeks of cash at current burn. It had a signed term sheet for a $1.2M bridge from two existing angels, priced as a convertible note with a 15 percent discount and a $9M cap. It also had three new funds asking for a seed extension at a $16M pre-money valuation. On paper the extension looked better. In practice it took 11 weeks to close the last time they ran a round.

What We Measured First#

Before any calls, we built a one-page cash model. Three rows: payroll plus contractors, cloud and tooling, inventory deposits. One column per week for 26 weeks. The model made two things obvious. First, a bridge bought 6 months of runway, not a year. Second, the seed extension, if it closed on schedule, bought 14 months and a lower dilution number. The gap between those two futures was about $240K in founder equity.

We ran the same model against three closing scenarios: 8 weeks, 12 weeks, and 16 weeks. At 16 weeks the bridge was the only path that did not miss payroll. The seed extension required a bridge of its own, which the new funds would not lead.

The Decision the Founders Actually Made#

They split the round. This is the part most writeups skip. Instead of choosing, they asked the existing angels to close $600K immediately at the same note terms, then run the seed extension in parallel. The new funds agreed to a $500K side letter with the same cap. Total raised on paper: $1.1M. Total closed in 10 days: $600K.

That split created a new problem. Two instruments with different caps and different pro rata rights. The cap table went from one page to four. Their lawyer billed 19 hours in a week. The CFO spent three days reconciling the two sets of terms into a single waterfall model so the board could see what a $40M exit would pay each holder.

Where the Numbers Landed#

, Cash in the bank at decision time: $310K., Weekly burn before the bridge: $34K., Weekly burn after the bridge: $29K. They cut two contractor slots and moved cloud spend to a reserved instance plan., Weeks of runway added: 26., Legal and admin cost of the split structure: $21K., Time the founders spent on fundraising during those 10 days: 62 hours, measured from calendar blocks.

That last number matters more than it looks. Sixty-two hours is roughly 1.5 weeks of one founder's output. The hardware team shipped a firmware update late because of it. The customer who was waiting for that update renewed anyway, but the support ticket volume that week doubled.

What the Test Logs Showed About Both Paths#

We ran a simple sensitivity check on each option. The bridge had one variable that mattered: whether the angels wired on time. They did, in 6 business days. The seed extension had four variables: diligence completion, partner sign-off, legal docs, and wire timing. Each one had a plausible slip. We modeled the chance of closing inside 8 weeks at about 55 percent based on the company's own history. That number, not the valuation, drove the decision.

The Counterfactual We Ran#

After the close, I asked the CFO to model the bridge-only path. If they had taken the full $1.2M note and skipped the extension, they would have hit 32 weeks of runway, kept the cap table clean, and gone out for a priced round in month seven with better metrics. The dilution would have been heavier on paper. The admin cost would have been $4K instead of $21K. The founders disagreed on whether that was better. One said the clean structure would have saved them a month of legal back and forth. The other said the split let them keep the valuation anchor for the next round.

Both are defensible. The data does not settle it.

Three Things I Would Track Next Time#

First, wire timing, not term sheet timing. A signed term sheet is not cash. The company had two signed documents that took 3 weeks to fund.

Second, the founder hours. Put fundraising on the calendar as a cost line. It competes directly with shipping.

Third, the cap table page count. Every new instrument adds legal work at the next round. The cost shows up 18 months later, not in the current quarter, which is why nobody prices it in.

The Part That Surprised Me#

The hardest decision was not bridge versus extension. It was whether to tell the team. The founders waited 9 days before the all-hands. In that window, two engineers heard rumors and started interviewing elsewhere. One left. The replacement search took 7 weeks and cost more than the legal bill. The company now puts a cash runway number on the wall every Monday. That single change did more for retention than either funding instrument.

More articles