Some Investors Sit In On Your Sales Calls Before They Wire. Here Is How To Pass That Test.

Reference calls you expect. Customer calls you expect. What a growing number of early-stage investors now want is different: they want to sit on the line while you sell, live, to a stranger who has not agreed to buy anything.

Sean Jacobsohn, a partner at Norwest, described the practice to Crunchbase News on September 9. His portfolio spans fifteen active companies, from pre-revenue to more than $300 million in revenue, mostly finance and HR software. Before he invests, he said, he will go on a lot of sales calls that he sets up with the CEO, specifically to see how good they are at selling. Asked whether he has ever passed on a founder for failing that test, his answer was yes. When prospects consistently decline a second meeting, he walks.

That last detail is worth sitting with. The metric is not whether you were articulate. It is whether the person on the other end wanted to talk to you again.

A pitch meeting is a rehearsed performance for an audience that has already agreed to listen. A sales call is an unrehearsed performance for an audience that has not. Those two situations test almost nothing in common, and investors have figured that out.

Jacobsohn frames his sales background as the thing that separates him from most venture investors. He held senior roles at WageWorks, Cornerstone OnDemand and Upwork, all three of which went public. His stated view is that nearly every CEO he backs comes from product or engineering, and that this alone is not enough, because the job requires selling to customers, partners, investors and employees.

The reason this matters more in 2026 than in 2021 is capital concentration. Crunchbase data shows fintech funding rose nearly 23% year over year in the first half of 2026 while deal count fell 25.7%, from 2,161 deals to 1,605. Fewer, larger checks means each investment carries more weight in a fund, and each diligence process gets more invasive. When an investor writes ten checks a year instead of thirty, four hours on your sales calls is cheap.

Founders preparing for a ride-along tend to over-prepare the demo and under-prepare everything around it. The demo is the least diagnostic part of the call. Here is what the person on mute is actually scoring.

Whether you diagnose before you prescribe. The most common tell is a founder who opens the product within ninety seconds. Investors watch for whether you ask enough questions to know which problem you are solving for this buyer before you start solving it.

Whether you can hold silence. After a hard question, do you wait, or fill the gap with a softer version of the question? Filling the gap reads as discomfort with the buyer's discomfort, the same discomfort that stops founders from asking for money later.

Whether you handle the objection or route around it. Rerouting is easy to spot. The prospect raises pricing, you acknowledge it in a sentence and pivot to a feature. An investor hears a founder who has not built a real answer.

Whether you ask for the next step and name it. A vague "I will follow up" ending is the version most likely to produce no second meeting, and no second meeting is what made Jacobsohn walk.

Failure one: selling the roadmap. A founder under observation feels pressure to look impressive and starts describing what the product will do in six months. Buyers discount roadmap claims heavily. Investors discount them more, because the roadmap is the part you control least.

Failure two: talking to the investor. The prospect asks a question and the founder answers in a register aimed at the person on mute, with market sizing a buyer never asked for. Every minute performing for the investor is a minute not spent qualifying the buyer, and the investor knows the difference.

Failure three: stacking the deck. Some founders schedule ride-alongs only with warm intros from friendly design partners. It produces calls that close easily and prove nothing, and an experienced investor will ask how the meeting was sourced.

Failure four: no consistent second meeting. This is the aggregate failure and the one stated out loud. Individual calls go badly for reasons outside your control. A pattern of prospects declining to continue does not.

Treat the ride-along as a consulting engagement where the payment is the investor's attention.

Set the sample size before the first call. Three calls is noise. Five to eight across different buyer profiles produces a pattern. If you do not know whether your pattern is good, find that out privately before you find it out in front of a term sheet.

Mix sourcing deliberately. Include at least two cold or lightly warm prospects. If every call is a referral from an existing customer, the exercise measures your network rather than your selling.

Debrief immediately, out loud, before the investor forms a private opinion. Say what you thought went wrong. A founder who diagnoses their own call accurately reads as coachable, which is worth more than one who ran a flawless call and cannot explain why.

Keep call notes structured enough that the pattern is legible later. Founders working through this on Foundra log the objection raised, the question that changed the buyer's tone, and why a second meeting did or did not get booked, because those three fields turn eight calls into an argument rather than eight anecdotes.

Ask for the investor's read. Most will give it, and that feedback is often worth more than the check.

If the prospect of this test makes you uncomfortable, the discomfort is the finding. The underlying skill responds quickly to volume and structure.

Run twenty calls yourself before delegating any. Founders who hire a first salesperson before personally closing customers hand over a process nobody has debugged, and the salesperson then fails for reasons the founder cannot diagnose.

Record and review your own calls for one thing: the ratio of time you spent talking to time the buyer spent talking. Early-stage founder calls routinely run seventy percent founder. Getting that under half is the single highest-yield change most technical founders can make.

Write down every objection verbatim for a month. Then write one honest answer to each, including the ones where the answer is that your product does not do that yet. Founders who handle objections well are not improvising. They have answered it forty times.

Practice asking for the next meeting with a specific date and attendee. It feels aggressive the first ten times and stops around the fifteenth.

There is a common defense that goes: the product will sell itself, and the founder's job is to make the product good enough that selling becomes unnecessary. In some markets this has been true.

Jacobsohn's framing on where durable companies get built cuts against relying on it. He told Crunchbase News that a simple horizontal workflow for small businesses is now easy enough to build that it invites heavy competition, while complex products for the midmarket and enterprise, or anything vertical requiring deep domain expertise, are hard for competitors and hard for customers to build internally. The second category is where he sees durability. It is also the category where nothing sells itself, because the buyer is a committee with a procurement process.

The uncomfortable implication is that the defensible markets and the sales-intensive markets are increasingly the same markets. Choosing to build somewhere selling does not matter increasingly means choosing to build somewhere anyone can copy you.

Not every founder should agree, and not every request is reasonable.

If you are pre-product and every call is a discovery interview, a ride-along measures your interviewing, a different skill that should be framed that way rather than scored as selling.

If you sell through a channel or product-led signup with no human touch, the request is a category error. Offer the equivalent artifact: activation and conversion data, and a walkthrough of where users drop.

If the investor wants to observe calls with fragile named accounts you are actively closing, decline that set and offer a different one. Protecting a live deal is defensible, and an investor who cannot hear it is telling you how they behave post-investment.

And if an investor wants twelve calls across six weeks while giving no signal on interest, you are being used for market research. Ask for a conviction level and a timeline before call three.

Is this new, or has it always happened quietly? Operator-turned-investors have done versions for years. What changed is that it is now described publicly as a formal gate, and tighter deal counts give investors the time and incentive to run it.

Should I offer a ride-along before being asked? If your calls are good, yes. It converts a defensive test into a proof point, and very few founders offer, so it differentiates. If you have not reviewed your recent calls, do that first.

What if I fail? Ask for the specific reason and treat it as free coaching from someone who has watched hundreds of founders sell. A pass for sales weakness is recoverable in a way that a pass for a broken market is not, because sales ability is trainable.

Does a strong technical cofounder cover for a CEO who cannot sell? Sometimes, since the company can hire around it later. But the test is run on the CEO, because the CEO has to sell to investors, candidates and partners as well as customers.

How many calls is a reasonable request? Five to eight is typical. Beyond that, ask what the extra calls would produce.

Does this apply outside enterprise software? The literal ride-along is most common in B2B. The underlying question, whether a founder can create demand in a live unscripted situation, gets tested everywhere. In consumer it shows up as scrutiny of organic acquisition.

1. The Sales Test This Norwest Partner Gives Founders Before He Will Invest, Crunchbase News, September 9, 2026
2. Fintech Funding Surges 23% In H1 2026 As Investors Concentrate Their Bets, Crunchbase News
3. Mapping The Future Of HR Tech, Norwest
4. Global Startup Investment Hit Record $510B In H1 2026, Crunchbase News
5. The Failure Museum, Sean Jacobsohn