Framework
The barbell of cancelability: be ridiculous or buttoned-up, never middle
Immad's heuristic: people in the middle of the spectrum can be canceled, but if you're either completely ridiculous or completely buttoned-up you're cancel-proof. You have to pick an end of the barbell.
“There's like a barbell-ness to being cancelable. Like if you're in the middle, you can be canceled, but if you're completely ridiculous or completely like buttoned up, then you can't be canceled. So you gotta pick one.”
Idea
Build an API-first, developer-first Salesforce competitor
Immad has wanted a good Salesforce competitor for 4-5 years. His favorite angle: an API-first, data-first CRM that becomes the single source of truth for customer data scattered across databases, analytics, Stripe, and marketing tools, since Salesforce's API access is gated behind ~$200/seat tiers.
“a Salesforce competitor that was actually good. Every time I say this, like lots of people are like, it's not possible. Salesforce is so like entrenched and all this stuff, which is quite similar to things that people were saying to me when I started Mercury in 2017. But I think everyone hates using Salesforce, right?”
Steal thisBuild a CRM that's API-first and the single source of truth for customer data, instead of locking the API behind a $200/seat upgrade.
Idea
A 'Clubhouse for business' to recreate water-cooler serendipity remotely
Immad pitches software-plus-hardware that recreates office serendipity for remote teams. He's invested in Sidekick, which gives you a tablet next to you so colleagues feel like they're sitting beside you, with settings for happy hours, and believes the right form factor is still up for grabs.
“I like someone's phrasing for this is like a Clubhouse for business. So it doesn't make sense in like a tiny 3 or 4 person company, but how do you, once your companies get bigger, like how do you create like the serendipity where like people can have conversations.”
Steal thisPair lightweight always-on hardware with software so remote colleagues feel physically adjacent, recreating the water-cooler moment.
Idea
A newsletter ad network for creators Substack won't serve
Immad predicts someone will build a newsletter-for-newsletter ad network because Substack refuses to do advertising on principle but can't stop creators from wanting it. He notes Facebook makes roughly $4/user/month on a free product, showing how lucrative ads are with the right audience and data.
“And actually, this is a good startup idea. Someone is probably going to make like a really nice newsletter for newsletter ad network, right? Like, because Substack's not going to do it. I don't think they can stop the creators doing— wanting it.”
Steal thisBuild the ad-sales and matching layer Substack refuses to touch, so newsletter creators can monetize with ads without hiring a sales team.
Idea
Embedded fintech: let any SaaS platform offer loans on its own data
Immad describes the 'embedded fintech' trend: a SaaS platform with access to business data and payments can offer its users loans (or cards), historically only possible for $5B+ fintechs. He invested in LendFlow, which stitches this together for software companies.
“There's this whole trend called embedded fintech. I don't know if you're familiar with it, but the idea is like, let's say you run like a SaaS platform and you have like all this access to like business data, maybe you're helping with payments and things like that. What if you offered these people on your platform a loan?”
Steal thisIf your software already holds a niche's business data, embed lending or cards to add revenue and 10x stickiness.
Fact
Embedded financial products make a SaaS tool 10x more sticky
Immad's reason embedded fintech matters beyond revenue: once your platform powers a business's cash flow, not just a workflow, it becomes roughly 10x more sticky and far harder to churn out of.
“I mean, there's— revenue is definitely part of it, but the other thing is the stickiness, right? Like, if you're— if this platform is not just like something I use as a SaaS tool, but it's something that's like really powering like the cash flow of my business, Now that's like 10x more sticky.”
Number
Mercury took 18 months and 8 people to get good
Immad started Mercury in August 2017, got its first alpha user in January 2019, and considers the April 2019 launch the point it was good, about a year and a half in, with 8 people total: 6 engineers, 1 designer, and 1 product/BD person.
$8
Team size to build Mercury to launch · people
“So we started in August 2017, and the first alpha user was January 2019. So a year and 3 months before we had any alpha users, and even then it was pretty crappy. I would say when we launched in April 2019, which was basically a year and a half from start, that was when it was good. For most of that time, we had 8 people working on it, including me, 6 engineers, 1 designer, and 1 kind of product BD person.”
Fact
Banks treat depositors as a cost center, not a customer to win
Immad explains why challenger banks exist: traditional banks make money on lending, so they view depositors purely as a cost (branches, signups, support) rather than potential. That blind spot is exactly what lets Chime, Mercury and others win.
“the reason like all of these challenger banks exist is because banks don't care about deposit customers. Like, that's like the fundamental issue in banking where like they think of depositors as like like cost center, because it costs them a lot of money to like have someone walk into a branch, like sign up for a bank, and then like worry about all this stuff. Where they make money is lending.”
Take
Win the bank account first; loans are a commodity, deposits are the relationship
Immad's strategy: a freemium model where a few customers with $1M+ subsidize thousands of small ones, plus owning the weekly-touch bank relationship. From the deposit relationship and its data you can easily build lending; going the other way from commodity loans to a relationship is much harder.
“It's much harder to go the other way where you're just like a loan provider and you try to build like a long-term relationship. Because loans are like commodity, you know, at the end of the day, you want to get the cheapest loan. Whereas the bank is like something you use every day, you want to have the best product, I think.”
Take
Seed diversification: 1,000 losers are fine if one is a $100B company
Immad's case for spraying seed bets (he has 180+): no investor can tell at the $10M stage which company becomes $100B, but a single $100B hit covers a thousand failures. So he optimizes for diversification and the shot at the outlier.
“really what you really want one as a seed stage investor is $100 billion company. And I don't care how good you are as a seed stage investor, you can't tell when it's a $10 million company, if it's going to be $100 billion company. But if you can hit one of those, like you can have 1,000 investments fail with like one $100 billion company and you're going to be like”
Steal thisAt seed stage, diversify aggressively and optimize for the one outlier; one $100B winner pays for a thousand losers.