What 22 Startups Showed Me at the NICAT Rapid Mentor Gala

Product Partner at Biz of Dev. A decade building Products with Founders, one Discovery at a time.
I've been mentoring Founders for about three years now. Mostly it's the informal way: a call that turns into a second call, a voice note late at night, a coffee that runs two hours over. This November I got to do it in a structured way for the first time, with NICAT (National Incubation Center for Aerospace Technologies). I'd count it as one of the most respected startup bodies in the country.
So this is a retrospective on a very long call, and a field report on Startup Product Discovery. Two afternoons, 22 Founders, 15 minutes each, about six hours in one video room. I pointed at a handful of common startup mistakes so many times that writing them down felt like the only sensible thing to do with the notes.
TL;DR:
- Strong Tech, Weak Business: most of the 22 teams had built something that works. Across both days, the tech was ahead of the business.
- The Six Common Startup Mistakes: two or three businesses on one small team, and replacing infrastructure instead of improving it. Global market numbers with no route behind them, and go to market that waits on someone else. Pitching the user instead of the buyer, and pricing that was assumed rather than tested.
- The Six Fixes I Kept Giving: pick your one year battle and name the five year one. Enhance what the buyer already owns. Start with local numbers and one paying customer. Run one channel yourself, and go to the rooms. Speak the buyer's language. Test the price before you set it, then show the loss before the price.
- Where They Live: every one of these sits in the business behind the Product. Each one lands in one of four places: the Business, the Market, the User and the Execution Plan. We break down every Product at Biz of Dev into those same four.
How I Ended Up in That Room
A while back I was chosen as one of the mentors on NICAT's board, and I was really glad about it. It might not be a big achievement for other people. But I really believe in the local ecosystem, and I've been trying to push it forward through my own community work at The Wandering Pro. So getting to extend that help to another body with a similar mission meant a lot more to me than another line on a profile.
I wrote about joining the mentor list on LinkedIn when it happened. The post has the twenty product lessons I'd want every Founder to hear first.

The Rapid Mentor Gala is how NICAT puts those mentors in front of its startups quickly. Founders book short one to one slots, and a mentor takes them back to back. You can see the startups in the NICAT directory, and the mentor directory has the rest of us (I'm in there).
The Format, and What Was Already Working

Day 1, Monday, November 16: 11 Founders, 15 minutes each, from about 2:15 to 5:15. By the end of it I said the pattern out loud. Most of them were trying to replace infrastructure, and most were pitching several businesses at once.
Day 2, Tuesday, November 17: 11 more, from about 2:00 to 5:00, and a different pattern. Most Founders were stuck on pricing, and on who the buyer actually is.
Founders often joined early or ran over, so both afternoons ran the full three hours.
Sitting three hours is a bit boring. Fun people though.
Before I get into what went wrong, the room deserves the credit, because it was strong. These five went at the top of the report I sent NICAT afterward:
- Real Builders: most teams had built something that works. I saw detection models, airflow simulations, security audits, cattle tags, radio devices and home safety sensors. One team had a projector that lets you walk through a house before it exists.
- Problems Rooted Here: the best ideas started from a Pakistani pain the Founders know firsthand. Think tube well theft, grain spoilage, gas leaks at home, power costs in data centers, and hospital systems that don't talk to each other.
- Open to Hard Feedback: Founders took direct feedback well, and a few changed course inside the 15 minutes.
- Honest About the Stage: many were upfront about being early, with no prototype or paying customer yet. That makes them easy to help.
- Resourceful With Money: several fund their R&D without waiting on investors. They use services revenue, competition prizes, and pilots that double as their first case study.
One thing about the quotes before we get into them. They're all mine, from the recordings. A lot of what I said was in Urdu, so the translated lines read a little looser than I'd write them.
Six Common Startup Mistakes From the Gala

Every one of these came up again and again across 22 very different startups. And every one is a business decision that hadn't been made yet. I'll keep the examples short here, because most of the stories belong with the fixes further down. (The mistakes that hit right at launch are their own list, which we covered in MVP launch mistakes that kill startups.)
1. Two or Three Businesses on One Small Team
This was the first thing I noticed on day one, and it never stopped.
It always starts the same way. You get one piece of tech working, and you start seeing it everywhere. The camera that spots birds on a crop can also spot someone stealing a solar tube well pump. The sensor rig that fits a silo also fits a hospital, and the agent that books sales calls can also find the leads. So the pitch grows a second market and then a third, each with its own buyer and its own way of saying no.
An agritech Founder was running exactly that. On the same camera stack, he had a bird deterrence product for markets abroad and a theft alert product for Pakistan. My first question was the obvious one:
Why are you fighting two fronts at the same time as a startup?
A construction tech Founder had six modules, each priced low. An ERP team was selling to small shops and large enterprises at once. And an AI lead generation team was going after several niches in parallel. It keeps happening because the tech can do all of it. As I told the construction Founder, "I'm sure you can research and find ten other use cases for your technology". That won't sell, because the tech travels and the business doesn't.
2. Replacing Infrastructure Instead of Improving It
Around the eleventh session on day one, I finally said the second pattern out loud:
Most people are trying to change infrastructure, and let me tell you one thing, especially in Pakistan, good luck selling new infrastructure.
New silos, a new diagnostic robot, new systems for buildings that already have them. And the buyer's answer is always some version of: "Brother, I've already spent lakhs, crores of rupees on this, built everything, now should I put in a new one?"
On day two it came back from the enterprise side: "Cost saving is not a big enough reason for them to change from a system they are already using".
3. Global Numbers With No Route to the Market
Most market size slides were global figures, billions of dollars. Nothing behind them showed how a team in Islamabad actually reaches that market. The word I heard most on day one was global. An AI lead generation team called themselves a global solution. I told them:
Rather than saying we are a global solution. That is the worst way you can put something.
Part of what pushes Founders there is that building with AI makes it feel like you can do everything at once. I told that team what I'd told a community member with the same idea: "AI makes life easy, but it's very expensive. So we have to focus".
4. Go to Market That Waits on Someone Else
Go to market was weak almost everywhere. Nearly every plan depended on someone else opening a door: an intro, a partner, one well connected person. Very few teams had a channel they ran themselves.
A lot of Founders were also planning on ads and online outreach. For most B2B buyers here, that barely works: "In the Pakistan market, one thing you will see is that people don't take online seriously". And a surprising number had a US launch on the roadmap before they had a paying customer at home.
5. Pitching the Technology to the Wrong Person
Founders kept pitching the person who'd use the product instead of the person who'd pay for it. They spoke the user's language instead of the buyer's, with every feature on the table at once. I reminded one of them what a real sales conversation is like: "here it's fine, I'm your mentor, I can discuss this. But the sales guy doesn't give you a second chance to explain".
On day one, three Founders pitched a "digital twin", and each one meant something different. One was a one-time airflow model, one a 3D model of a building with live sensors, and one a simulation engine. Buyers can't price a buzzword.
A few were also inviting the wrong comparison. The handmade marketplace was up against social tools that give accounts away: "Free accounts are being given out, literally, you can't compete with free".
6. Pricing That Was Assumed, Not Tested
This was the most common mistake on day two. The business model was the weak side across both days. Few Founders had tested whether anyone would pay, or how much. So most of the pricing in the room was assumed:
- One-Off Prices: on products that suit a yearly plan.
- Per Unit Pricing: on hardware, which invites the buyer to compare unit costs instead of their losses.
- Hidden Prices: kept off the website entirely.
- Interest Mistaken for Payment: demos booked and free plans full, with nobody yet asked for real money.
On the missing prices, I was blunt: "Usually if people don't have pricing on their website, my assumption is that it hasn't been studied that much". Buyers are just as lost on the other side of the table. That's why we unpacked why app development cost quotes vary so much.
Six Fixes for Common Startup Mistakes
1. Pick Your One Year Battle, and Name the Five Year One
To the agritech Founder I said it about as plainly as I could:
You need to pick your battles. Right now you have to see what my battle is in the next one year, and then what my battle is in the next five years.
Because "if you sit there chasing both, you won't get anywhere".
I used PakWheels to explain why, since everyone in the room knew it: "Why do different marketplaces exist? Because when you do marketing, or selling, or distribution, the techniques differ". The same idea of listing a car runs as PakWheels here and as Craigslist or eBay somewhere else. Nobody thinks those are the same business.
I also pointed him at KeepTruckin, which most people know as Motive now. It started with the electronic logbook a US regulation made trucks carry. Only years later did it grow into fleet management and dashcams. It's one product first and the family after it, the same logic as the MVP skateboard rule.
So a team of three running two markets is basically running two sales motions, on half the attention each. Neither one gets to revenue. The fix is a time split rather than a cut. Keep the other markets on the roadmap and out of the pitch. Pick the battle for the next year, the one that reaches running revenue fastest. And name the five year battle out loud, so nobody thinks you forgot it. That agritech Founder parked the international product before the 15 minutes were up.
For the construction Founder with six modules, I used a restaurant. A new place opens and tells everyone the breakfast is good, the dinner is good, everything is good. And "people say that means nothing is good". So cut to one or two modules and make one of them earn, "until you're making, from any one module, let's say ten thousand dollars total revenue". Get testimonials and stats behind it before you add the next.
And then own it:
If you want to be known in that niche as that person, you need to sell that niche and be that person.
The ERP team got the same thing on day two, for SME versus enterprise:
Don't get distracted. They say the hard thing is saying no.
Pick one for the next six to eighteen months. If it's enterprise, know what you're signing up for, because "people underestimate how hard enterprise is". I looked into it afterward for the follow-up. A Gartner survey put public sector technology purchases at about 22 months on average, the longest of any sector. Vendor estimates put enterprise ERP sales at nine to eighteen months.
Then specialize. The example I gave was coffee shops. New chains here run an expensive American POS, because nobody local had built one just for them. A specialist product wins a vertical like that here, and then it can travel. The investors agree. Bessemer looked at public vertical software companies and found they trade at higher revenue multiples than horizontal ones. That's mostly because customers stay longer and spend more over time. If you're stuck choosing which vertical or module comes first, start with the Product Prioritization frameworks guide.
2. Enhance, Don't Replace
So if you can enhance the existing thing and promise, like, even 90% of the results, that's the better approach.
That was for a grain storage team building smart silos with three people. Retrofit the silos that already exist, and sell the monitoring. And lead with an angle the owner cares about more than spoilage: "Primarily make it, okay, the bank, you need a loan, this is how it can help you. Talk about money a bit more". Monitored stock can back a loan, and that's worth more to a silo owner than the spoilage it saves.
A health hardware Founder had built an impressive device that runs several diagnostics in one. I asked the uncomfortable question, "Why a robot? Why not software?", because in Pakistan, sending a doctor with the right equipment can be cheaper than sending a robot. That becomes your real competition. The bigger gap he'd talked about was that hospital systems don't talk to each other. And "no hospital is gonna be like, yes, we already have ten machines of ten different brands, now we'll add an eleventh".
So I pointed him at the XKCD comic about competing standards. In it, somebody tries to fix fourteen standards and ends up with fifteen:
If you're trying to standardize something, you can't do it by adding a new standard, you have to standardize the existing stuff.
Software that connects the machines a hospital already paid for is a cheaper first product than new hardware. It also tells you what to build next. It's the same call we walk through in our Build vs Buy guide, seen from the buyer's side. They already made that decision once, and you're asking them to unmake it.
On day two the ERP team heard it as "don't try to replace their systems. Try to enhance their systems". So did a radio hardware team building a bridge for the radios that security and facilities teams already carry: "everyone is trying to change the infrastructure, and my advice is always: please just enhance, don't change". Connect to what the buyer runs, and offer a replacement only where there's real pain. Then give them a reason to start small: "start here, you'll enjoy it so much that then you'll say change everything".
3. Start With Local Numbers and One Paying Customer
Constraints around your product actually help it become price competitive and a better fit for those customers who don't want a global solution.
I told the AI lead generation team about a community member who wanted to build a lead gen tool for the whole world. We narrowed it to the UK, then London, then the trades, then barber shops with strong ratings on Google Maps. That's when it clicked: "Now his product became very easy to sell, because now he knows exactly which businesses he has to go and call".

The same goes for the market numbers. They need to be local and real before they're big: "make that a realistic figure. And when you hit that, then you can expand. Expansion is always in the books". We laid out the method for building those numbers in startup market research.
Sometimes the local number is sitting in the news. Take the construction Founder pitching structural risk prevention. The Sindh Building Control Authority had listed about 741 dangerous buildings across the province, 588 of them in Karachi. Demolitions and a fresh survey were under way. That's a buyer, a number and a news hook in one line, and it beats a global market slide every time.
And size works differently here. With a home safety Founder I put it this way: "a one million dollar business is not the same as a ten million dollar business, is not the same as a hundred million dollar business", and "in Pakistan, what I do is divide these by four, because of our economies of scale". A house built for two floors won't hold eight. So set the first goal for the foundation you have. For him, that meant a quarter of a million dollars a year here first: "once you're at 250k a year, go all in GCC".
Then get the first customer. It's a smaller goal than most Founders set, and a much more useful one:
Get one good paying customer, which meets your cost. The second customer will make you profit. The third customer will make you scale.

4. Run One Channel Yourself, and Go to the Rooms
For most B2B buyers here, I told the Founders to take ads out of their heads: "Here, no online methods are effective, only in person methods work. So take it out of your mind that I'll run ads, I'll do this, I'll do that". Instead, "the best way to get leads is just walk into hospitals and talk to people, or go into events where these people are".
For health tech specifically:
In health tech, marketing works like this: paper work and network. The ads kind of stuff is rare.
Paper work meaning research, an advisory board, partner logos and results measured with real clinics.
Network meaning walking into hospitals, and going where clinicians actually meet. Some sectors in Islamabad have ten or fifteen hospitals in one place. It means joining the industry bodies, the chamber of commerce and P@SHA. Then it means getting into the same rooms as your buyers at professional events. A name someone recognizes opens doors here that a cold message never will:
But in Pakistan, names work.
The rooms matter for a reason that's easy to miss, too: "we're already bombarded with messages, and then the problem is, how do we distinguish what's real versus what's fake? Being in the same room matters a lot".
But rooms alone aren't a channel you control. "There are only four ways you can get customers. You create content, you run ads, you get referrals, or you do cold outreach. Other than that there's no rocket science". So pick one of them and run it yourself, so sales don't stall waiting on a single contact. And give it real time: "You need to pick one and at least run it for six months". The DevOps team wanted to drop Sales Navigator after a few weeks, and I told them, "I'm not a big fan of switching channels before trying out one thing fully". I also gave them a play that works right now: "a self-assessment, or build some audit-type tool, and you can do lead gen with it". And for a climate team with research locked in reports: "I'm not saying go on TikTok and start dancing, but you know, blogs". There's a longer version of the low budget playbook in our poor man's GTM strategy.
And whatever the channel, never lean on one relationship:
One person is never the solution. Don't bet your business on one person.
Then the US. I give every Founder with a US launch on the roadmap the same answer:
First show me that you've bought the tickets to America.
I'm not saying never: "You have to go from here to whatever is closest, which is GCC, and then from there you can maybe expect to go to America". For an HR software team, the US would mean fifty states of payroll compliance, and "every state has its own headaches". So Australia and the rest of APAC first.
A short-let compliance team, already selling to the US through a partner, got the narrow version. You can't take the whole US at once. So pick one county or one less aggressive state, because "New York and California are always a little bit iffy options, because you have to understand you're competing with startups that are based there". Then use recent fines in the news as outreach. When someone in your county just got fined, that's your pitch. New York City shows how concrete it gets. After Local Law 18, registered hosts dropped to about 3,500, against roughly 23,000 listings before 2023. Penalties run up to $5,000 for unregistered stays.
Deadlines work the same way. A security audit team selling to software houses here has a buyer with a date on it. The EU Cyber Resilience Act requires a machine-readable software bill of materials for products sold in the EU from December 11, 2027. Every software house exporting to Europe will need one.
5. Speak the Buyer's Language
To the data center cooling Founder:
The guy who's making the decision, he cares about money, revenue, how much money is it saving me, tell me in PKR, tell me in dollars.
And talk about the ground here rather than the brochure. The multinational vendors model airflow for free to sell their equipment. Their standards might not apply to Pakistan, "the electricity here, the concrete here, the materials here". So "you have to be realistic about the ground reality, that is the experience you're selling". Don't even let the comparison start, because "the way the comparison goes, the talk will then come to price justification".
The DevOps team got the same push from another angle. Startups don't care much about DevOps. They deploy on Vercel and move on. Inside a scale-up, the engineers who use the tool "are not the ones who are gonna be purchasing that". So pick one ICP, "please pick one ICP. Otherwise no one will take you seriously". Find the leadership role over tooling and spend, and sell the saving in their terms: "Your developers, you don't need four people, you need one". Then have a one-line answer ready for the comparison every buyer will make. In their case, it's the internal developer platforms and AI DevOps tools already on the market, because "why not them?" is the first question a scale-up's CTO asks.
We wrote about the user and the buyer splitting apart at length in identifying user requirements. It showed up in almost every room.
The short-let compliance team needed it said plainly too: "Saying there are compliance issues. Okay, but which ones, and why are they important, what is the loss, how much money am I saving by using your product?" Because "just saying we prevent compliance issues is not enough in the current day and age".
A floor plan projection team kept getting compared to VR headsets. The fix was choosing the comparison yourself: "Don't compare yourself to VR. Whenever someone asks that question, that's the wrong answer, because your comparison isn't that". And then, more simply, "You are a marketing tool. You are not a sales tool".
I told them about an animatronics company I worked with back in university. It sold poorly as "animatronics", and took off once it called itself an event experiences company. A walk-through of a home that doesn't exist yet is a billboard for the project.
And for the digital twin Founders, plain words. Say what it actually does (a one-time model, a live sensor feed or a simulation) and what the buyer gets. Keep the technical depth for when they ask. Or, as I told the home safety Founder about his extra features, "you don't even have to mention it, unless asked".
6. Test the Price, Then Price the Loss
On day two I started opening the pricing conversations with this:
In business you have only two kinds of products: one that makes a business more money, one that saves a business money.
"And both are right, but the marketing and GTM and acquisition of the two is very different". A product that makes money gets bought by someone chasing a target. A product that saves money gets bought by someone defending a budget, and "when we come in on the saving side, we need really hard proof of what wastage is happening".

The home safety Founder wanted to scale into the Gulf. I didn't think scaling was his problem at all:
I don't think you have a scaling problem. I think you have a pricing model problem.
He was selling a safety device at a one-off price. But every popular car tracker here runs on a yearly subscription, and "until the recurring revenue element comes into this, you won't become a scalable business". The worry that people here won't pay monthly doesn't hold up either. We already pay yearly tax, membership fees and insurance.
The dairy Founder had built an ear tag for cattle and quoted it per tag. "You don't take it to per unit pricing. You take it to packages", because farms have sizes the way businesses do. And the order of the pitch matters as much as the price:
Don't mention the price till the very last. First you tell them how much money they are losing today.

"All the tech Founders, they aren't sales trained", I told him. So talk about outcomes first and the unit economics after. Keep per unit pricing out of the slide deck too.
Before any of that, check willingness to pay. It was the first thing on my list for a speech therapy software team with plenty of demos and free users. Interest isn't payment until someone pays. We wrote up how to run that test properly in how to validate a startup idea.
Trust is part of the price, too. That same team found some clinics holding back their patient data. Pakistan has no data protection law in force yet, so a clinic can't lean on the law for that reassurance. It has to come from your own written terms. And for the AI lead generation team, the trust problem is the whole category right now. TechCrunch reported that one well funded AI SDR company lost most of its early customers over unrealistic expectations. So buyers now ask for references and renewal rates. A written result from a design partner is worth more than any feature.
Then publish the price. "In this day and age it needs to be on the website, full transparency". And keep it simple, because "people make it very hard, at the start they go and build six or seven tiers. Don't do that". Use three tiers at most, and let them do different jobs. The entry and mid tiers "meet your costs", and the custom plan "is there to make profit". With each customer, you push the price up a little.
And split buyers whose loads behave differently. For the radio hardware team, "a spike type customer is a totally different game from a steady load type customer". So that's two setups and two prices, with spare units for failover on the spike side, because "the day that spike happens and your device fails, that's gonna kill your company". The industry already draws that line. It sells mission-critical and business-critical communications as separate grades. So buyers will recognize the split.
Hardware brings its own pricing problem. I put it to the floor plan projection team as plainly as I could:
Look, welcome to the world of hardware startups.
Projectors are expensive, rentals want payment upfront, and clients want credit. So the money has to come from somewhere cheaper than equity. Concessional loans come first, like the PM Youth Business and Agriculture Loan Scheme and CM Punjab's Asaan Karobar Finance. Investors charging high returns come after. Abroad, the companies doing this lease the projection setup per project instead of selling it. That eases the cash flow on both sides.
The last pricing point surprises Founders who feel guilty about services:
Product being a front door to services is actually very common. That services needs to be high quality, high price. Needs to be low effort from the Founder.
Three or four high paying services clients can pay the bills while the product does the acquisition. And the product is the part that travels to the Gulf or the West later. For the AI lead generation team it went one step further: "For you, every client is a services engagement, which eventually converts into product. You have to think like that and monetize like that and have your process like that".
After the Gala: A Card for Every Founder
Fifteen minutes goes fast, and most of what gets said in it is forgotten by dinner. So I didn't want the advice to end when the call did.
After the Gala I put together a card for every Founder I mentored, and sent it to them on LinkedIn. Each card had what was working, the issues I saw, and my advice, which is only what I actually said in the session. It also had a short block of outside research I did afterward on their market, their competitors and their pricing. A lot of the numbers in this piece came from that research. I also opened up a few free slots for NICAT fellows who wanted a longer conversation.
It's how we work on every engagement anyway: documentation that explains the why and not just the what. A decision nobody wrote down is a decision you'll have to make again.
Why the Business Behind Development Matters
Every Founder in that room could build. Very few had decided which battle to fight this year, who exactly pays, how much, and how the price compares to what the buyer loses today.
I said it to one of the Founders on day one. It's the shortest version of what we do at Biz of Dev:
I do similar work in the software field, I do Product Discovery, so what's its pitch: know what to build and build it right.
Building got cheap, and Knowing What to Build is the expensive part now. That's why we start every engagement with the business behind the Product. We break it into the same four parts every time, the Four Phases of Foundational Product Discovery. If you're new to the idea, What Is Product Discovery and Why Does It Matter? is the short version. For a Product that's already live, the same four keep running as Continuous Product Discovery. Every mistake from those two afternoons lands in one of them.

Business Understanding: Why Does This Exist, and How Does It Make Money?
This is where the business model lives.
- Saves Money or Makes Money: decide which one you sell before you write a line of copy. That choice changes the marketing, the go to market and the acquisition.
- The Pricing Model: one-off or recurring, units or packages, public or hidden, and whether anyone has actually been asked to pay.
- The One Year Battle: which business, out of the two or three on the slide, this team is actually running.
- Services as a Front Door: high price, low Founder effort, paying the bills while the Product grows.
Market Understanding: Where Does It Live, and Where Can You Win?
- Local Numbers First: with a route to the market behind every figure.
- Constraints Over "Global Solution": a niche is where a price holds.
- The Real Competitor: often the system the buyer already owns, or the free model a vendor throws in. Sometimes it's the doctor who's cheaper than the robot, or the free tool you can't compete with.
- The Comparison You Invite: VR headsets, a free tool, a multinational's brochure, so pick it yourself.
- The Nearest Next Market: the GCC before the US, with tickets bought.
User Understanding: Who Uses It, and Who Pays?
- Two Different People: the engineer loves the tool, while the person over tooling and spend signs for it.
- The Buyer's Language: PKR and payback time over the engineering metric.
- Willingness to Pay: tested with real buyers before the price is set.
- Different Behavior, Different Product: spike load and steady load, SME and enterprise.
- Do the Homework Before the Room: "when you sit down, the person across will be more of an expert than you".
Execution Plan: What Do You Build First, and How Does It Reach People?
- Enhance, Don't Replace: promise 90% of the result on what the buyer already owns.
- One or Two Modules: until one earns, then own that niche.
- One Channel You Run Yourself: for six months, plus the rooms where buyers decide.
- One, Two, Three: the first paying customer covers your cost, the second makes profit, the third scales.
- No Single Point of Failure: don't bet the business on one person, in sales or anywhere else.
If a few of these common startup mistakes sound like yours, the answers are business answers. Startup Product Discovery is where you find them. Finding them is a lot cheaper before the next build than after it. That's why what to build next is a business question first. The quickest way to see where your own Product stands is the Discovery Lens. It's 16 short steps across these same Four Phases, with a report the moment you finish, and I read every result myself.
The Cobbler's Own Shoe
The last session on day one was the AI lead generation team. By then I was tired enough to say so. We agreed to pick it up properly another day, "when I'm not this tired".
Earlier in that same session, we'd been talking about where services businesses are heading. I gave them my take, which is a personal one: "It's a dying breed, especially the ones starting right now. The future is product companies or consulting business". The services I meant are shops selling hours. That's a different business from services as a front door to a product.
And then, because it's true:
Five years ago, consulting, what was it? Everyone called me an idiot.
At the end, while they were thanking me for the ICP advice, I ended up saying something that's really about me. We'd spent half the session on their customers. I'd already asked them, "you're helping other people judge their ICP, do you know your own ICP?" So I left them with this:
Your own ICP will never fit, you'll be fitting other people's ICPs. The cobbler's own shoe is broken, remember this.
With or without my help - I wish you the best.






